<%BANNER%>

Sex and the Recently Single: Cultural Models of Mature Sexuality and HIV Risk by Physicians and Midlife Women


PAGE 1

SEX AND THE RECE NTLY SINGLE: CULTURAL MODELS OF MATURE SEXUALITY AND HIV RISK BY PHYSICIANS AND MIDLIFE WOMEN By KATHRYN ELAINE GRANT A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Kathryn Elaine Grant

PAGE 3

iii ACKNOWLEDGMENTS There are many people whom I would lik e to thank for their assistance and patience. I am profoundly gratef ul to my parents, Joe and Jean Grant, who provided the emotional and financial support that allowed me to pursue my educational goals. To my husband, Pedro Montes de Oca, I want to express my appreciations for his unending encouragement, unwavering faith in my abilitie s, and the patience to see me through this long, arduous process. I am grateful to Dr. Virginia Swartz, of the family practi ce clinic for her enthusiasm for my research project. Without her help in recruiting physicians to participate, I may have never completed the proj ect as planned. I greatly a ppreciate the directors and coordinators of the state-s ponsored displaced homemaker programs in north Florida for allowing me to visit the program meetings and to recruit my sample from among the participants. Special appreciation is rese rved for the women participants from the displaced homemaker program, who allowed me into their homes and lives to complete this project. Their interest in what I was doing was continually inspirational over the long duration of the project. I am also very thankful to all of the physicians who made time in their busy schedules to be interviewed for this project. I offer my sincerest thanks to Drs. Ja mes P. Stansbury, Leslie Sue Lieberman, Sharleen Simpson, and Barbara Zsembik of my supervisory committee. Their timely responses and instructive cri ticism kept me on track for co mpleting the degree. I thank Dr. Lieberman for including me in a res earch project on mature sexuality, which

PAGE 4

iv cultivated my career and research interest. I am grateful to Dr. Zsembik for her help with statistical analysis and for nurturing an interest in midlife women, so that my focus expanded on traditional aging research. Many thanks go to Dr. Simpson for her very timely responses and feedback during the disser tation-writing process. Finally, my most profound gratitude is reserved for Dr. Stan sbury, my supervisory committee chair, who gave me instructional advi ce, practical research o pportunities, and unwavering encouragement. I am most fortunate that Dr Bryan Page took the time to assist me early on with my proposal. I credit his guidance and that of Dr. Stansbury for my ultimate funding from the National Science Foundation an d the Woodrow Wilson Foundation. Many friends and colleagues helped me dur ing the research and writing phases of this project. I want to express my appr eciation to Dr. Kathleen Ragsdale, for her encouraging comments at difficult moments, and practical suggestions. I am also deeply appreciative to my co-worker and friend, Elsp eth Keller, who picked up the slack when I needed to be away from work for research. I thank my friend and supervisor (Dr. Mary Ann Burg, director of the Womens Health Rese arch Center at the Un iversity of Florida) for giving me research opportuni ties that enhanced the skills I needed for the dissertation, for letting me take time off to work on the doctorate and all of her helpful advice. I want to remember my dearly departed grandmother, Ethylene Dixon Grant, who helped to embolden my resolve to complete th is process. Finally, I want to thank my friend Susan Buhler, who passed away recently. In the concluding phases of my project I missed her uplifting sense of humor and her di rect way of encouraging me to always consider the utility of my research.

PAGE 5

v TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................ix LIST OF FIGURES.............................................................................................................x ABSTRACT....................................................................................................................... xi CHAPTER 1 INTRODUCTION........................................................................................................1 Objectives and Research Questions..............................................................................3 Genesis of the Project...................................................................................................4 Contributions of the Project..........................................................................................9 Obtaining Institutional Review Board Approval........................................................12 Funding.......................................................................................................................1 2 Overview of Chapters.................................................................................................12 2 LITERATURE REVIEW...........................................................................................17 Introduction.................................................................................................................17 Part One: Issues in Clinical Communi cation for Midlife and Older Patients.............19 Ageism and Health Care......................................................................................20 Gender Issues in Physician-Older Patient Communication.................................22 Women Physicians/Men Physicians....................................................................23 Gender, Ethnicity and Social Class in Medical Encounters................................25 Minority Elders and Health Communication:.....................................................26 Improving Health Communication......................................................................29 Part Two: Issues in Physician-Pati ent Communication About Sexual Health...........30 Patient Perspectives on Physicia n Communication About Sex and HIV Prevention........................................................................................................32 What Physicians Do: Discussion of sexuality and Other HIV Exposure Possibilities in Clinical Contexts.....................................................................36 Part Three: Epidemiology of HIV/AI DS Among Persons Aged 50+ and Age, Gender and Ethnicity in HIV/AIDS -Related Knowledge and Risk........................41 Surveillance Reports on HIV/AIDS Among Persons Age 50 and Over.............42 Epidemiology in Florida......................................................................................45

PAGE 6

vi Intersection of Age, Gender And Ethni city in HIV/AIDS Risk & HIV/AIDSRelated Knowledge..........................................................................................46 HIV/AIDS in older persons.................................................................................47 Midlife and Older Women and HIV/AIDS.........................................................49 Midlife and Older AfricanAmericans and HIV/AIDS.......................................52 Perception of Risk: The perils of the at-risk label...........................................56 Testing the Behavioral Change Theory for Older Adults...................................58 Targeting Older Persons with HIV/AIDS Prevention Messages........................60 3 DATA COLLECTION: THEORIES AND METHODS............................................64 Theoretical Considerations.........................................................................................64 Introduction.........................................................................................................64 Ethnography........................................................................................................65 Cognitive Theory and Methods...........................................................................66 Consensus Theory and Cultural Schemas...........................................................68 Methods and Sampling...............................................................................................72 Introduction.........................................................................................................72 Vignette-based interviews............................................................................73 Analysis of the interview and survey data...................................................74 Cultural consensus modeling techniques.....................................................75 Analysis of cultural consensus data.............................................................75 Suitability of Methodology..................................................................................77 Sampling Frame...................................................................................................78 Sampling and Recruitment..................................................................................81 Research Setting..................................................................................................82 4 CULTURAL CONSENSU S DATA AND ANALYSIS............................................90 Free listing..................................................................................................................9 1 Pile Sort Procedure.....................................................................................................94 Multidimensional Scaling of Pile Sort Data........................................................96 MDS Results........................................................................................................97 Clustering of Pile Sort Data...............................................................................103 Ratings Procedure.....................................................................................................108 Property Fitting of Ratings and Pile Sorts................................................................115 Conclusion................................................................................................................118 5 AGE, GENDER AND ETHNICITY IN CLINICAL TALK...................................120 Hypotheses................................................................................................................120 Age and Gender Constructs: Reactions to the Hysterectomy Vignette....................121 Physician Views of the Hysterectomy Vignette.......................................................126 Sensitivity to Treatment Effects on Sex-life.............................................................128 Physician Gender......................................................................................................135 Ethnicity Concordance in the Clinical Encounter....................................................141

PAGE 7

vii Physician/Patient Responsibility to Raise Sexual Health Topics.............................152 Discussion and Conclusion.......................................................................................156 6 PERCEPTIONS OF HIV RISK................................................................................161 Introduction...............................................................................................................161 HIV Vignette............................................................................................................162 DHP Womens Views of the HIV Vignette......................................................163 Physicians Views of the HIV Vignette............................................................166 Single, But Not for Long..........................................................................................167 Dating and HIV/STD Risks......................................................................................170 Marriage Doesnt Always Mean Monogamy....................................................172 The Issue of Practicing Safer Sex......................................................................174 Cultural Consensus Data Related to HIV/STD Risk................................................175 Higher Than Expected Perceptions of Risk..............................................................176 Testing for HIV/AIDS..............................................................................................177 Know or Knew Someone with HIV/AIDS and HIV Testing...................................179 Other Women in Age Group and Perception of HIV/STD Risk..............................180 Suggestions for Physicians.......................................................................................184 Physicians Perceptions of HI V Risk for Their Patients..........................................185 Racial/Ethnic Perceptions..................................................................................187 Age and Marital Issues......................................................................................189 Medical Education and HI V Prevention Readiness..................................................190 Talking About HIV Risk..........................................................................................192 Discussion and Conclusion.......................................................................................193 7 DISCUSSION AND CONCLUSION......................................................................197 Synopsis of Demographic Data and Overview of Chapters.....................................198 DHP Sample......................................................................................................198 Physician Sample...............................................................................................199 Synopsis of Findings.........................................................................................200 Cultural Models of Matu re Sexuality and HIV..........................................200 Issues in the Physician-Patient Dyad.........................................................205 Limitations................................................................................................................209 Generalizations to Larger Groups.....................................................................210 Sampling Issues.................................................................................................211 Objectives and Interview Instrument Limitations.............................................213 Limitations on Analysis and th e Problem of Too Much Data...........................213 Significance and Future Research Considerations...................................................214 APPENDIX A SHORT DEMOGRAPHIC SU RVEY FOR DHP WOMEN....................................217 B PHYSICIAN SURVEY............................................................................................219

PAGE 8

viii C VIGNETTE INTERVIEW FOR DHP WOMEN.....................................................222 D VIGNETTES AND INTERVIEW FOR PHYSICIANS..........................................226 LIST OF REFERENCES.................................................................................................230 BIOGRAPHICAL SKETCH...........................................................................................242

PAGE 9

ix LIST OF TABLES Table page 3-1 Disrupted Marital Status of State and Research Community...................................80 3-2 Mean Age of DHP Participants, Year s in Union and Household Composition.......82 3-3 Frequencies and Percentages of DHP Women Sample (n=44)................................83 3-4 Frequencies and Percentages of Physician Sample (n=31)......................................85 3-5 Mean Age of Physicians and Y ear of Medical School Completion.........................85 4-1 Free List Frequencies with Both Samples................................................................93 4-2 Terms Rating Highest Level of Agreement for DHP Women...............................112 4-3 Comparison of Ratings Keys..................................................................................112 4-4 Terms of Agreement between Physicians and DHP Women.................................113 4-5 Terms of Disagreement Between Physicians and DHP Women............................113 5-1 Cross-Tabulation of DHP Women and Physicians and Assuming Older Patients Are Sexually Active...............................................................................................132 5-2 Cross-Tabulation of Ethnicity a nd Opinions on Blacks and Whites Being Treated the Same in Clinical Settings....................................................................149 5-3 Cross-Tabulation of Physician/Patient Responsibility...........................................152 6-1 Cross-Tabulation of Age Range and Partner Status...............................................169 6-2 Cross-Tabulation of Partner Status and Personal HIV Risk Perception................171 6-3 Cross-Tabulation of Marital Status and Risk Perception for HIV/STDs...............173 6-4 Cross-Tabulation of Know/knew Someone with HIV and Ethnicity and Tested for HIV...................................................................................................................180 6-5 Cross-Tabulation of Physicia n Status and HIV Education....................................191

PAGE 10

x LIST OF FIGURES Figure page 4-1 MDS for DHP Women Pile Sorts..........................................................................100 4-2 MDS for Physician Pile Sorts.................................................................................101 4-3 Hierarchical Clus tering of DHP Women................................................................105 4-4 Hierarchical Clus tering of Physicians....................................................................106 4-5 PROFIT Analysis on DHP Wo men Pile sorts and Ratings....................................116 5-1 DHP Womens Views of the Hysterectomy Story.................................................122 5-2 Physicians Sensitive to How Medical and Drug Treatments Effect Sex-life.........129 5-3 Female and Male Doctors Talk to Their Women Patients in the Same Way.........135 5-4 DHP Womens Preferences Regard ing Physician Gender by Ethnicity................140 5-5 Physician Race/Ethnicity Affects Communication with Women Patients.............143 5-6 Physician Ethnicity Effects Communication by Respondent Ethnicity.................144 6-1 Frequencies of View of the HIV Vignette.............................................................163 6-2 Partner Status and Dating Interest (Percents).........................................................169 6-3 Frequency of Partner Stat us by Age Group and Ethnicity.....................................170 6-4 Percentage Tested for HIV by Ethni city and Age Range (Self-Reported).............178 6-5 Percentage Who Think Other Recently Single Women Know of HIV Risk.........181 6-6 Others At Risk by Ethnicity, Age Group, and Marital Status (Percents)...............182 6-7 Patients Aged 45+ Inform ed about HIV/STD risk.................................................186 6-8 Percentage of Males and Females At Risk.............................................................186 6-9 Medical Education and HIV Preparedness.............................................................191

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SEX AND THE RECE NTLY SINGLE: CULTURAL MODELS OF MATURE SEXUALITY AND HIV RISK BY PHYSICIANS AND MIDLIFE WOMEN By Kathryn Elaine Grant August 2005 Chair: James P. Stansbury Cochair: Leslie Sue Lieberman Major Department: Anthropology This dissertation explored differing cultura l models of midlife female sexuality by recently single women from a displaced homem aker program in north Florida versus primary care physicians in the same communit y. Research about clinical communication suggests that patients and physic ians may differ in their percep tions to such a degree that effective health communication is hampered. As baby boomers age, senior sexuality is becoming more socially and clinically important than it was previously. Additionally, throughout the 1990s and 2000s, the national HIV/AIDS infection rate has rema ined around 11% for those aged 50 and over, and in Florida that percentage is higher. The proportion of women infected continues to rise, particularly among African Americans. These statistica l trends are of concern for women who are re-entering the dating s cene after divorce, separation, or widowhood.

PAGE 12

xii My study considered how 44 midlife wome n conceptualize sexuality and sexual health for themselves and other women thei r age, and how 31 physicians conceptualize sexuality and sexual health for older women patients. These sexual health conceptualizations included a focus on HI V/AIDS knowledge and risk perceptions, and their ideas about discussing sexual health in clinical contexts. African American and Euro-American women between the ages of 45 and 68 completed demographic surveys, healthcare encounter vignette-based interviews, and one of three cultural consensus modeling techniques: free list, pile sort, and rating procedures. Attending and resident physicia ns, mostly from a family practice clinic completed a similar interview format. Results showed that the recently single women and the physicians shared many of the same concepts about midlife sexuality for women, with a few notable exceptions. Most of the women believed th at they were at-risk for HIV; whereas most physicians considered their younger pa tients to be at-risk and thought that risk declined with patients advanci ng age and with female status. Women believed that physicians should discuss sexua l health issues with older patients, and create a clinical atmosphere conducive to discussing sensitive topics. Physicians felt that they should do a be tter job of talking about sexual health issues with older patients, but thought th at time constraints and their own biases about aging and sexuality hindered their progress.

PAGE 13

1 CHAPTER 1 INTRODUCTION I argue that, in a basic sense, all medicine is cross-cultural -th at patient and healer inevitably conceive of the world, communicat e, and behave in ways that cannot be reasonably or safely assumed to be similar or readily compatible. Use of a common language, concepts, and behavior between patient and physician gives the appearance of shared understa ndings and effective comm unication; but appearances may be deceptive. --Robert Hahn, 1995, p. 265 My study deals with the tensi on among differing constructions of clinical talk about sexual health, and the impact of that talk on health and well-being. The quote above suggest that patients and physicia ns may differ in their per ceptions in such a way that effective health communication is hampered a nd effective health care is illusive. That is, although the physician and patient may shar e a common language, or for that matter share any other number of characteristics, th ey could understand and interpret the domain of health care in ways different enough to have an ineffectual interaction. For these actors, the cultural domain of health care ma y only be partially shared, and for patient health and well-being, these differe nces can be minor or critical. I explored the cultural domain of sexuality for midlife and older women and if that domain is culturally shared between the women themselves and with primary care physicians. I considered how midlife and older women conceptualize sexuality and sexual health for themselves and other wo men their age, how pr imary care physicians conceptualize sexuality and sexual health for midlife and older women, and how these perceptions impact the clinical encounter. Th ese sexual health conceptualizations include

PAGE 14

2 a focus on Human Immunodeficiency Virus (H IV) knowledge and risk perceptions, and their ideas about discussing sexual health in clinical contexts. In the wake of the HIV/AIDS (Acquire d Immunodeficiency Syndrome) pandemic, research on provider-patient communication about sexuality issues has taken on a new significance. There are numerous studie s with women and men of various ages, ethnicities, and sexual orientation, and with th eir health care providers, to determine how and if sexual health issues are addressed (Coa tes 1999; Epstein, Morse, Frankel, Frarey, Anderson, and Beckman 1998; Haas and Coe 1997; Makadon and Silin 1995; Metz and Seifert 1990; Kerr, Valdiserri, Loft, Breso lin, Holtgrave, Moore, MacGowan, Marder, and Rinaldi, 1996; Nusbaum, Gamble, Ski nner, and Heiman 2000; Nusbaum, Gamble and Pathman 2002; Rich 2001; Sadovsky 2000; Sk iest and Keiser 1997). Many studies have explored what physicians do in their prac tices with regard to HIV prevention as well as more general discussions of sexuality, sexual function, and sexual health. Unfortunately, despite a broa d range of partic ipants, many of these studies do not delineate their finding by age, class, ethnic ity or sexual orientation; and large-scale studies on HIV and/or sexuality communication in clinical contexts usually focus on the younger, higher-risk groups, excluding mid life and older persons altogether. Health care providers would be an id eal source of HIV prevention information since adults in the U.S. average almost thr ee visits a year to a physician, and most people believe physicians are credible sources of h ealth information (Makadon and Silin 1995). However, this resource usually goes unta pped partly because of physicians narrow notions of what constitutes appropriate medical care, their discomfort with discussing sensitive issues such as sexuality and drug us e, limited conceptions about who is at risk

PAGE 15

3 for HIV/AIDS, and time and resource constr aints (Strombeck and Levy 1998). Many providers were trained before HIV/ AIDS became widespread, and/or the sociodemographic shift in risk groups t ook hold, compounding the issues mentioned above. Additionally, as is true of the gene ral population, some physicians have ageist attitudes about older persons in terms of sexuality and other risk behaviors and do not perceive them as at risk for HIV infec tion (Strombeck and Levy 1998). However, the continuing rise of cases among those typically not considered "high risk" indicates that many physicians are (whether they know it or not) interactin g "with HIV-positive patients and other patients who are at risk for infection that may not fall into the 'traditional' high-risk categories" (Haas and Coe 1997: 403). Objectives and Research Questions My study focused on midlife and older wome n's concepts of sexuality and aging, and their notions about HIV and STD risks, comparing those concepts with those of primary care physicians in an urban north Florida community. The primary objective was to describe cultural models of matu re sexuality by analyzing semantic domains emerging from in-depth interviews with "recently single" (separated, divorced or widowed) midlife and older women (African American and Euro-American), and from interviews with primary-care physicians. The central research question is how two cultural models of mature sexuality might differ in such a way as to inhibit effective patient-physician communication. Do phys icians hold ageand gender-based assumptions about discussing sexuality and/or HIV/STD-related topics with midlife and older women? Are physicians concepts about sexuality and aging different from women's own sexuality and aging concepts? What do women think about the impact of the age, gender, and ethnic characteristics of physicians on c linical communication? Are

PAGE 16

4 there differences in women's perceptions about discussing sexuality issues in clinical contexts based on age, mari tal status and ethnic groups? Genesis of the Project My search for a research topic was not a search at all; rather it was the culmination of several research experiences, personal experiences, and a long-term interest in womens issues. I became interested in health care communication about sexuality through several experiences and activities. Firs t, the health care encounter stories from family and friends brought the struggle to be heard over the noise of medicine to my attention. Not all of the stor ies were negative; the positive encounters were a relief and surprise me and to the participants. Secondly, experience with several resear ch projects shaped my interest in gerontology and health care communication. After comple ting my masters degree, I worked as an interviewer on a project (through the University of Florida (UF), Formal Services Project of NE Florida in 1996) assessing older ad ults attitudes toward home health services. I gained e xperience using vignettes to explore respondents feelings and knowledge about available services, and what they believed to be the best path in solving difficult situations afflicting the elde rly. Working on that project cultivated my interest in the health and we ll-being of older persons. After a few years of teaching undergra duate courses in anthropology, sociology, and gender studies, I returned to UF to pur sue my Ph.D. in applied medical anthropology with a focus on gerontology. As I pursued my coursework, I had the opportunity to work as an assistant on a nursing anthropological research project w ith prostate cancer survivors ( Gender Ideology and Prostate Can cer: Veterans Cultural Model of Masculinity) that raised my concern about the lack of attention to quality-of-life issues in

PAGE 17

5 healthcare, particularly sexuality issues for older persons. That project aimed to determine the impact of male gender roles on Veterans coping skill s, health outcomes, and seeking of medical services. My interest in clinical communication about sexuality was further developed when I served as a research associate on a projec t for the Association of Reproductive Health Professionals ( Mature Sexuality: Patient Real ities and Provider Challenges) reviewing the literature on physician-pa tient communication about sexu ality. During that time I also assisted with the training of first-year medical students, and served as a standardized patient on local and national examinations th rough the Harrell Professional Development and Assessment Center, UF College of Medici ne. That experience gave me a glimpse of the rigors of medical education and the wa ys that it shapes th e perspectives of participants. Finally, through my work and studies, I al so came across the story of Jane P. Fowler, an average Midwestern woman who co ntracted HIV in her mid-50s while dating a few years after her divorce. She never t hought of herself as at-risk, but once she had started dating again, she asked he r doctor if she should be te sted for HIV and other STDs; her doctor said not you, you don't need that test (Linsk, Fowler, and Klein 2003: S249; personal communication with Ms. Fowler). Two years later, an exam for health insurance revealed that she was HIV-positive. After reading this story and corresponding with Ms. Fowler, I searched for all I coul d find on HIV prevention (clinical and public health) for midlife and older women who we re recently single due to divorce or widowhood. It was soon clear that this popula tion was not considered in most research on HIV prevention, or in clinical communica tion research, particularly about sexuality.

PAGE 18

6 Despite their growing numbers, midlife rece ntly single women seemed to be largely ignored. Thus the research topic of c linical communication about sexuality and HIV/STD risk for recently single midlife women finally melded into focus. The next question was where to find a sample of women to work with. Preliminary Research and the Research Setting Because of my interests in gerontology, I had expected to stay in Flor ida for my dissertation research, since it is demographically the oldest state in th e union. Although there are other areas with higher HIV infection rates, Fl orida ranks third highest according to the CDC. As it turned out, choosing a locale cl ose to home for me was not a stretch theoretically, since Florida has higher rates of HIV/AIDS among women, minori ties, and persons over 50 than the national average (Chapter 2). In 2002, I had the good fortune, through my work at the UF Womens Health Research Center in Gainesville, to meet th e director of the local Displaced Homemakers Program. I told her about my research inte rests and she explained that most of the program participants were midlife and ol der women who were divorced or widowed. This was my window of opportunity, and provided my first entre into the field. To gain a better understanding of both the Displaced Homemakers Program and its participants, I interviewed the director and her assistant a bout the program, and researched the program through the state website. According to State of Florida documents, The Displaced Homemaker Program was esta blished by the Florida Legislature in 1976. . The Displaced Homemaker Program is a state funded program designed to address the job training and educational needs of Florida citiz ens who: are thirtyfive years of age or older; have work ed in the home providing unpaid household services for family members; are not ad equately employed; likely will have difficulty securing adequate employment; a nd, have been dependent on the income

PAGE 19

7 of another family member but is no longe r supported by such income, or have been dependent on federal assistan ce. (State of Florida 2003) The programs goal is to help participants attain independence and economic security. The program is administered through the Division of Community Colleges. There are rarely any male participants in the pr ogram, and most of the women are divorced, separated, or widowed. In order to be sure that this was a re asonable sample for what I wanted to do, I conducted a preliminary pilot project with 18 women in a Displaced Homemakers Program. I gathered free-lists from several women in the Gainesville program to test the free list elicitation procedure a nd to explore the initial para meters of the cultural domain of aging and sexuality. For th is pilot project, a very short anonymous survey was used to gather demographic information and free lis t data. The survey was handed out at a support group meeting and later re turned to me. Participants were told that the survey was voluntary and that the data would neve r be published in any way. Using these surveys, I obtained information about the ag e, marital status and living arrangements, health status, educational level, and recentn ess of doctor visits of the respondents. The free lists on the surveys cons isted of 4 to 10 words or phrases (and in few cases, a paragraph or two) about how they define aging and sexuality for midlife and older women. Although there was not enough complete data for formal analysis, the results of that small pilot study showed the continued importance of sexuality for most of the women, particularly their desire to be consid ered sexual beings rega rdless of their age and marital status. Several women were se xually active or had r ecently been sexually active, though many lamented the difficulty of finding someone to share intimacy with.

PAGE 20

8 This pilot study also indicated that there was a cultural domain that needed to be explored further. The director of the program in Gainesvill e introduced me to a nearby program and the people involved there. This other north Florida program served a larger population with more minority women, which was ideal in providing a more diverse sample. The director of the nearby program was not as inte rested in my research topic as the other director who had helped me initially, but she finally agreed to allow me to visit the program from which I eventually recruited the recently single women of my sample (Chapter 3). Once I had the oppor tunity to interview the dire ctor and her assistant from this program, I understood a bit more about he r reluctance to participate. She explained that the program used to focus more on the whole person, but that the state now wanted a stronger employment focus. She also found that in years past, the health-related topics were not well received by the participants and had been removed from the programs content. She seemed to feel that the women would not be interested in my health related project, but would allow me to give it a try. Fortunately, many women in the program found the topic of interest a nd wanted to participate. The director interviews generated some basic information about how the displaced homemaker programs operate and some informal demographic information about the participants. Although demographics variy from year to year, most of the women are in their 40s and 50s with a range from 35 to about 75 years old; a nd about 65-70% identify as White, 20-25% identify as Black, under 10% identify as being a member of a Latino subgroup, and the remaining 1 to 3% identify as one of the Asian nationalities, American Indian, and other. Most are either divorced or widowed; fe wer are separated or married

PAGE 21

9 with a recently disabled spouse. Both direct ors defined their participants similarly in terms of why the women choose to join the pr ogram and the benefits of the program. The vast majority of the women come because th ey need to work and need to develop the skills and support to help them find work. Ju st under half continue in the support-group programs year after y ear or sporadically. Once the site had been established for the women in my sample, I contacted the Research Coordinator in the department of a Family Practice Residency program (a friend of the director of the center where I wo rked) in the area to determine the protocols for working with the residents and the atte nding physicians in the department. We met for lunch and I explained the project to her, and we discussed how she could help. The research coordinator was extremely helpful, and eagerly introduced me to all of the physicians she could, each time I came by. I was very fortunate to find someone as helpful as she, and her utility became even clearer as I attempted to contact physicians in other departments. My success with recrui ting other physicians was poor, and I believe that the lack of such a contact person as I had with the family medicine program was the main reason. Although the few physicians that I managed to recruit outside of the Family Practice program told me of the importance of the issue and their delight at seeing someone doing this kind of research, the la ck of a contact person to serve as a go-between for the physicians and myself si gnificantly hindered my ability to recruit doctors from other departments. Contributions of the Project My research project has a pplied contributions that can benefit physician-patient communication about sexual health concerns, an d public health initiatives directed at midlife and older women concerning sexua l health with culturally appropriate

PAGE 22

10 educational campaigns. This project is impor tant because it will inform improved health care communication about sexuality issues, as well as clinical HIV/STD prevention efforts with midlife and older women by disse mination of results to the research and clinical communities through medical and social science publications and presentations. Additionally, in the future I plan to use th e results to develop physician-training modules on communicating about mature sexualit y. My study explicitly focused on underrepresented groups because of its rese arch/sampling location at a community college, its understudied sample of divor ced, separated, and widowed women age 45 and over, and its balanced sampling of majority and minority groups. As a means of "giving something back" to the research participan ts (and to those who participate in the Displaced Homemakers Program, but who did not participate in the project) I will give post-project presentations of th e results. I also plan to de velop tip-cards that women can keep in their wallets with respondent reco mmendations on how to get the most out of their doctor-office visits. My project also contributes to the la rger body of scientific knowledge. An ethnographic investigation with a small samp le provides an opportuni ty to explore the nuances alluded to in larger studies on physician-patient comm unication about sexual health topics and HIV risk perceptions. My project builds on the work of Weller and Baer (2001), Chavez and colleagues (2001) and Trotter and colleagues (1999) by exploring between-group and w ithin-group variation in health beliefs. In addition to specifically comparing mature women with physic ians in terms of their ideas about aging and sexuality, and HIV/STD risks using cult ural consensus methods, the systematic cultural consensus data is woven with the interview and survey data to explore the

PAGE 23

11 cultural and structural factors, and the issues related to interpersonal dynamics of the physician-patient dyad. This s ituates my micro-level expl oration within macro-level phenomena, including the hierar chical nature of the medical interview and the hegemony of the biomedical model of disease that of ten structures healthcar e interactions (Baer, Singer and Susser 1997). Applied medical anthropology and other soci al science endeavors in the medical arena can serve to highlight social and cultu ral factors that affect health care outcomes and to incorporate better unde rstandings of these factors in disease prevention. The anthropological tradition of coll ecting detailed data is well suited to situating micro-level events within their larger macro-level determin ants. Epidemiologists and social scientists have shown repeatedly that th e biological and social aspect s of health are interrelated (Hahn 1995; McElroy and Townsend 1989). Un fortunately, most medical education minimizes the importance of social and cultu ral factors in health care (Good and Good 2000; Hahn 1995). The body of research that specifically addresses physician-patient communication about sexual health concerns including HIV risk, has been steadily growing, yet many questions remain unanswered. My study examined how physicians conceptualize sexuality for midlife and older women re lative to how women conceptualize it for themselves, and how those concepts may affect health care interaction with regard to HIV/STD prevention. Despite the growing ev idence that midlife and older persons are continuing to engage sexually (some in a high-risk manner), many physicians do not address sexual health issues with these pa tients. My study considered this issue and

PAGE 24

12 related issues in an effort to raise awaren ess about physician biases and the social and cultural factors that contribute to them. Obtaining Institutional Review Board Approval Instruments and forms for my project were approved by the UF Institutional Review Board (Protocol# 2002-U-721) September 6, 2002. In compliance with standard Institutional Review Board (IRB) regula tions, these materials were resubmitted on August 15, 2003 to extend the approval time pe riod for data collec tion until August 25, 2004. Participation was voluntary and res pondents were compensated for their time (Chapter 3). Funding I am very grateful for the funding to conduc t the fieldwork for my research project. In February 2003, I received notification that I had been awarded two grants. Funding was provided by a dissertation improvement grant from the National Science Foundation (BCS-0241675) and a Woodrow Wilson/Johns on & Johnson Dissertation Grant in Womens Health. Overview of Chapters The following pages provide an outline of the chapters in this dissertation. Chapter 2 is a literature review that focuses on i ssues in physician-patient communication about sexual health. The first secti on highlights some of the stru ctural, cultural, and economic issues that broadly effect health care co mmunication, focusing on the research studies and reviews that are most relevant for this dissertation research pr oject. Following this introduction to health care communication issu es, I focus on issues in physician-patient communication about sexual health. This incl udes a description of research findings on the role of patient-physician communication in HIV prevention. I reveal the disconnect

PAGE 25

13 between patient and physician perspectives on clinical communicat ion about sex, aging and sexuality, and HIV/STD prevention, particul arly as it relates to midlife and older persons. Next, the relevance of this research project is clarified by situating it in the epidemiology of HIV/AIDS among persons aged 50 and older in the U.S., in Florida, and finally in Northeast Florida where the research takes place. I further show the projects relevance by exploring age, gender and et hnic trends in HIV/AIDS transmission and HIV/AIDS-related knowledge. Finally, chapter two concludes with a brief discussion of the primary prevention strategies that shape the nations public health initiatives and how primary prevention models relate to older pe rsons. My premise is that ethnographic research that looks at intraand intercultural variation in health-related perspectives can contribute to the development of more e ffective health care communication and HIV prevention strategies. Chapter 3 describes the methods used in th is research. A brief introduction outlines the methods used and is followed by a more detailed discussion. The explanation of methods continually links back to the prev ious chapter by showing how the goals and design of the research comes out of the que stions raised by previous research and epidemiological findings. I give a descri ption of Cultural Cons ensus Modeling (CCM) and health beliefs and then show support for using CCM in combina tion with traditional qualitative data collection. The next section describes the surveys and interviews instruments and their strengths and weakne sses in terms of addr essing the research questions in chapter one. I give a synopsis of the vignette-based interview and refer the reader to the appendices for more detail, a nd then discuss the CCM procedures and how they link with the other data collection methods. The sampli ng and the interview settings

PAGE 26

14 are explained in this chapter. I explain how participants were recruited for the project, and how the selection of this sample is s upported by the epidemiological data. Finally, I give a brief introduction to the final samples and their key characteristics. Chapters 4, 5 and 6 present the research fi ndings. Chapter 4 presents part one of the data analysis, focusing on sexuality concep tualizations. I provide a description of the types of analysis used, including software progr ams, and an explanation of the theoretical considerations for using th ese techniques. This chapter focuses on the Cultural Consensus Modeling data (CCM), presenting the data from the free li sts, pile-sorts and the rating test with the women in the Di splaced Homemaker Program (DHP) and the primary care physicians. These data are presen ted textually and graphically for their own merits and are linked to the data from the in -depth interviews in the following chapters. The interview data contextualizes the CCM da ta and shows how it is linked to the other issues in the research project, discussed in chapters five and six. The data on sexuality conceptualizations are consider ed in terms of their support of the hypotheses, as well as their shortcomings for support of the hypotheses. Chapter 5 continues with pa rt two of the data analysis, focusing on physicianpatient communication regardi ng sexual health issues. After a brief introduction describing the various issues addressed and hypotheses, I in troduce the data gathered on physician-patient communication. This chapte r contrasts the views of women in the DHP program and the views of primary care physicians This section starts with a focus on the responses to the first vignette and point out the differences and similarities between the two groups. This follow by gender-, ageand et hnicity related issues in physician-patient

PAGE 27

15 communication about sexual health:. I conc lude with an overview of the findings presented in this chapter, their support of the hypotheses or lack of support. Chapter 6 concludes the data analysis chap ters with part three, which focuses on HIV risk and prevention issues. After an intr oduction to the topic of HIV risks, I present the data on HIV related knowle dge from women in DHP progr am. I introduce some of the issues raised with the HIV vignette, and then follow it with their comments on personal risk perceptions and their perceptions of the risk for other women in their age group and partner status. Next I present the data on HIV risk per ceptions from primary care physicians regarding their patient popula tion. The risk for patients is further delineated by age, sex, ethnicity and then ot her various possible variables suggested by me or the physicians, such as drug use, se xual orientation, sex wo rk, marital status, number of partners, and so on. This chapte r concludes with a co mparison of the views the women in the DHP program and the physic ians, linking these findings with the CCM data. Chapter 7 concludes the disse rtation with an overview of the findings discussed in previous chapters and a discussion of how the findings square with the literature in chapter 2. The next section unites the CCM findings with the vignette-based interview, showing the utility of combined methods and links the findings about HIV risk perceptions to other issues cl inical communication about sexu al health. Limitations of the research project are discu ssed on several levels: 1) some of the project goals were not achieved due sampling issues; 2) some of the data collection methods limit the generalizeablity of the results; 3) not all of the objective were addressed with the interview instrument, and 4) time and space limitations on analysis of all the data

PAGE 28

16 gathered. This project yielded a rich dataset despite its limitati ons; so rich th at not all of the data could be analyzed at this time. Th is chapter concludes with a discussion of the significance of the data and its utility for improved health care communication regarding sexual health issues and HIV prevention in clinic al context. Finally, th e utility of the data is situated within the framework of anthropol ogical methods and theory and how they are relevant to health care an d health care communication.

PAGE 29

17 CHAPTER 2 LITERATURE REVIEW Introduction The end of womens reproductive lives and facile assumptions about their difficulties in finding partners frequently lead some health care providers to the mistaken conclusion that sexual issues are unimportant for midlife and older female patients, (particularly for those who are no longer married due to divorce, separation, or widowhood). However, evidence suggests th at many mature women and men remain sexually active, some into their 80s and beyond, that some have multiple partners, and practice a full range of sexual behaviors (A merican Association of Retired Persons (AARP) 1999, 2003; Association of Reproductiv e Health Professiona ls (ARHP) 2000; Jacoby 1999; Janus and Janus 1993; Levy 1994; Levy and Albrecht 1989; National Council on Aging (NCOA) 1998; Stall and Cata nia 1994). As baby boomers age, senior sexuality is expected to become socially and clinically more important than it was previously perceived to be (ARHP 2000; Janus and Janus 19 93; Yates, Stellato, Johannes and Avis 1999). Despite this projected tre nd, in American society sexuality is often viewed as a concern of the young. This cult ural emphasis on sexuality in youth, and the concomitant implication that older persons s hould not be concerned a bout sexuality or the potential health risks associat ed with some sexual behaviors, is poor preparation for agerelated changes in sexuality (Hotvedt 1983). Furthermore, it does not prepare older persons or their health care providers for c linical communication abou t sexual issues.

PAGE 30

18 This literature review focuses on issues in physician-patient communication about sexual health. The first secti on highlights some of the stru ctural, cultural, and economic issues that broadly effect health care co mmunication, focusing on the research studies and reviews that are most relevant for this dissertation research pr oject. Following this introduction to health care communication i ssues, I focus on studies that reveal the disconnect between patient and physician pers pectives on clinical communication about sex, aging and sexuality, and HIV/STD prevention, particularly as it relates to midlife and older persons. Next, the relevance of this rese arch project is clarified by situating it in the epidemiology of HIV/AIDS among persons aged 50 and older in the U.S., in Florida, and finally in Northeast Florida where the research takes place. I further show the projects relevance by exploring age, gender and et hnic trends in HIV/AIDS transmission and HIV/AIDS-related knowledge. Fina lly, this chapter concludes with a brief discussion of the primary prevention strategies that shape the nations public health initiatives and how primary prevention models relate to older pe rsons. My premise is that ethnographic research that looks at intraand intercultural variation in health-related perspectives can contribute to the development of more e ffective health care communication and HIV prevention strategies. My study focuses on midlife and older wo men, but it should be noted that the literature concerning aging and health care, aging and HIV related issues, and aging and sexuality issues is not consistent in how it frames the categories of midlife, older, and elder. Age ranges for what is considered midlife and what is considered older or elderly vary widely from one study to th e next, making comparison between studies

PAGE 31

19 difficult. In this literature review, whenever possible, I clarify th e age groups considered in a given study. Part One: Issues in Clinical Communi cation for Midlife and Older Patients Physician-patient interaction is shaped a nd influenced by cultural, structural and institutional factors and th e interpersonal dynamics of the physician-patient dyad. Physicians communicate with and react to patien ts differently on the basis of the patients age, gender, ethnicity, educational level, and social class. By the same token, patients also communicate with physicians differen tly based on a similar set of factors. Additionally, many of the difficulti es patients face in their experiences of health care are related to some aspect of health care systems, which are becoming increasingly competitive and resorting to impersonal and routinized care to remain institutionally viable (Coulter and Fitzpatrick 2000). Interactions between physici ans and patients are often hi erarchical as well as intercultural. These interactions may be construed as hierarchical because of the hegemony of the biomedical model of dis ease (Baer, Singer and Susser 1997) and the structural power dynamics of modern medicine (Baer et al. 1997; Ca nt and Sharma 2000; Henley 1977; Lorber 1997); and as interc ultural because of sociodemographic and structural differences affecting the lived experiences of the participants (Cline and McKenzie 1998). The hierarchi cal and intercultural nature of the inte raction is carried out in the medical discourse where patients struggle to be heard over the noise of medicine (Elgin 1999). Patients enter medical in teractions from a position of relative weakness . They enter unfamiliar surroundings in which all the participants seem to share a common language . Me dical practitioners, on the other hand, are in their home court in the medical se tting (Fisher 1993: 168).

PAGE 32

20 Doctors have greater power in the medical en counter by virtue of the fact that they control access to health resources as well as the dynamics of the health care interaction, and that control is institutionally regulate d and sanctioned (Baer et al. 1997; Cant and Sharma 2000; Henley 1977; Lo rber 1997). However, recogn izing the asymmetry of the relationship does not mean that patients are pa ssive objects in the ex change; rather they respond within the possibilities of the existing social structure by reac ting to or resisting medical directives and procedures (Baer et al. 1997; Lazarus 1988). Additionally, the increasing medical pluralism of U.S. society offers alte rnatives to the dominant biomedical paradigm that sometimes serv e to empower patients who may rely on multiple medical systems at one time (Cant and Sharma 2000). There are a host of communication challenge s in the interaction between midlife and older patients and their health care providers. Ageist, sexist, and racist attitudes and behaviors on the part of providers an d patients, age-associated physiologic communication difficulties, generational, ethnic and gendered differences in communication patterns, and the influence of a th ird party all relate to the various layers of difference that may influence physician-pa tient interactions (A dler 1998; Cline and McKenzie 1998; Damron-Rodriguez 1998; Jack son and George 1998; Ikels 1998; Lorber 1997; Ryan and Butler 1996). The complexity of the interaction is of ten greater for older than for younger patients because they frequently present multiple medical problems that are mostly chronic, incurable, and thus cont rary to the treatment paradigm in modern medical training (Adelman et al. 1991; Hahn 1995). Ageism and Health Care Ageism results in stereotypes of older individuals as a homogenous group with declining productivity, healt h, independence, sexuality, memory, and communication

PAGE 33

21 skills (Ryan and Butler 1996). The ageist at titudes of care providers are complex and reflect a societal preoccupati on with productivity and youth th at effectively stigmatizes aging and illness (Kreps 1990). Ageist attit udes arise from fears about aging and death (Ryan and Butler 1996; Nussbaum, Thompson and Robinson 1989) and even the patients own ageist assumptions and beha viors (Adelman et al. 1991; Coupland and Coupland 1999; Ryan and Butler 1996). Ageist be haviors arise in clinic al settings partly because providers there are more likely to encounter older adults who are ill, frail, confused, and hospitalized rather than vigorous healthy older individu als (Adelman et al. 1991). Ageism among providers also results in attributing disease processes to normal aging, leading to inappropriate medical tr eatment (Chiang 1998; Davidhizar and GigerNewman 1996; Ryan and Butler 1996). The problems of ageism are compounded by age-related changes in communication skills, such as hearing and vi sion impairments, changes in th e ability to process various types of information, and memory impairme nts (Ryan and Butler, 1996). Some hearing loss is common for 50% of adu lts over age 65, and some degree of visual impairment is also common (Adelman et al. 1991). The actua l incidence of dementia increases with age from about 2.5% for people in their sixties, to about 22% for t hose in their eighties (Adelman et al. 1991). These sensory a nd cognitive difficulties have important implications for communication and can im pose serious limitations on health related communication for otherwise healthy adults. Because these changes are socially stigmatized and seen as symbols of a gene ralized decline, they are also sometimes inappropriately assumed to be indicator of the patients capacities in other unrelated areas

PAGE 34

22 (Adelman et al. 1991). Adding to the comple xity of ageist perceptions and physiologic changes are gender stereotypes, a nd racial and ethnic stereotypes. Gender Issues in Physician-Ol der Patient Communication Gender issues in physician-patient interac tions are of tremendous importance for the aging patient due to the age differentials in longevity. Consequently the problems of aging are largely the problems of women (A llman et al. 1999). Women of all ages generally visit the doctors office more often than men for multiple reasons relating to health and social roles (Jenks 1993). A CDC report on womens use of ambulatory medical care in 1997 and 1998 found that even af ter excluding pregna ncy-related visits, regardless of race or ethnicity "women were 33 percent more likely than men to visit a doctor, although this difference decreased with age. The rate of doctor visits for such reasons as annual examinations and preven tive services was 100 percent higher for women than for men and medication patterns differed significantly (Brett and Burt 2001: 12). Women often see doctors in the pr ocess of caring for children, husbands, and parents as non-patient clients because of thei r social role of family caretaker (Jenks 1993). Several studies and reviews suggest that both physicians and patients are influenced by gender stereotypes in the clinical enc ounter (Cline and McKenzie 1998; Hall and Roter 2002; Haug 1996; Jenks 1993; Street 2002). The gender of the physician relates to how they interact with patients and to how patients perceive them, and the gender of the patient impacts how physicians relate to them as well (Hall and Roter 2002; Street 2002). West (1984) found that doctors consistent ly underestimate the medical knowledge of their patients, and tend to talk "down" to th eir patients. They frequently underestimate the abilities of patients w hose ethnicity, gender, or so cioeconomic status differs

PAGE 35

23 significantly from that of mo st physicians. Many physicians expect women to be difficult and demanding, and to have more psychosomatic illness episodes than men (Cline and McKenzie 1998; Haug 1996; Jenks 1993). Women frequently receive less thorough diagnostic work-ups and less aggressive trea tment than men, and women typically also receive less technical/medi cal information (Adler 1998; Cline and McKenzie 1998; Lorber 1997). Female patients often must struggle to be considered competent individuals when interacting with male physicians, since th eir concerns are often taken less seriously than t hose of male patients (Cline and McKenzie 1998). Women spend more time asking questions of their doctors (Jenks 1993), but the amount of time women spend asking questions is "disproportionate to the information yielded" (Cline and McKenzie 1998: 64). Women Physicians/Men Physicians The vast majority of studies on the influence of the physician's gender on interaction patterns in the me dical interview indicate that the gender of physicians and patients make a difference in medical practi ce (Adelman et al. 1991; Cline and McKenzie 1998; Hall and Roter 2002; Haug 1996; Lorber 1997; Roter and Hall 1998; Street 2002; West 1984). In most studies that consider the influence of the sex of the physician, female physicians appear more apt to provide support and informati on to their patients, are more egalitarian, and give more time to consultation than th eir male counterparts (Adelman et al. 1991; Haug 1996; Lorber 1997; Ro ter and Hall 1998). They also tend to interrupt their patients less than male physicia ns, and show more inte rest in what their patients are trying to communicate (Cooper-Pat rick, Gallo, Gonzales, Ve, Powe, Nelson, and Ford 1999; Haug 1996; Lorber 1997; We st 1984). West (1984) found distinct differences in interruption frequency between in teractions with female patients and male

PAGE 36

24 patients, as well as differences between ma le physicians and female physicians. Male physicians initiated 67% of interruptions, wh ere as patients initiated 33%. When the gender and race of the patient was thrown into the equation the asymmetries were magnified. Women physicians spend more time with patients and engage in more positive communication practices, such as agreement, approval and laughter, and are more likely to build partnerships with their patients by asking for their opinion, and expressing understanding (Lorber 1997). This has led to a preference for women physicians in primary care, as well as the traditional primary care specialties of obstetrics and gynecology, and pediatrics, where direct hands-on, first-line care is needed. Although the research indicates that communicative differences do exist between female and male physicians, the difference s are typically not momentous, and on the whole women and men physicians are more alik e than they are different (Roter and Hall 1998; Street 2002). These gender differences in communication styles are most likely an effect of life-long socialization and intera ction patterns (Lorber 1997). Physician and patient gender stereotypes a nd attitudes likely generate a priori assumptions about their capacities and desires for both gender conc ordance and discordance in clinical communication (Street 2002). For example, people generally believe that female physicians are less likely use aggressi ve communication strategies (commands, directives, negative opinions, controlling beha viors) than are male physicians (Street 2002: 204). While women physicians also do not tend to interr upt their patients as often as their male counterparts, women physicia ns are interrupted more often than men physicians by their patients (Hall and Rote r 2002; West 1984). Stereotypes and beliefs

PAGE 37

25 about women and men physicians, combined with the assumption that women are more interested in psychosocial aspects of hea lth (Roter and Hall 1998), could explain why patients tend to talk more, divulge more emotional and relational information, ask more questions, and are more involved in health care decisions when in teracting with women physicians (Street 2002). Some of the problems associated with fre quent interruptions by physicians in the medical interview are related to contempor ary time constraints imposed by HMOs and hospital administrators (Lazarus 1988; Paget 19 93), but also because of the standard way that physicians shape the meaning of illne ss for their patients through their use of question and answer practices (Clark 1996; Good and Good 2000; Paget 1993). Importantly, in the traditional clinical encoun ter the representation of the reality of the patients condition is (re)defi ned in the cognitive structur es of the discipline of medicinea reality that may be very di fferent from that understood by the patient (Clark 1996: 751). This hierarchical exchange is in part developed in medical schools where students learn to reconstr uct the clinical narrative in the context of diagnostic and therapeutic procedures (Good and Good 2000). Clark (1996) suggests that as the patient tells his or her story to the physicia n, the physician begins to manipulate the information conveyed by interrupting to pose strategic questio ns that shape the information to conform to the format of the medical report. These structural and institutional impediments to positive and empa thetic health care communication are often magnified by racial/ethnic and socioec onomic differences of the participants. Gender, Ethnicity and Social Class in Medical Encounters The issue of social class intersects with age, ethnicity and gender, because minority/ethnic groups are more likely than the white majority to be in poverty, women

PAGE 38

26 are more likely to be poor than men, a nd older ethnic/minority women, are the most impoverished of all (U.S. Bureau of the Census 1995). More older women than men are Medicare and Medicaid benefi ciaries, and since Medicare pays for only 80 percent of physicians services, physicians do not have a financial incentive for spending optimal time and delivering optimal careemploying optim al communication with this class of patients (Allman et al. 1999: 334); and not al l physicians accept Medicare payments effectively trapping some women in an unproductive physician-patient relationship. A study of 20 female patients age 65 a nd older found that women who had only Medicare coverage had the per ception that their phys icians did not really care about them as individuals (Allman et al. 1999). Ad lers study (1998) of an ethnically and economically diverse group of ol der women with breast cancer, found that some women tended to behave more passively with their physicians because they lacked other health care options due financial cons traints and poorer health lead ing to a sense of forced reliance on their physician. Some women may no t attempt to assert themselves in the medical encounter and may feel trapped due to various social, si tuational and economic reasons. These issues of interpersonal dynamics and economics are perhaps most troublesome for ethnic/minority groups. Minority Elders and Health Communication: Not only is the elderly p opulation growing, the older ethnic/minority population is growing faster than the majority elderl y population (Damron-R odriguez 1998; Jackson and George 1998; Tripp-Reimer 1999) and will likely reach 25% of the citizens aged 65 and older by the year 2030 (Davidhizar and Giger-Newman 1996). Recent studies show that for patients with similar clinical fi ndings there are differences by ethnicity in treatment, amount of information provided, a nd empathy displayed by physicians (Adler

PAGE 39

27 1998; Cline and McKenzie 1998; Cooper-Patrick et al. 1999; Schneider, Zaslavsky and Epstein 2002; Woo et al. 2004). Jackson a nd George (1998) argue that Blacks, women and members of lower social classes are ofte n assigned lower social worth by health care professionals, and receive the worst care in hospitals. As a result, more impoverished groups, including the elderly, often delay seekin g treatment to avoid negative appraisals by health care providers (Jack son and George 1998, Lazarus 1988). Cooper-Patrick et al. (1999), in their c onsideration of race and gender concordance and discordance and its impact on patients perceptions of physicians participatory decision-making (PDM) styles, found that patie nts aged 40-65 years ra ted their visits as more participatory than patients younger than 30 years. However, African Americans and other minorities rated their physicians as having lower PDM scores than did white patients. Although the differences were lessened when there was race and gender concordance between physician and patient, Afri can Americans still rated their visits as less participatory than whites (Cooper-Patrick et al. 1999). Even after controlling for SES African American and other minority patien ts had less participatory visits with white physicians. Another more recent study confir ms the above findings. In a cohort study using pre-visit and post-visit surveys and audi otape analysis, race-c oncordant visits tend to be longer and were rated as having more positive affect (Cooper, Roter, Johnson, Ford, Steinwachs and Powe 2003). However, race/ethn icity is not the only factor affecting the quality of physician-patient inte ractions. One nationally repr esentative survey that oversampled for Black, Hispanic and Asian househol ds found that the quali ty of interactions was lower for non-White patients, particularly Hispanics and Asians, even after controlling for differences between respondents with and without health insurance (Saha,

PAGE 40

28 Arbelaez and Cooper 2003). Yet race/ethnicity discordance contributed minimally to the quality of these interactions, and was better explained by ph ysicians behaviors such as treating patients with greater respect, spe nding more time with patients, listening and involving patients in the d ecision-making process. Although communication may be enhanced when the physician and patient belong to the same ethnic group (or for that matter the same gender, or age group, or social class), this does not insure that both will ha ve the same explanatory models of illness or the same expectations for the medical vis it (Cooper-Patrick et al 1999; Ikels 1998). A study by Green et al. (1994) of older patient satisfaction shows the interrelatedness of SES, racism and ageism in health encounters. In this sample of mostly poor, non-white women patientsthere may be little expectation that the young wh ite physician working in th e highly respected tertiary care medical center study site would shar e in decision-making. These data should not be construed to mean that older patient s do not desire to participate in decisionmaking. Rather, these data may reflect the past experiences of these older patients with other physicians and their percep tions of what they can expect from physicians in this setting. (Greene et al. 1994a: 1284) Therefore, differential interaction patterns between physicians and their patients are not only shaped by cultural, generational and gende r differences; they are also constructed out of the experience and histor y of social inequality. Although different genders, age groups and et hnic groups vary in their expectations regarding medical encounters, researchers s hould be cautious in making cultural value generalizations that do not consider how the experience of social inequality shaped those values. The health care system is itself cu lturally constructed and functions within the parameters of the larger society. Placing the h ealth care transaction into its larger cultural context clarifies the social processes th at shape the meaning and outcome of the encounter.

PAGE 41

29 Improving Health Communication Effective (and affective ) communication is key to all human relationships and yet it is frequently not achieved in health care relationships. Despite good intentions, doctors often do not listen to or u nderstand their patients con cerns, and therefore, do not incorporate those concerns in to the healing process. Re garding the physician-patient transaction as hierarchical and intercultu ral highlights the social and structural distinctions between individuals and groups within the larger society. Problems in the doctor-patient relationship have to be understood as a reflection of the power relations of the larger society (Rhodes 1996). Regardless of ethical pronouncements by the American Medical Association to treat all of humanity without dist inction by race, gender, age, religion, socioeconomic status, or sickness c ondition, there is evidence of substantial institutionalized inequity (Hahn 1995: 143). Health communication scholars, lin guists and medical sociologists and anthropologists have suggested a multitude of ideas and techniques to improve health care communication and ultimately health care outcomes. The approach that appears to incorporate most of these suggestions is the shift from the traditional biomedical problem-based approach to a goal-oriented holistic approach (Clark 1996: 762). In the goal-based model, health is defined not as an outcome, but as a process, incorporating the values of the individualas embodied in his or her personal and life goals . which can best be determined through a dialogue between patient and pr oviders (Clark 1996: 763). This processual approach to health a nd health care empowers patients to assume greater responsibility for thei r health, and ideally allows both provider and patient to transcend some of the barriers of difference that separate th em (Clark 1996). The key in this approach is listening. Medical profe ssionals are often not taught how to listen and

PAGE 42

30 believe that it amounts to something a kin to good manners or note ta king (Elgin 1999). Elgin, a linguist, argues that the crisis in health care is a crisis of language, because of misunderstandings, a failure to listen and believe the stories of patients, and not communicating effectively and empathetically. Kleinman suggest, If we think of the clinical encounter as a trans action between patient and practi tioner conceptualizations of sickness and clinical reality, then a core c linical task becomes making these explanatory models explicit and thereafter translati ng and negotiating between them (1979: 61). However, Kleinman cautions that this appro ach to health care may require practitioners to recognize that biomedical conceptua lizations may be inappropriate in some circumstances or require adjustments to fit w ith the patient's experi ence. Thus culturally competent care is not about memorizing the at tributes of different ethnic and cultural groups, but about learning to use a patient -centered approach that incorporates fundamental skills and attitudes that may be applicable across ethnic boundaries (Saha et al. 2003:1718). These skills are crucial to improving health care communication about sexual health issues, which the following s ection notes is poorly attended to for older persons. Part Two: Issues in Physician-Patien t Communication About Sexual Health An interest in exploring the dynamics of provider-patient communication about sexuality issues has taken on new significance in the wake of the AIDS pandemic. This interest is manifest in numerous studies w ith women and men of all ages, and with their health care providers, to determine how and if they address topics that are sometimes difficult to articulate, embarrassing, and even stigmatizing (Epstein et al. 1998; Gerbert, Maguire, and Coates 1990; Haas and Coe 1997; Makadon and Silin 1995; Metz and Seifert 1990; Kerr et al., 1996; Nusbaum, Gamb le, Skinner, and Heiman 2000; Nusbaum,

PAGE 43

31 Gamble and Pathman 2002; Skiest and Keiser, 1997). These studies have explored what physicians do in their practices with regard to HIV prevention as well as more general discussions of sexuality, sexual function and sexu al health (Epstein, et al. 1998; Haas and Coe 1997; Frank and Harvey 1996; Kerr, et al 1996; Makadon and Silin 1995; Penson, et al. 2000; Ross, et al. 2000). Health care practitioners have the potential to be an excellent source of HIV prevention information for a broad spectrum of persons since adults in the US average almost three visits a year to a physician (Makadon and Silin 1995). However, this is seldom the case. "In spite of the recognize need for provider involvement in prevention education, many obstacles exist, including clin icians' narrow conceptions of medical care, discomfort with discussing sexuality and dr ug use, attitudes about HIV/AIDS, and time and resource constraints" (Strombeck and Levy 1998: 918). Many physicians simply lack good interviewing skills, especially when dealing with the older patients (Putnam 1996). Furthermore, many providers practicing today were trained before the HIV/AIDS epidemic began, and thus lack sufficient know ledge and skills to address HIV risks. Additionally, as is true of the general populat ion, some physicians ha ve ageist attitudes and assume that those older patients are not at risk for HIV infection (Strombeck and Levy 1998: 918). However, the continuing rise of cases among those typically not considered "high risk", indicat es that many primary care physic ians are, whether they are aware of it or not, interacting "with HIV-pos itive patients and othe r patients who are at risk for infection that may not fall into the traditional' high-risk categories" (Haas and Coe 1997: 403).

PAGE 44

32 Although many health care practitioners have begun to incorporate important disease prevention issues into their practices sex and drug-related discussions are still relatively low. In a recent study, the CDC reported that the overwhelming majority of primarycare physicians 'usually' or 'always' as ked their adult patients about cigarette smoking (94%) and alcohol use (84%). In ar eas most directly related to HIV risks, physicians were more reticent, however. Many fewer inquired a bout illicit drug use (49%), condom use (31%), number of sexua l partners (22%), or sexual orientation (27%). (Strombeck and Levy 1998: 918-19) Despite some improvement in preventive medicine, physician training has traditionally given higher priority to immedi ate acute needs rather than to long-term behavioral interventions. Narrow conceptualizat ions of the physician's role tend to focus on patients' complaints rather than offering health promotion services (Makadon and Silin 1995). Patient Perspectives on Physician Comm unication About Sex and HIV Prevention The vast majority of patients surveyed and interviewed over the past decade have indicated that they would like to talk with physicians abou t sexual health concerns, but most have never done so. These studies show that most adult women and men want to discuss sexual health issues, including HIV, with their health car e providers, but would prefer that their doctor ra ise the issue (Haas and Coe 1997; Loehr Verma and Seguin 1997; Makadon and Silin 1995; Marwick 1999; Me tz and Seifert 1990; Nusbaum et al. 2002; Waterhouse 1993; Zweifler Padila and Schafer 1998). One nationwide survey of patients revealed that only 15% reported that they had discussed HIV risks with their doctors in the previous five years, even though 94% had vi sited a physician during that time. Nearly three-quarters (72%) of the disc ussions about HIV that did take place were at the patients' urging (Gerbert, Maguire, and Coates 1990). Other smaller studies

PAGE 45

33 focusing on reports from both physicians and pati ents reveal similar finding with regard to clinical communication about sexual health issues. In a nursing study, Waterhouse (1993) survey ed a convenience sample of 53 men and 20 women (age 29-58 years, mean 42 year s) who were either current or retired university employees. This study examined th e attitudes of healthy individuals toward physicians and nurses discussing sexual concerns with clients. The majority of this mostly white (96%), married (77%), and welleducated (68% with bachelor's degree or higher) sample thought discussing sexual concer ns with nurses was appropriate, and all subjects thought that physicians should discuss sexual concerns with patients ( p < .0001). Twenty-six percent felt nurses should "alw ays" or "almost always" discuss sexual concerns with their patients, and 66% re sponded "always" or "almost always" that physicians should discuss sexual concerns w ith their patients. Women showed more positive attitudes towards discussion of sexual concerns by physicians and nurses than males, but this difference was not significant (data not shown). Age, marital status, and educational level had no signifi cant influence on attitudes. Nusbaum et al. (2002) assessed women's in terest and experience with discussing sexual concerns with their docto rs, and identified what they perceived to help or hinder this discussion. The 964 women who complete d the questionnaires ranged in age from 18 to 87 years old and were seeking routine gyne cological care. Most participants (78%) wanted to discuss their sexual concerns w ith their physicians, and 97.2% felt that their physicians should be knowledgeable about sexuality. More than half (57%) had discussed some of their sexual concerns with a physician, a nd 89.3% of these reported the discussion was helpful. Another 24% had been interested in discus sing concerns but had

PAGE 46

34 not done so, and 35% reported missed opportuni ties to discuss concerns when they attempted to raise the subject by physicians did not seem in terested or to understand. Although most respondents (69.7%) reported f eeling too embarrassed to bring up the topic; these women were over 10 times more likely to respond that they would have discussed their concerns if the physician had asked compared to those not reporting embarrassment. Patients reported feeling comf ortable talking about se xual concerns with both female and male physicians, but physicia n gender was reported as effecting their willingness and ease in discussing these concerns. In a much smaller, qualitative study, Loeh r et al. (1997) inves tigated the attitudes of women over 60 toward sexuality and intim acy, including how the medical profession addresses sexuality and intimacy issues. F ourteen English-speaking women age 60 to 85 in Kingston, Ontario, Canada, from senior ap artment complexes or nursing homes, senior community organizations and volunteer groups participated in focus groups to discuss sexuality issues. On the whole, the participan ts felt that sexuality is, and will continue to be, an important part of their lives. Howeve r, like the women in the Nusbaum et al. study there was uncertainty and hesitation a bout the prospect of volunteering sexual information in clinical context. The women in this study suggested that they would feel more apt to talk about sexuali ty with their physicians if th ey were not being rushed; if they felt their complaint was important to the doctor; if doctors used layman's terms to explain things, and; if they were not addressed by terms of endearment or age-related terms like "Granny". Men's expectations about physician inte ractions mirrored those of women, particularly in the desire for physician-in itiated communication a bout sexuality. Metz

PAGE 47

35 and Seifert (1990) surveyed by mail a small sa mple of primarily white, middle-class, well-educated men (N = 62) ranging in age fr om 18 to 73 years. While 97% of the men reported prior sexual concerns, only 19% had di scussed these issues with their doctor. However, 35% had consulted one or more professionals for assistance with a sexual concern (psychologists/ psyc hiatrists 26%, primary care phys icians 19%, sex therapist 12%, marital and family therapist 8%, and th e clergy 3%). Many were hesitant about talking to their doctors, but wa nted to deal with sexual prob lems. Only 31% of the man reported that they were likely to initiate a conversation about se xual concerns with a physician. They overwhelmingly preferred that the physician initiate the discussion of sexuality issues (85%). Only 18% indicat ed they did not appreciate their physician inquiring about sexual matters. Fewer than a quarter of the men reported that their physicians had inquired about sexual matters, and only a quarter of the patients have raised sexual questions themselves. Zweifler et al. (1998) also focused excl usively on men, exploring the reasons that diabetic Latino men do not discuss their sexual difficulties with their doctors. Zweifler and colleagues conducted interviews with 82 Latino (mostly Mexican-American) male patients who were diagnosed with diabetes me llitus. The men ranged in age from 19 to 86 years, with a median age of 56 years. Most of the men reported erection difficulties, 59% always had erection difficulties and 82% indicated that they frequently (more than one-half the time) had erection difficulties. Only 47% of patients had discu ssed these issues with their physicians. Significantly, among the patients with frequent erectile dysfunction, 90 % were interested in receiving information about impotence a nd 85% were interested in trying a new

PAGE 48

36 method to improve erections. However, less than half of had discussed it with their physician. Men did not initiate discussion in most cases because they were either embarrassed or because they expected th eir physician to initiate the dialogue. These studies indicate that in most cases half or fewer than half of patients had actually discussed any of their sexual concerns with their doctor. Repeatedly, one study after another indicates that mo st patients would like for thei r physician to provide a more comfortable environment for discussing sexual ity issues by gently raising the topic. These studies also reveal that many women a nd men have sexual issues that they would like to discuss with a physicia n (Loehr et al. 1997; Nusbaum et al. 2002; Metz and Seifert 1990; Zweifler et al. 1998); in some cases a par ticular type of doctor, therapist, or nurse is felt to be more appropriate than others (Metz and Seifert 1 990; Waterhouse 1993) and gender concordance between patient and physic ian was preferred more often than not (Nusbaum et al. 2002; Metz and Seifert 1990; Zweifler et al. (1998) Indeed, physicians are routinely credited as the most trusted s ource of healthcare information by the general public (Haas and Coe 1997; Makadon and Silin 1995). Although physicians can be good educators for their patients, most have littl e or no training in how to design, developed, and present information effectivel y to others (Haas and Coe 1997). What Physicians Do: Discussion of sexua lity and Other HIV Exposure Possibilities in Clinical Contexts Several qualitative and quantitative studi es, and review articles, indicate that physicians and other health care providers are uncomfortable talking about sexuality issues, including HIV, with their patients (Epstein et al. 1998; Kerr et al. 1996; Makadon and Silin 1995; Penson et al. 2000; Ross et al. 2000). In particular, primary care general practitioners, who increasingly have to a ddress HIV prevention and sexual dysfunction

PAGE 49

37 issues in their practices, report difficulty a nd discomfort in addressing their patients' sexual concerns (Epstein et al. 1998; Sadovsky 2000). This is particularly evident with older patients. In 1996 a survey of prim ary care physicians (n = 124; 72 internal medicine and 52 in family practice) in Dallas County, Texas, reported that most physicians rarely or never disc ussed HIV/AIDS with patients older than 50 years of age, and were less likely to discuss symptoms of HI V infection or to counc il older patients for HIV testing than their younger patien ts (Skiest and Keiser 1997). Certainly the physician's speci alty is predictive of thei r likelihood of engaging their patients about sexuality, but the sex and ethnici ty of the physician and patient, as well as the recentness of medi cal training also appear to be important factors. A national random-sample survey of 2,545 primary care physi cians to determine the extent to which they provide HIV prevention and clinical se rvices found that male physicians and the physicians' perception that pati ents would be offended by se xual history questions were predictive of not asking new patients about th eir sexual behaviors and drug use (Kerr et al. 1996). OB/GYNs were the most likely to ask these questions, whereas GPs were the least likely to ask these ques tions. OB/GYNs were more th an six times and those in Pediatric/Adolescent medicine we re four times more likely to ask about condom use than all other specialties. Additionally, physicians who treated mostly White patients were less likely to report asking their patients a bout their sex and drug behaviors. Black and White physicians were less likely than Hispanic Asian and other physicians to ask adults about sex partners. In general, White physicians, male physicians, physicians whose patients are mostly age 65 and older, physic ians who treat mostly White patients, and physicians who believe that their patients would be offended if they asked questions

PAGE 50

38 about sexual behaviors and /or drug use were less likely to engage their patients on HIV/AIDS related topics. Although, physicians appear more likely to discuss sexual issues, including HIV prevention, with their younger pati ents than with their older pa tients, Epstein et al. (1998) found that physicians' discomfort with discus sing these topics re sults in ineffective evaluation of the risk status of even their younger patients. In a convenience sample of 17 Family Physicians and Internists and 26 patients age 18-45 who indicated a concern about HIV risk on a pre-visit questionnaire, videotaped interaction revealed that physicians did not elicit enough information to ch aracterize a patient's HIV risk status in 73% of the encounters. This appears to be directly related to physician discomfort with discussing sex related issues. Nonverbal barriers to effective communication were frequent. For example, physicians were obs erved to shake their heads no, avoid eye contact, or turn their back to the patient wh ile asking about sexual beha vior and to initiate discussions about HIV during a genital examination (E pstein et al. 1998: 438). Typically, physicians avoide d HIV discussions by changing the subject or simply ignoring patients stated concer ns and cues that they were interested in discussing HIV risks. Patients in our study wanted their phys icians to introduce th e topic of HIV, but few physicians had a comfortable standard quest ion or statement with which to introduce the topic (Epstein et al. 1998: 440). One 1990s study, however, indicates that more recently trained physicians may be more comfortable discussing sex and drug-relate d issues with their patients. The mail-in survey indicated that most (81%) recent medi cal school graduates indi cate that they are moderately or very comfortable discussing HIV/AIDS topics, incl uding sexual behavior

PAGE 51

39 issues, with their patients, but assessed only 0-25% of their patients to be "at-risk"(Haas and Coe 1997). This survey included bot h recent medical school graduates and preceptors from all areas of medicine. Th e 103 physicians who graduated between 1950 and 1984 were less likely to discuss HIV-rela ted topics than thos e who graduated after 1985. Both recent graduates and well-establish ed physicians discuss HIV/AIDS with the patients they perceive as at risk, however recent medical school graduates consider more of their patients at ri sk, and felt more comfortable with HIV/AIDS topics and discussed these topics with a higher pe rcentage of their patients. For the patient discussion questions, most selected the 0-25 percentage of their patients category with which they discussed the following: ge neral HIV/AIDS; HIV/AIDS risk factors; modes of transmission; "safe sex"; and, abstinence. Only 9.8% answered None for percentage of patients with whom they discussed general HIV/AIDS. However, this number increased to 10% did not discuss risk factors, 12.1% mode of transmission, 17.3% di d not discuss 'safe sex' with patients, and 25% reported not discussing abstinen ce with patients. For all of these questions the [St. Louis University School of Medicine] alumni report discussing these HIV/AIDS-related topics with a highe r percentage of thei r patients than the preceptors group. This only reached statis tical significance in the risk factors and modes of transmission questions, however. (Haas and Coe 1997: 405) It is encouraging to learn that more recent medical school graduates are more comfortable with discussing HI V/AIDS, though their mispercep tions about who is at-risk is not so encouraging. The reluctance of hea lth care providers to discuss HIV/AIDS with their patients, especially th eir older patients is a signifi cant disadvantage to prevention efforts with midlife and older persons. Indeed, larger quantitative studies indicate that Family Practice physicians (FPs) do not routinely recommend HIV testing for the ma jority of their sexuall y active patients. A survey of FPs in Rhode Island and Mississi ppi showed that although 93% tested their

PAGE 52

40 high risk patients for HIV, only 7% of th eir sexually active pati ents aged 18-50 had been tested in the previous year (Simm ons, Ma, Stumpff and Flanigan Routine HIV Testing in the Primary Care Outpatient Se tting, poster presente d at the October 2004, North American Primary Care Research Group (NAPCRG) annual meeting, Orlando, FL.). The difficulty appears to be narrow conceptualizations about which patients that are at-risk for HIV. These authors conclude, In order to not stigma tize any specific risk groups, nor to miss any patients who are unable to be identified as being at high risk, routine testing in the primary care setting should be encouraged. These studies indicate that most primary care physician s who treat patients aged 45 and older do not discuss HIV prevention topics or sexuality topics with their older patients (Frank and Harvey 1996; Kerr et al. 1996; Skiest and Keiser 1997; Wilson and Kaplan 2000). However, those that consid er the physician's gender (Frank and Harvey 1996; Kerr et al. 1996; Wilson and Kaplan 2000) noted that women physicians discuss sexuality issues more than men physicians regardless of physician specialty. These studies point to the fact that many physicians do not perceive of their older patients as sexual beings who may have sexual concerns or questions, or may be at risk for STDs. Indeed what often comes to mind when the dating scene for single women and men is mentioned are folks in their 20s and 30s. Ho wever, this perception is contrary to the reality as noted in a very recent AARP su rvey of 3,501 single women and men between ages 40-69. The survey focused on their lifestyles, dating and romance patterns, including risky behaviors. The study notes that over a third of all persons in the US ages 40-69 are single, and divorcees make-up a plur ality of this group (AARP 2003). About a third of single women and men in the AARP st udy were in exclusive dating relationships,

PAGE 53

41 and almost the same percentage (32%) were dating non-exclusively. Many others indicated an interest in finding a new rela tionship. This report shows how out-dated physician and even societal views are regard ing sexuality and aging and HIV/STD risks for older individuals. In spite of the recognized need for pr ovider involvement in prevention education, many obstacles exist, including clinicians narrow conceptions of medical care, discomfort with discussing sexuality and drug use, attitudes about HIV/AIDS, and time in resource constraints (Strombeck and Le vy 1998: 918). Although there are excellent articles (Dustow 1995; Wilson and Castillo 2004) and books (Miller, Mayer, and Makadon 1999; Ross et al. 2000) that offer practical suggestions for discussing the subjects of sex, sexuality, safer sex and HIV with a diverse population of patients, few providers appears to use them. Below is a review of the epidemiology of HIV/AIDS among midlife and older individuals and the trends in HIV/AIDS infection and HIVrelated knowledge, indicati ng the salience and timeliness of incorporating these communication tools into clinical practice. Part Three: Epidemiology of HIV/AIDS Among Persons Aged 50+ and Age, Gender and Ethnicity in HIV/AIDS-Related Knowledge and Risk In 1989, Catania et al. argued persuasively that the prevention of HIV transmission "is the single most important AIDS-related public health goal," and that greater attention should be paid to preventing transmission of HIV among late-middle-age and elderly individuals (1989: 77). Since th at time the need to target prevention efforts at mature persons, has only increased, with adults ag e 50 and older consis tently representing approximately 10% of the cumulative AIDS caseload (Strombeck and Levy 1998). In 1998, Ory et al. continued this theme by poin ting out that persons age 50 and older are

PAGE 54

42 still largely ignored in both HIV/AIDS scientif ic research and intervention efforts, despite being among one of the fastest growi ng segments of the AIDS caseload. In the mid-90s, when the number of adults diagnosed with AIDS showed their firstever decline, new incident cases among the midlife and older populat ion (50 years of age and older) rose twice as fast as those 49 years of age and younger (22% and 9% respectively) (Ory et al. 1998). Nevertheless, calls for prev ention efforts, particularly among health care providers, go largely unh eeded. Although sexuality issues are increasingly visible in the media, news and entertainment, as well as taught in schools of medicine, nursing, social work, public health and psychology, the skills for talking about sex and diagnosing sexual problems or couns eling patients with se xual concerns remain under emphasized for many medical practitioners (Ross et al. 2000). Due to the lack of intervention efforts that have been direct ed toward midlife and older populations ( see Strombeck and Levy 1998), both within and outsi de the clinic, it is not surprising that these populations tend to ha ve less information and more misinformation about HIV/AIDS risks, and that physicians of ten underdiagnose and misdiagnose their older patients because they do not perceive them to be "at-risk" for HIV (Ory et al. 1998). As among individuals in other populations, gende r issues, sexuality issues, structural, cultural, sociodemographic and economic issu es impact how mature individuals are affected by the HIV/AIDS epidemic. Recognition of the diversity of the older adult population needs to replace ageist assumpti ons that misrepresent and homogenize this group for more effective interven tion and prevention strategies. Surveillance Reports on HIV/AIDS Among Persons Age 50 and Over Surveillance data from the Centers fo r Disease Control and Prevention (CDC) estimates the number of persons living w ith AIDS increased 7.9% from 1999 to 2000

PAGE 55

43 (CDC 2001b: 3). By the end of 2002, an es timated 384,906 persons in the United States were known to be living with AIDS (CDC 2002: 5). Death rates for persons with AIDS have continued to decline, mostly due to the use of highly active antiretroviral therapy (HAART). During 1998, the estimated nu mber of deaths among persons with AIDS declined 14% (CDC 2002:5) Although the estimated number of AIDS cases diagnosed annually declined substantially from 1996 through 1999, the rate of decline observed between 1999 and 2000 slowed relative to previous years (C DC 2001b), by 2002, AIDS diagnoses increased two percent (CDC 2002). Additi onally, these declines are not significant among all age, sex, ethnic and regional categories. The leveling in overall AIDS incidence is occurring as th e composition of the epidemic is changing. AIDS incidence decl ined in most populations but increases were observed in some groups, notably women and persons infected through heterosexual contact.From 1999 to 2000, th e number of estimated deaths among persons with AIDS declined 11%; this decline, however, was less than that observed from 1996 to 1997 (42%)when th e widespread introduction of highly active antiretroviral therapy (HAART) had its greatest impactD eclines in deaths, which have been attributed in part to the impact of HAART, also reflect declines in AIDS incidence which in turn reflect bot h underlying trends in new HIV infections as well as delays in disease progre ssion among HIV infected persons who are receiving treatment. Other factors associated with the slowing of declines in deaths from 1999 to 2000 compared with those pr eviously observed may include delayed test seeking among certain populations, lim ited access to or use of health care services, and limitations of current therapies among persons in care. (CDC 2001b: 3) Among the issues mentioned above, delaye d test seeking and access to health services are particularly salient for minoritie s, older age groups, and women. Even with delays in testing and reporting, the per centages of minorities, minority seniors and women infected with HIV, and diagnosed with AIDS continue to increase (CDC 2001a, 2001b; Florida Department of Health 2003a 2003b, 2003c). As of 1996, the estimated adult/adolescent AIDS incidence, by r ace/ethnicity was 46.1% for the White, non-

PAGE 56

44 Hispanic population, Black, non-Hispanic s accounted for 34.8%, and Hispanics accounted for 17.8%. However, by the year 2000, the incidence of AIDS cases had undergone a significant shift in ethnic make-up. The percentages for White, nonHispanics dropped to 37.8% of all estimated cases, Black, non-Hispanics rose to 41.2%, and Hispanics rose to 19.6% (CDC 2001b: 30). Newer surveillance da ta from 30 areas in the U.S. and territories that have instituted confidential-name-based HIV infection reporting since 1998 through 2002, shows that th e estimated number of diagnoses of HIV/AIDS by race is now 54% Black, 32% Wh ite, 13% Hispanic, and 1% other (CDC 2003). From 1998 through 2002, AIDS incidence decrease 61% among children and 24% in the age group 25 years, and remained stable in the age group 35 years. AIDS incidence increased in the fo llowing age groups: 13, 15, 45, 55, and 65 years and older (CDC 2002: 6). In 2001, 10.9 % of all cumulative AIDS cases (based on age at diagnosis) were of men and women age 50 and older (20.7% of all persons age 45 and older). Women age 50 and over accounted for 9.5% of all reported AIDS cases in 2001; African American women alone accounted for 50% of this total. Men age 50 and over accounted for 11.2% of all reported cases (CDC 2001b). If we drop the cut off age down to 45, then women accounted for 17.3% of all reported AIDS cases in 2001, and men age 45 and over accounted for 21.4% of all reported cases (CDC 2001b). The numbers of midlife and older women and men infected, particularly minority women, is more significant in the southern United States. Throughout the epidemic Florida continues to be one of a handful of stat es with the highest HIV/AIDS rates.

PAGE 57

45 Epidemiology in Florida The South has remained the region with th e highest estimated number of persons living with AIDS, accounting for 38.2% in 2000, though the Northeas t follows with a close behind at 29.8% (CDC 2001b: 38). Fl orida ranks third behind New York and California in total a dult/adolescent cases (Flo rida Department of Health 2003b). Florida accounts for about 11% of the estimated number of persons living with HIV/AIDS in the US (Florida Department of Health 2003b). However the gender and ethnic trends in HIV/AID incidence are notable. African Americans account for 55% of cumulative reported HIV cases in Florida compared to 40% nationally, and women account for 36% of cumulative reported HIV cases in Florid a compared to 29% nationally (Florida Department of Health 2003b). Of women in Florida diagnosed with HIV and with AIDS through 2003, 72% were Black, 16% and 17% (r espectively) were White, and 11% and 10% (respectively) were Hispanic (Flo rida Department of Health 2003c). In Florida, through 2003, 14% of AIDS cases and 12% of HIV cases were reported in persons age 50 and over. Th e racial trend noted for all HI V/AIDS cases in Florida are mirrored in its midlife and older population. Significantly, Black, non-Hispanics account for almost half (51%) of all seniors (age 50 and over) that have been diagnosed with AIDS in Florida (Florida Department of Health 2003a). White, non-Hispanics account for 33% and Hispanics represent 16% of senior AIDS cases. Unfortunately, most persons age 50 and over have not been tested for HIV. This is in part related to age and ethnic trends in HIV/AIDS knowledge and risk perceptions.

PAGE 58

46 Intersection of Age, Gender And Ethnicity in HIV/AIDS Risk & HIV/AIDS-Related Knowledge In the early years of the HIV/AIDS ep idemic, the source of infection for many older adults was often through the receipt of contaminated blood or blood products. This source of infection distinguished them from other age groups who were not as likely to be blood product recipients (Strombeck and Le vy 1998). Unfortunately, when these older adults became infected through blood transf usions, many health care providers and policy makers falsely assumed that the infection w ould not spread based on the stereotype of older adults as disengaged from sexual ac tivities (Laskin Siegal and Burke 1997). Laskin Siegal and Burke argue that "[t]hi s myth that older ad ults infected through transfusion would not go on to infect others se xually has been one basis for the lack of education of the older populati on regarding the risks from th is phase of the epidemic." (1997:156). The good news is, due to improved blood screening methods in place by 1985, the AIDS caseload of Americans diagnosed at ag e 50 and older who had become infected through blood transfusions, decreased from 17% in 1988 to approximately 6% by the end of 1996 (CDC 1988, 1996, as cited in Strombeck and Levy 1998). The bad news is, HIV transmission for newly identified AIDS cases in adults over 50 years of age is primarily through high-risk sexual behavior (Stromb eck and Levy 1998), and despite this trend most midlife and older adults are largely stil l not targeted by clini cal and public health prevention efforts (Ory et al. 1998, Linsk, Fowler and Klein 2003) As for the general population, the trend to wards more heterosexual transmission of the virus that causes AIDS appears to be increasing among both sexes and all ethnic groups. By comparing the most recent trend in AIDS incidence by exposure category

PAGE 59

47 and race/ethnicity diagnosed in between 1998 and 2002, each ethnic group showed slight increases in the heterosexual mode of transm ission and slight declines in intravenous drug use (IDU) transmission, with the exceptions of Asian/Pacific Islanders and female and American Indian/Alaska Native whose I DU mode of transmission increased (CDC 2003). HIV/AIDS in older persons Mack and Bland (1999) present four interre lated issues that tend to distinguish older persons from younger pers ons with regard to HIV/AI DS: 1) the progression from HIV infection to AIDS occurs faster among older persons, and the progression of AIDS opportunistic illnesses often progress at a fast er rate as well; 2) AIDS, sometimes called the "new great imitator", causes many of th e same symptoms typical of the chronic diseases that are common among older person s, making correct diagnose challenging; 3) health care workers do not expe ct their older patients to be at-risk for HIV/AIDS, and therefore miss many prevention and diagnos is opportunities, because of ageist assumptions about sexual and/or drug use behaviors among older persons; and 4) more older people die from AIDS re lated illnesses in the same m onth that they are diagnosed than younger people because of the three issues listed above. The problem of late diagnosis and misdiagnos is is illustrated in some retrospective analyses of hospital records for older patients who died without HIV/AIDS diagnosis. Of 170 patients (ages 60 to 79) in Harlem Hospita l Center in New York City, blood samples showed that 6% of the men and 9% of th e women were HIV positive. Nearly all the cases were misdiagnosed, and few had know n risk factors for HIV (AIDS Alert 1995:79). Another retrospective study at Gr ady Hospital in Atlant a found that 32 HIVpositive elderly patients (age 60 or older, mostly Black and male) that presented with

PAGE 60

48 signs or symptoms of AIDS were not test ed for HIV until months later (AIDS Alert 1995). Most of the patients acquired HIV th rough sexual intercourse (38%), primarily through homosexual contact, and injection dr ug use (16%). Blood transfusions accounted for only 9 % of the infect ions (AIDS Alert 1995: 79). Mack and Bland (1999) point out that age-related issu es in HIV infection and diagnosis are perhaps more problematic for older women because of social and physiological factors that place them at greater risk of in fection, and earlier death from infection. Physiologic changes in vaginal ti ssues make tearing dur ing intercourse more likely, and this increases the risk of HIV inf ection, particularly when barrier methods are not used. Also, life expectancy for older wome n with AIDS is less than it is for older men, "in part, because older men tend to make more antibody that older women" (Mack and Bland 1999: 688). Compounding these physiologic factors is the fact that older women are less likely than younger women to use condoms since they typically associate them with pregnancy prevention and/or they do not perceive themse lves to be at risk for HIV and other STDs that barrier methods help to prevent (Mack and Bland 1999, Yates et al. 1999, Zaboltsky 1998). Additionally, women (of all ages) are so metimes at a disadvantage in protecting themselves from HIV because of their struct ural position in heterosexual relationships and cultural values defining their sexual inte ractions with men (L ear 1997). Financial dependency on a male partner may put some women at a disadvantage in negotiating condom use (Florida Department of Health 2003c), and the mere s uggestion that condom protection is needed puts rela tionship trust into jeopardy.

PAGE 61

49 Midlife and Older Women and HIV/AIDS The trends in the AIDS incidence rate fo r midlife and older persons are especially troublesome for women. Strombeck and Levy (1998) report that from 1991 to 1996 the number of AIDS cases attributed to he terosexual contact increased 106% and IDU increased 75% for women age 50 and older, whereas the number of reported AIDS cases in older men attributed to heterosexual cont act increased 94% and I DU increased 53%. For mature women, particularly ethnic minority women, the trends of HIV/AIDS infection among those aged 50 and older are particularly devastating. Zaboltsky (1998) reports that in 1989, more than three-fourths (77 %) of women age 65 and older with AIDS were White (non-Hispanic). This had de clined to slightly le ss than half (48%) as of 1997. At the same time, the proportion of women in the same age group who were Black increased from 16% in 1989 to 38%, a nd Hispanic women increased from 7% to 12% over the same time period. Significantl y, younger age groups in the 50+ category show similar patterns; from 1989 to 1997, non-Hi spanic white women age 50-54 declined from 37% to 23%, whereas Black women increased from 47% to 54% and Hispanic women increased from 14% to 21%. This patt ern is continued in each 50+ five-year age cohort in the number and percenta ge distribution of AIDS cases. The reason for these patterns are unclear; th ey may be due to women of color aging with their infection and/or being diagnosed at older ages, or from more of these women contracting HIV later in life from an HI V-positive man (Zaboltsky 1998). Results from one national study of women age 50 and olde r, with heterosexually acquired AIDS, compared to those age 49 and younger (most of the women from both age groups were Black and with incomes below $10,000), found that the older women were more likely to be widowed, separated, or divorce d; to live alone; to have less than 12 years of schooling;

PAGE 62

50 and to report having been exposed to HIV th rough sex with a man whose risk of exposure was unknown (Schable, Chu, and Diaz 1996). Compared to younger women in the study, women 50 years of age and older were less likely to have had a male sex partner in the past year, less likely to have a history of other STDs, more likel y to have never used condoms, and more likely to been tested fo r HIV while hospitalized. The age-related difference in condom use is consistent with other survey data indicating very low prevalence of condom use among midlife and older women and relates to age-related difference in perceived risk for HIV infecti on, and the use of barrier methods for disease prevention rather than pregnancy preventi on (Binson, Pollack and Catania 1997; Theall, Elifson, Sterk and Klein 2003). Among a subsample of women age 40-75 w ho participated in the 1990 (wave 1) and the follow-up 1992 (wave 2) National AIDS Behavioral Surveys (NABS), only 8% of women had any risk factors for HIV in fection (Binson, Pollack and Catania 1997). While this means that 92% had no risk factor s, the remaining 8% represents 4.5 million midlife and older women in the U.S. The findings were less optimistic for minority women, and African American women, in par ticular, had the highest prevalence of at least one risk factor (12.4%) if they lived in a "high-risk city" (high-risk city (HRC) refers to 23 metropolitan areas that have large numbers of AIDS cases and large Hispanic and African American populations, which were used to over-sample minorities). The length of sexual relationship had little rele vance for risk behavior, because women who had been partnered for 40 years or more were equally as likely to repor t a risk behavior as women who had been partnered for less than 5 years. When this finding is combined with the fact that 13% in the national samp le and 22% in the HRC sample did not know

PAGE 63

51 the risk status of their primary sex partne r the potential for expos ure goes up. The authors note that when women who did not know the risk status of thei r primary partner are included in the definition of risk, "the pr evalence of risk incr eased to 19.6% in the national sample and 22.7% in the HRC sample (Binson, Pollack and Catania 1997: 350). This study also points to a low percep tion of risk among these women; 90% reported never having been tested for HIV, and 90% reported never having used condoms (75% of those who lived in "high-risk citi es"). "Moreover, about 90% of the women who did not know the risk status of their primary partner indicated that they were not at-risk for HIV" (Binson, Pollack and Catania 1997: 349) In a study comparing two age cohorts (a ges 18-29, and ages 40-71) of mostly unmarried African American women (85%) on their perceptio n of HIV risk, found that most of the older women (55%) reported no ch ance of infection compared to 29% of the younger women (Theall et al. 2003). Several i ssues increased the pe rception of risk for both groups women including: th ose who had been cheated on by their most recent partner, or they had cheated on their partne r; know someone with HIV/AIDS or someone who has died of AIDS; sexual preference for males only; having been tested for HIV; were treated for an STD in the past year; and perception of low efficacy of condoms or disliked male condoms. However, most of the women in the study regardless of age perceived themselves to have no risk or to be at low risk for contracting HIV, even those reporting at least one risk factor. Although midlife and older women have b een at relatively low risk of HIV infection throughout the epidemic there overall rate of e xposure has been climbing over the past decade and this is especially tr ue for ethnic minority women. Age, gender,

PAGE 64

52 ethnicity, and SES all play a role in determ ining risk behavior, but these demographic characteristics should not be the foundati on for categorizing one group "at-risk" and another "not-at-risk", partic ularly for prevention education strategies. Instead, the prevalence of behaviors, along with the unde rlying beliefs, customs and behavioral history, should be understood so that interv ention and prevention efforts are relevant. Midlife and Older African-Americans and HIV/AIDS Rodgers-Farmer (1999) examines the probl em of categorizing an ethnic group as "at-risk" without contex tualizing that risk in her anal ysis of AIDS knowledge, attitudes and behavior among African Americans aged 55 and older. Rodgers-Farmer points out, African-Americans are not disproportionate ly affected by AIDS because they are African-Americans, but because prevention programs aimed at this population have not been culturally specificNor have these programs adequately addressed pervasive beliefs in the African-America n community that may prevent AfricanAmericans from participating in AI DS prevention programs (1999: 2) African-Americans may be re luctant to pa rticipate in government -sponsored programs on HIV/AIDS because of common knowledge about past medical abuses, such as the Tuskegee experiment, and pervasiv e beliefs that AIDS is part of the white conspiracy to commit genocide against African-Americans (Rodgers-Farmer 1999). Additionally, misconceptions about who is "at-risk" for HI V also impede AIDS prevention efforts in this community (Rodgers-Farmer 1999). Based on analysis of an African-America n subsample (n= 745) from the National Health Interview Survey (NHIS-AIDS), which explored AIDS knowledge and attitudes from half of the families who participated in the NHIS, Rodgers-Fa rmer examined AIDS knowledge, prevalence of HIV antibody testi ng, and prevalence of HIV-related risk among those age 55 and older.

PAGE 65

53 Ninety-eight percent of th ese respondents perceived their chances of getting infected with HIV as "low/no chance", and the same percentage perceived their chances of currently being infected with HIV as "low/no chance" (Rodgers-Farmer 1999). This may help explain why African-Americans at risk for HIV infection are underutilizing HIV antibody testing, with only 7.6% of those age 50 years and older with a risk factor for HIV having been tested compared to 28.9% of those 18-29 years and 26.4% of those age 30-49 (Grinstead et al. 1997) The Grinstead et al. st udy also used the NHIS-AIDS, but made comparisons using the whole samp le of 2,717 heterosexual African-Americans between age 18 and 75. Only 14% of the NHIS-AIDS respondents age 55-75 reported that they had been tested for an tibodies to HIV (Rodgers-Farmer 1999). Overall, the results suggest that those persons with less than 12 years of education, those persons whose family income is less than $20,000.00 per year, and those persons who perceive themselves as not be ing at risk for HIV infection have low levels of AIDS knowledge A substantial percentage of respondents failed to recognize that persons could still look healthy and have the AIDS virus. This particular finding is of interest becau se it has implications for engaging in unprotected sexual intercourse and the initiating of HI V antibody testing prior to engaging in sexual intercourse (Rodgers-Farmer 1999: 11-12) These African Americans tended to be ve ry knowledgeable about the major modes of transmission (sexual interc ourse, perinatal contact, and sharing needles for drug use), with 80% choosing the correct answers for que stions about modes of HIV transmission. However, they still had many misconceptions about transmission th rough casual contact. For example, over 90% believed that there was a possibility of contracting AIDS from being cared for by someone who had AIDS; from being bitten by mosquitoes or other insects; and from being coughed or sneezed on by someone who has AIDS. Moreover, 80% of the respondents believe that there was a possibility of contracting AIDS from eating and restau rant with a cook who has AIDS; from using public toilets; and from sharing eati ng utensils with someone who has AIDS. Finally, over 70% believed that there was a possibility of contracting AIDS from working near someone who has AIDS, and from attending school with a child who has AIDS. (Rodgers-Farmer 1999: 8)

PAGE 66

54 Many respondents (45%) also believed that th ere was vaccine to prevent AIDS. Yet, few respondents had any knowledge a bout the available drugs to treat HIV, or knowledge about the benefits of testing and early treatment (R odgers-Farmer 1999). As for the perceived effectiveness of c ondoms, 40% indicated that they did not know if condoms were 'effective or not '; 35% believed that condoms were somewhat' effective; 21% believed that condoms were very effective'; and 4% believed that condoms were not at all' effective... Only 15 percent of the respondents knew the oil-based lubri cants can cause latex condoms to break(Rodgers-Farmer 1999: 8). The implications of misinformation in HIV/AIDS knowledge are supported in the recent data on the incidence and prevalence of HIV/AIDS among African Americans. Although lack of income and lowe r education status appear to be correlated with lack of accurate knowledge about certain HIV/AIDS i ssues, an ethnic comparison of people with AIDS with different income levels and edu cation suggest that misi nformation is only one of many factors that affect HIV know ledge and transmission rates. As noted in the previous section on epidemiology of HIV/AIDS, African Americans suffer disproportionately for this epid emic, particularly in Florida. A recent study that looked at basic demographic a nd social information about people with HIV/AIDS (PWA) over the age of 40 year s in one Florida community, found some notable differences between middle-age and older African-Americans and Whites (Speer et al. 1999). Not surpri singly, the two groups differe d significantly on several sociodemographic variables. Sixteen percent of the participants from both groups were 60 years of age or older, and 28% of the Af rican Americans and 16% of the Whites were female. African-Americans were more likely to have high school education or less, whereas Whites were more likely to have co llege education. The average annual income of African-Americans is half that of White s in the sample at $7,015, and 81% of African-

PAGE 67

55 Americans fell below the poverty level versus 37% of Whites. As for sexual orientation, 82% of African-Americans described themselv es as heterosexual, whereas only 28% of Whites described themselves this way, and 59% of Whites described themselves as homosexual or bisexual as opposed to 11% of African-Americans (Speer et al. 1999). More African-Americans thought they had been infected by an opposite-sex partner (20%), an infected sex worker (14%), a st ranger (16%) or by sh aring needles (29%), while most Whites thought they had been infected through sexual relations with the same-sex partner (44%), and a stranger (23%). Although the groups did not differ significantly on pre-infection HIV preventi on education, 43% of African-Americans and 32% of Whites reported not receiving any prevention educat ion (Speer et al. 1999). The dramatic differences between these two groups on both socioeconomic factors and risk behavior factors, suggest that preven tion education efforts need to be sensitive to social and ethnic differences in order to be e ffective. In particular, African Americans in this sample were concerned about being assume d to be gay because of their HIV status. The social stigma associated with HIV/AIDS may be stronger in certain communities and is often related to misconceptions about the virus and who is "at-risk". The most recent data from the Behavioral Risk Factors Surveillance System (BRFSS) for the state of Florid a suggest that in terms of HIV testing, African Americans are ahead of the other two ma jor ethnoracial groups. The BRFSS shows that women in Florida are slightly more likely to have been tested for HIV than men (49.2%, 47% respectively) and 59.7% of Black, non-Hispanics reported have been tested, compared to 44.2% of White, non Hispanics, and 52.7% of Hi spanics (Florida Department of Health 2003d). These data do not include persons ove r age 64, and age is clearly an important

PAGE 68

56 factor in the likelihood of bei ng tested in the percentages ab ove with 57.3% of those ages 18-44 being tested, compared to 33.2% of t hose ages 45-64. Lower income and higher level of education were also correlated with a greater likel ihood of being tested. Notable for this research project ar e the data on health professi onals discussing STD prevention through condom use. In the past 12 mont hs, only 14.9% of men and 17.5% of women reported that a health professional had talk ed to them about preventing STDs through condom use. Blacks and Hispanics reporte d more of this co mmunication (29.9% and 29% respectively) than did Whites (10%), and Black women and Hispanic men reported the most (both 34%). Again, age is a significa nt factor with 21.3% of those ages 18 to 44 years reporting this communication compared to 8.8% of those 45 to 64 years of age. Income is also influencing the communicati on patterns of health care professional with 31.6 % of those with incomes below $25,000 co mpared to 7% of those with $50,000 or more in annual household income reporting that their providers talked to them about STD prevention. Clearly, health care professi onal are targeting specific age and ethnic groups, and poorer populations with this message. While it may be appropriate to target these groups, it should not be at the exclusion of other groups. This sort of focus tends to maintain the stereotypes about those at-ris k and not-at-risk so that older, more affluent populations may be missing this important message. Perception of Risk: The perils of the at-risk label Misperceptions about HIV risks and tran smission can be dangerous. Although national level data suggest that most older pe ople in the general population do not engage in HIV-related risk behaviors, the approxi mately 10% who do are less likely to report using condoms during sex, to have ever been tested for HIV, or to of change their

PAGE 69

57 behavior in the past 12 months than younge r persons who are "atrisk" (Ory and Mack 1998). This behavior is due in part to older persons not be ing assigned to the at-risk category by public health campaigns to combat the spread of HIV. At the closing of the second decade of the HIV/AIDS pandemic AIDS-related knowledge has increased for all ages, however older people are st ill less knowledgeable and younger people. For example, persons age 50 and older are more likely to think that HIV/AIDS can be contracted th rough casual contact; they are more likely to believe that they can be infected by being coughed on, shar ing food utensils, or using public toilets than younger people (Ory and Mack 1998). The studies detailed in the previous pages note that risk s behaviors and perceptions are often incongruous for many midlife and older persons. Many women and minorities that are in the older age groups studied about do not think that they are at risk even when they are engaging in high-risk behaviors. A study by Dancy (1996) on African Ameri can women and their AIDS-related knowledge, showed that despite adequate know ledge about HIV risks, these women often did not apply that knowledge to their own beha viors because they de cided that it did not apply to them. Theall and colleagues, suggest that this is the case in their study of risk perceptions between two age c ohorts of women. Engagement in risky behavior despite HIV/AIDS knowledge and with no pe rceived risk of infection is likely a result of denial and rationalization of behavior (Theall et al. 2003: 424). These findings about the perceptions of HIV risks among midlife an older women are cautionary. As women in the baby boomer generation turn 50, it is unlikely that their sexual behaviors and attitudes will change radically (Yates et al. 1999). They will

PAGE 70

58 probably continue to engage in more risky se xual and drug-related behaviors than those in older cohorts. However, the effectiven ess of prevention efforts targeted at younger adults may carryover into middl e age (Ory and Mack 1998: 661). The question of how to change percepti ons, encourage early testing and safer behaviors with a diverse population of midlif e and older persons remains. Below some of the more common strategies are considered and how lif e-course stage may affect prevention strategies. Testing the Behavioral Change Theory for Older Adults There are several theoretical models that have strongly influenced the CDC's AIDS prevention research: the health belief model, social cogni tive theory, the theory of reasoned action, and the transtheoretical m odel of behavioral change (Fishbein and Guinan 1996). In general these models s upport the examination of beliefs underlying attitudes and behaviors, as we ll as the contextual factors th at encourage individuals to behave in a certain way. "According to thes e models, individuals who are at greater risk, perceive themselves at greater risk, have mo re resources, and percei ve or actually have health problems are more likely to be tested for HIV than their c ounterparts" (Mack and Bland 1999: 689). All of these models are fo cused on the individual rather than social groups, and therefore in many ways neglect the cultural and structural factors that affect beliefs and behaviors. Mack and Bland (1999) tested the assump tions of the various models with two different age groups to assess whether theories of health behavior should take life-course stage into account. Their models test whet her perceived risk, resources, and health problems affect the likelihood of being tested for HIV among persons age 50-64. Their analysis of data from the 1996 Behavioral Risk Factors Surveillance System (BRFSS)

PAGE 71

59 compares adults aged 18-49 (77.6% of sample ) with adults age 50-64 (22.4% of sample) on their self-reported chance of getting infected with HIV, change in sexual behavior the last 12 months, perspectives of condom efficacy in preventing HIV, and marital status. The percentage of older adults who had ever been tested for HIV was 26.6% and for the younger adults was 46.8%. The percentage of adults aged 50-64 who are voluntarily tested was less than half that of adults aged 18-49. Much of this difference is due to the fact that pregnanc y was not being given as a reason for the last test by older women The percenta ge of respondents w ho report that their chance of getting HIV/AIDS is high or medium declines with age. The percentage of adults aged 50-64 who changed their se xual behavior the last 12 months because of what they know about HIV/AIDS is one-t hird of the percentage for adults aged 18-49 (5% versus 16.6%). (Mack and Bland 1999: 690-691) Most older men and women di d not rate their chance of getting AIDS as being high or medium (4.1% and 3.8% respectively). Consistent with other research findings, Blacks were more likely than Whites to have ever been tested for HIV (38.4% versus 24.4%). Older Blacks rated their chance of gettin g AIDS as high or medium only slightly more so than older Whites. "Perhaps the most notable difference by race for adults aged 50-64 was that 16.3% of Black respondents report ed that they had changed their sexual behavior in the last 12 months on the basis of what they knew about HIV compared with only 3.4% of White respondents." (Mack a nd Bland 1999: 692). Although the model confirmed the research resu lts for the younger portion of the sample, (males, Blacks, those who perceived themselves as having hi gh or medium risk, those who have more resources, and who have health problems were all more likely to have been voluntarily tested for HIV other than respondents) the pr edictive model of health behavior was not very robust for older adults based on this anal ysis, as measures of resources and health status had to be dropped from the model b ecause they made no significant contribution. Only perception of greater ri sk was predictive of being te sted for HIV (Mack and Bland 1999). This suggest that life-co urse stage may be important when designing intervention

PAGE 72

60 strategies, and that current programs directed at the gene ral population ar e not reaching older persons. Theall and her colleagues note that their study of risk perceptions among two age cohorts of women had implicat ions for different types of intervention and prevention messages/efforts aimed at younger and olde r women. Among younger women, HIV risk behaviors were associated with younger wome n's perception of ri sk, suggesting that knowledge is high but not related consistently to practice... in contrast to the preceding, our data revealed that HIV risk behavior was not as salient an issue for the older women's perception of HIV risk they suggests that programs working with older women should focus on risky drug use behaviors and sexua l orientation issues, since older women associated drug-related risk with their level of perceived HIV risk, suggesting that these women were aware of the risks associated with drug using behaviors. They also suggest that educational efforts for younger women m ay include issues such as pregnancy, dating, and establishing new romantic/sexual relationships, whereas for older cohorts, topics such as partner comm unications/negotiation, dating/se x after divorces, and risks associated with HIV infection for older indi viduals may be highlighted (Theall et al. 2003: 426). Targeting Older Persons with HIV/AIDS Prevention Messages Several experts in the field of preventi on have noted that HIV/AIDS programs are seldom prepared to consider aging issues, a nd that most programs serving older adults are not prepared to address HIV in fection as an issue (Linsk et al. 2003; Ory et al. 1998; Stall and Catania 1994; Strombeck and Levy 1998). Few local or national programs include screening for HIV-related risk behaviors, su ch as alcohol and s ubstance abuse or STD history. Nor do they address the concerns of many older persons about life transitions,

PAGE 73

61 such as loss of a partner through divorce or d eath, which can lead to changes in social and behavioral patterns. Although senior focu sed programs may serve many newly widowed or divorced individuals, there is little information available addressing the issues of socialization and intimacy. Negotiating friendships with possible intima te partners is a new behavior for many, and educational and support programs may help older adults a void detachment and withdrawal. A need exists for informa tion for older adults about how to reduce HIV-related risk, maintain personal safe ty, and make decisions about sexual relationships. Given that ol der adults are vulnerable to STDs, a basic need exist for HIV/STDs prevention education integrated with information about meeting one's social needs safely and avoiding risks . HIV/AIDS service organizations, however, may have information about risk and even about making social contacts, but these materials may not be age sensit ive or even may be overtly ageist by excluding the possibility of HIV infection/ AIDS and older adults. (Linsk et al. 2003: S245) One area where rapid change needs to occu r is in the medical professions. Since older persons, particularly women, see heal th care providers more frequently than younger populations, this is an ideal place to target older persons with HIV prevention and testing messages. Clearly from the studies reviewed in section two of this chapter, physicians rarely assess their midlife and ol der patients as at-risk for HIV and other STDs. Many also express discomfort in disc ussing sensitive issues, including sexuality issues, with this class of patients. Linsk and colleagues urge the better education of primary care physicians and othe r providers, which are the front line of care and services for many older persons in the community, to understand and address the fact that HIV and other STDs can and do infect older pe ople. Good patient care for sexually active older persons includes disc ussions about STDs (e.g., chlamydial infections, gonorrhea, and syphilis) as well as HIV and how to prev ent infection. Health practitioners should do routine risk assessments and offer HIV/STDs testing to older as well as younger patients who may be at risk and also should support pa tient requests for such testing. However,

PAGE 74

62 this practice is rarely the reality. Prov iders and patients have described how older persons are routinely discouraged or even ridiculed when trying to initiate such services. One well-publicized story of a 54-year-old heterosexual woman, who asker her longtime family practitioner if she should be tested for HIV because she had engaged in a few intimate relationships after her divorce, only to be told not you, you don't need that test. Two years later, when she applied for new me dical insurance but was rejected because a routine blood tests... revealed a significant bl ood abnormality. It was then left to the family practitioner to inform her patient that [ it] was, indeed, HIV infection (Linsk et al. 2003: S246). In general, based on the information presen ted in the preceding pages, the rate of HIV infection and AIDS diagnosis in the U.S. is growing among those who engage primarily in heterosexual activity; this is most significant for women and minorities particularly midlife and older women minorit ies. This is compounded by higher rates of poverty, age-related social a nd biological factors, gende r-power dynamics, ethnic and cultural perceptions of sexuality, HIV/AIDS risks and public health initiatives. The lack of accurate knowledge about HIV transmissi on and prevention practices among midlife and older persons points to the need for bett er education efforts directed at a diverse aging population. While there is an obvious need for broad public health efforts to target a general population of senior s through retirement organi zations and more specific regional populations th rough community outreach; there is also a need to bring health care providers in to the front line of preven tion education in the course of their work. Although HIV/AIDS and aging often are not perceived as being interrelated, the increase in the number of cases of HIV inf ection/AIDS in older adults indicates that elders are often overlooked in terms of HIV prevention efforts.... As individuals live longer with HIV infection/AIDS, this new group of long-term survivors will

PAGE 75

63 join those affected by HIV infection/AIDS in their later years. Service models for older persons and HIV/AIDS service models can benefit from targeted education and better integration of prevention, care, and supportive services. (Linsk et al. 2003: S249).

PAGE 76

64 CHAPTER 3 DATA COLLECTION: TH EORIES AND METHODS Theoretical Considerations Introduction Many cultural anthropologists, linguistically trained or not, focus their attention on what people say and the words they use to expr ess themselves, in their efforts to explain culture and cultural patterns. James Spradley noted in 1979, in his treatise on ethnographic interviewing, that in most ethnographic fieldwork, researchers make cultural inferences from what people say; the way people act; and the artifacts people use. Since both tacit and explicit culture are re vealed in language, ethnographers routinely focus much of their efforts on what people say, in both casual conve rsation and lengthy interviews, in their work to describe cultu re (Spradley 1979). Lingui stic anthropologist, Deborah Cameron, likewise notes that many anth ropologists regard language as a key to understanding certain aspects of culture, with vocabulary sometimes serving as a summary of a peoples concerns and be liefs about the world around them (1992). This project relies on what people say to reveal their cu ltural perceptions. I combine qualitative and quantitative ethnographic methods to describe cultural models of mature female sexuality. Data collection in cludes a short demographic survey; in-depth interviews following the presentation of h ealth care encounter vignettes; and cultural consensus modeling techniques including freelist, pile-sort, and rating procedures. In the first sections of this chapter I explain the rationa le for the methods used and discuss some examples of these methods that relate to the study of health perceptions.

PAGE 77

65 Secondly, I describe the methodology, the sampling and setting and why it was an appropriate choice. Finally, I describe the entre to the multiple research settings and provide some basic information about the participants. Ethnography Ethnographic interviews begin with the ethnographer talking to one person and then going to the next, usually finding some thing different, as well as many things in common (Handwerker 2001). We keep track of similarities, note variability, and keep data until we decide that we've exhaustively identified significant cultural variation... Th en, we construct a story from the inferential generalizations we've arrived at about the people we worked with, their liv es, and the circumstances in which they have lived; about what those people now think, feel, and do; about who agreed with whom about what to and to what extent; and, so, about who is similar to whom and to what extent and how they differ from othe rs and to what extent. (2001: 11-12) To accomplish the activities described in this passage this project combine structured and unstructured interview techniques. Bernard ( 1995) describes unstructured interviews as the best for getting people to open up a nd express themselves on their own terms. However, he suggests that in situations wh ere you might only get one chance to interview someone, a semistructured interview is best. It has much of the fr eewheeling quality of unstructured interviewing, and requires all the same skills, but semistructured interviewing is based on the use of an interview guide. This is a written list of questions and topics that need to be covered in a particular order (Bernard 1995: 210). This approach proved to be the best fit for the pr oject I undertook, since in most cases I would only meet with each informant once. I will provide more detail about the interviewing techniques in second half of this chapter. As a cognitive anthropology project I also included the cognitive methods of cultural c onsensus modeling, which rely on structured interview techniques to re veal cultural domains.

PAGE 78

66 Cognitive Theory and Methods This project employs the cognitive met hod and theory of Cultural Consensus Modeling (CCM). CCM is based on using mathematical models that simultaneously record an estimate of a respondent's cultura l knowledge about a partic ular domain and an estimate of how much they agree with ot her respondents on what is culturally known about the domain (Bernard 1995; Weller a nd Romney 1988). The central theme of consensus theory is that the correspondence between any two informants is a function of the extent to which each has knowledge of th e culturally appropriate responses for the interview questions (Weller and Romney 1988). Owing to its cognitive science foundations the concept of cu lture in cultural consensus theory is one that emphasizes wh at people know and how it affects what they do, with less emphasis on behavior itself. Ha ndwerker (2002) draws at tention to what he describes as the largely overlooked and unde rvalued implications of Edward Tylors 1871 definition of culture: (1) the culture th at specific people us e to live their lives constitutes an evolving configuration of c ognition, emotions, and behavior unique to themselves; and (2) cultures consist of evol ving configurations of cognition, emotion, and behavior at the intersection of individually unique cultura l sets (2002:106). He goes on to explain that while culture is learned and shared by groups of people, only individuals learn, and individuals embody and constitute the only source of cultural data. To make the world we live in sensible, all of us asse mble out of our individual sensory experiences ways of thinking about what we have experienced (109). In other words, while culture is shared by groups of people, the experience of knowing ones culture is unique for each individual.

PAGE 79

67 CCM builds on the concept of culture as shared "cultural model" (D'Andrade 1995; Dessler and Bindon 2000). Dressler and Bindon (2000) describe a cultural model as a highly schematized, skeletal representation of some cultural domain in that it defines the elements that make up a cultural domain. A semantic or cultural domain is basically any subject matter that can be coherently de fined (Weller and Romn ey 1988). According to Weller and Romney a domain is an organize d set of words, concepts, or sentences, all on the same level of contrast, that jointly re fer to a single conceptu al sphere (1988:9). Cultural domains can be organized around almost anything; they can be kinship terms, animal categories, beliefs about symptoms of disease and so on. Among other things, understanding a domain helps us understand how the world around us works and how to function in it. Dessler and Bindon describe it this way, In deciding what to do in any given situa tion, or in deciding wh at the behaviors of others mean in any given situation, we draw on our understanding of the world as that understanding is structured by our cultur al model. So, if we are heading off to the bait shop, or a baseball game, or a graduation ceremony, or a committee meeting, we have a pretty good idea what, in broad outlines, to an ticipate in others' behaviors, and what repertoire of behaviors we will draw on. (2000: 247) Individuals are part of multiple cultural groups, in the sense that they may at any given time occupy several social role s that require them to have specific cultural knowledge about several domains. Each individual draw s on a large set of cu ltural models, models they have learned and shared with othe rshowever incompletelyto get through each day (Dressler and Bindon 2000). From one pers on to another, there is variation in the knowledge of the elements of any given in cultural model, because models, although shared, their components are also shaped and understood through individual experience (Handwerker 2001, 2002; Dressler and Bindon 2000). There are three sources of vari ability in cultural models w ithin a society. First, as noted, we incompletely share cultural m odels; therefore, some people will know

PAGE 80

68 more (or less) about some domain than others. Second, there may be systematic variability in cultural models across social contexts Third, individuals may be limited for a variety of reasons in their abilities to act on cultural models; that is, they may know the model, but they may be unable to act in accordance with it (by economic constraints, for exampl e). (Dressler and Bindon 2000: 248). The techniques of CCM provide a systematic way to highlight th e shape and scope of cultural domains and the de gree of agreement and variability within them. Since 1986, when Romney, Weller and Batche lder introduced the consensus model of culture, several other researchers have us ed and added to the th eory and techniques (Chavez Hubbell, McMullin, Martinez, and Mishra 1995; Chavez, McMullin, Martinez, Mishra, and Hubbell 2001; Dressler 1996; Garro 1986, 2000, Handwerker 1997, 2002; Weller and Baer 2001; Weller a nd Romney 1988). The consen sus model of culture has contributed to anthropological theory by highlighting the shared nature of culture as a defining feature of cultural know ledge (Dressler and Bindon 2000). For this research project I use cultural consensus modeling (CCM) to explore if there is a shared cultural domain of sexuality for midlife and older women, and if primary care physicians also share that domain, or to what degree they sh are it. The cultural domain of sexuality, or sex-li fe as some preferred to describe it, for midlife and older women is a bit of a departure from traditi onal uses of CCM. However, several other researchers have used consensus theory and methods on abstract concepts and have used the concept of schemas to shed light on th e reasons for cultural variation highlighted through CCM. Consensus Theory and Cultural Schemas As stated above, CCM focuses on shar ed knowledge and helps to show the variability within what is shared but not th e reasons for the variability. To explore the factors associated with intra-cu ltural variation it is helpful draw on traditional qualitative

PAGE 81

69 methods to illuminate the reasons for variati on. Ethnographic interviews provide a useful complement to CCM data collection and anal ysis. Below is a brief discussion of how several researchers have succe ssfully combined the quantitati ve and qualitative features of CCM with more tradi tional ethnographic methods. Linda Garro (2000) used consensus theory to look for patterns and variation when she studied a group of Ojibway's beliefs a bout hypertension, but unlike her earlier work using CCM to examine the shared knowledge of curers and non-curers in Mexico (1986), this time she used combined methods to get at the nuances of intra-cultural variation. Using interviews along with CCM, and draw ing on Kleinmans work on explanatory models, she mapped the consensus model for the causes of hypertension among the Ojibway. However in doing so she also reveal ed the patterns of variation in the model, notably demonstrating how individual experien ce cause some to reject specific elements of the model or to reframe it based on differi ng criteria. Garros wo rk with the Ojibway led her to develop an analytical approach to some of the shortcomings of consensus theory. With its focus on what is shar ed, CCM has limited utility for describing the elements of difference within a domain; di fferences do not just indicate variation in cultural knowledge about a domain, but perhap s a different way of using the domain or putting the domain together. Garro describes this in terms of schemas or cultural models theory where variation between individuals reflec ts different schemas or different aspects of the same schemas. CCM and schemas are not in conflict; ra ther they are complementary tools for understanding cultural similar ity and variation. In a study of gender schemas and prostate cancer among veterans Stansbury et al. (2003) used CCM to examine cognitive

PAGE 82

70 schemas about gender. With a reliance on pr imarily emic categories of masculinity the researchers mapped the masculine ideology of the veterans, showing both the shared ideology between prostate cancer patients and non-patients, and the variation in that ideology brought about by the experience of dise ase. This research did not have the rich ethnographic data to explore the nuances of the variation that Garro was able to investigate. Nevertheless the data ar e supported by schema theory in cognitive anthropology (DAndrade 1995; Garro 2000), by showing how the findings support earlier more qualitative research on the affect of illness on gender constructs. The research lends support to conten tion that schemas are flexi ble interpretive states that reflect the mixture of past experience and present circ umstances (DAndrade 1995: 142). Several other researchers have used CCM to consider health and health care beliefs, and some have ventured to determine if be liefs are related to health care behavior. Hurwicz (1995) compared physicia ns with elderly patients to s ee if the two shared similar ideas about what illness symptoms necessitate d office visits. She also explored how the patients models of when to visit the doctor s office related to thei r behavior. The two groups did share a similar domain about doctor visits, however the pa tient behavior did not conform to the health-care seeking norms de scribed in their model. Her analysis of the health seeking behavior showed that tre atment choiceswere more a result of the characteristics of the [illness] episodes than of the different criteria being applied to the decision (1995: 232). Hurwiczs analysis of this difference between the model and the behavior points lends further support to the theory that models of cultural domains are

PAGE 83

71 shaped by a life-time of expe rience and the factors of curre nt circumstance, influencing how individuals interpret cultur al models and react to them. Chavez et al. (2001) also addr ess the topic of cultural belie fs and behavior in their analysis of beliefs abou t cervical cancer risk factors and the use of Pap exams. For this, they drew on their 1995 research that looked at intra-cultural variati on in beliefs about the risk factors for breast and cervical cancer among three sub-groups of Latinas, Anglo women and physicians. They found Latinas, particularly immigrants as opposed to Chicanas born in the US, held beliefs about cervical cancer that di verged from those of physicians and those held by Anglo women. In the 2001 report, they found that the closer the Latinas beliefs were to Anglo wome n the more likely they were to have had a Pap test in the last two years, yet the closer their beliefs mirrored the physicians the less likely they were to have had and exam. The di stinction here appears to be related to the emphasis that physicians put on sex-related behavi or as an important ri sk factor vs. Anglo women who down-play sex-related factors in favor of other f actors, such as heredity. The researchers earlier ethnographic data s how that the immigrant Latinas hold more conservative moral values about sexual be havior than Chicana and Anglo women, and this influences their health-seeking beliefs and behaviors. Howeve r the researchers are quick to point out the strong in fluence of structural factor s on health-seek ing behaviors such as lack of health insurance, educational level, and acculturation. These findings point to the importance of the complementarity between structural and cultural explanations in unde rstanding health beliefs and be havior. Chavez et al. also credit their early ethnographic interviews with a small sample of Latinas for the richness of the data and its usefulness in clarifying the reasons for th e intra-cultural variation in

PAGE 84

72 the cultural model, and the generalizability of ethnographic findings on health-beliefs to much larger populations based on smaller ethnographic studies. The measurement of shared cultural know ledge has been operationalized in the cultural consensus model and verified for sm all sample sizes by several anthropological researchers (Chavez et al. 1995, 2001; Ga rro 1986; Handwerker 2002; Handwerker and Wozniak 1997; Romney, Weller and Batcheld er 1986; Weller and Romney 1988). This project uses a relatively small sample as we ll to explore the doma in of sexuality for midlife and older women. To address the nuanc es of intra-cultural variation within the model the CCM data is combined with an interpretive analysis of the vignette-based interview data to contextualize the domain. What follows is a description of the methods used in this project. Methods and Sampling Introduction I created vignette-based interviews to a ddress womens and physicians' ideas about the clinical encounter, as well as general perspectives on aging and sexuality. These semistructured interviews relied on open-e nded questions and probes to elicit a broad understanding of similarity and variation in aging and sexuality concepts among the recently single women in a displaced ho memaker program (DHP) and among primary care physicians (referred to as physicians ) in the same city. The vignette-based interviews with both samples explored variation in ideas about provider-patient communication, beliefs about agi ng and sexuality, and individua l and social factors that may enhance or inhibit communication about se xuality issues, includ ing HIV/STD issues in clinical context.

PAGE 85

73 I also administered a survey (Appendi x A) with the women that gathered demographic data on age, ethnicity, marital st atus, number of children, number in current household, level of education, years lived in th e area, health status, recentness of doctor visits, and type of doctor visited. Surv eys gathered from physicians (Appendix B) provided information on the physician's age, recentness of medical training, type of practice, demographic profile of their patient population, pe rceptions of time constraints on patient-physician interaction, perceptions of HIV risk for their patients, sources of information on sexuality and aging, and ideas about patient and phys ician responsibility regarding discussing sexual health issues. Vignette-based interviews I read each DHP woman two vignettes involving a women and her clinical encounter, 1) a story about a women who unde rgoes a hysterectomy and suffers painful intercourse after the surgery, and 2) a stor y about a women not being tested for HIV by her primary care physician and only being test ed after she is hosp italized. After each story, I asked the respondents to tell me their opinion of the case, other more detail questions, and then general que stions about health care co mmunication and sexuality and aging (Appendix C). Physicians participated by either completing the survey and one of the CCM procedures, or were interviewed using the same vignettes discu ssed above (Appendix D). However, in order to save time I divided the interview schedule into two parts: the first group of physicians who completed the full in terview (n=9) was read the HIV story, and the second group was read the hysterectomy story (n=8). Some of the interview questions were similar to those asked of the DHP women sample, but several medical

PAGE 86

74 practice questions were also asked concerning their current views and practices regarding communication about sexual health issu es with midlife and older women. Analysis of the interview and survey data All of the survey data and key responses from the interview data were coded and input in to the Statistical Package for the Social Sciences (SPSS, v.12.0, 2003). The interview data were coded for text units in N6 (QSR N6, v 6.0, 2002) so that the appropriate context of the aggr egate responses could be explor ed. In several instances I sorted the interview data in N6 so that it could be added to SPSS. For example, the question about whether or not the DHP thought that other women their age were aware of the risk for HIV/STDs was an interview question and generated long responses, which were analyzed and grouped under yes, yes most, no, no, not most, and dont know. This interpretive analys is was applied to all interv iew responses that could be ordered in such a way based on the aggregate of replies. The SPSS data are analyzed for frequencies of the survey and interview questions and presented graphically in chapters five and six. The two samples are analyzed separately using frequencies a nd Chi-square tests to measure internal agreement, and the samples are combined on variables that are common to both groups to measure agreement between groups. For the DHP women, marital status, age groups, ethnic groups, and health status are us ed to determine if these vari ables influence their responses to the vignette interview questions. The physic ian sample is analyzed in much the same way as the DHP women, but on fewer vari ables (residency status, age groups, and gender). The coded N6 text data are used to highlight the nuances of the statistical data. The text data also provides crucial insi ghts to understanding the consensus data,

PAGE 87

75 described below, ultimately linking the various data sets to reveal the cultural schemas of mature sexuality. Cultural consensus modeling techniques The interview format for the women and the physicians also included the cultural consensus modeling (CCM) procedures. Appr oximately half of the physician sample participated in the survey/interviews/CCM pr ocedure and the other half were only asked to participate in the CCM proce dures and complete the survey. The respondents were asked to participate in one of thr ee systematic data collection procedures for CCM. The first procedure, fr ee listing (qualitative) of terms and phrases associated with the domain of mature sexual ity, occurred with th e first group of DHP women interviewed (n=18), and physicians (n=14). These free-lists were then be combined and distilled into a shorter version of the lists by collapsing repeated terms into a single term and/or removing terms and phr ases mentioned by only one respondent (see Borgatti 1999), to be used in the second st ep, pile sorting. This second procedure the next group of DHP women (n=15) and physicians (n=7) to sort the terms and phrases into categories or "piles." The list is then used in the final step, the rating procedure. The remaining participants (n=14 DHP women a nd n=10 physicians) were asked to rate the relevance of the terms to describe sexuality for women age 45 and older. However, due to the small sample size for this study, I asked three of the DHP women who completed the free list procedure to assist me by al so completing the rating procedure. These procedures and their results ar e described in chapter four. Analysis of cultural consensus data The results of the CCM data are analyzed using a variety of methods, all of which are described in greater detail along side the results in chapte r four. All of the CCM data

PAGE 88

76 are imported and analyzed using ANTHROPAC (Borgatti 1996a). Th e initial free listing techniques elicits e ach individuals notions about th e particular domain in question without imposing the researche rs own concepts. The results of the combined lists from all of the respondents that completed the free li st task are then anal yzed for frequency and salience of terms to find the core of the domain. The pile sort and ratings data are used to highlight the structure of the domain and help to determine if there is one cohesive domain and/or if there are subdomains and multiple domains. The single pile-sort ta sk (Borgatti 1999; Weller and Romney 1988)) elicits the way that the indivi dual respondents organize the 48 most frequently mentioned terms from the earlier freelist data. The results provide the coordinates for a twodimensional representation, using multidimen sional scaling (MDS); showing how the respondents believe the domain ought to be stru ctured. These results are also analyzed with cluster analysis, which is another way to show how items are grouped together. The ratings data are used to show the degree of agreement among the respondents, in this case, their perception of the relevan ce of the terms for describing womens midlife sexuality. The results of this task generate a profile matrix for cultural consensus analysis (Romney, Weller and Batchelder 1986). This provides a way to systematically measure the degree of shared perceptions am ong respondents and ascert ain the reliability of the data from the average intercorre lation among respondents (Weller and Romney 1988). Additionally, this analysis can be used to determine degree of agreement within and between groups; in this case between the DHP women and the physicians. Finally, the ratings and pile-sort data can be used together to test hypotheses about the characteristics that shape respondents notions about th e similarities among the items

PAGE 89

77 tested using property-fitting (PROFIT) (Bor gatti 1996b). PROFIT is a multiple regression technique that uses the coordi nates from the MDS pile-sort da ta as independent variables and the attribute used in the ratings test as th e dependent variable. Th is analysis assesses whether or not the hypothesized attribute wa s influencing the suspected similarity among the items. These methods and their analyses wi ll be described in greate r detail in the next chapter. Suitability of Methodology Due to the sensitive nature of this resear ch topic, the combined methods of cultural domain analysis and in-depth interviews w ith the vignettes were appropriate for this project and this population because they permit respondents to depersonalize their responses if they so choose, on this highly pe rsonal topic. This is because the vignettes describing a hypothetical encount er were followed with ge neral questions regarding respondent opinions on the pros and cons of the encounter. Respondents were also told that their responses would be kept confidential, that there were no right or wrong answers, and that their responses did not need to be based on their personal experience, but instead may just reflect their genera l knowledge and opinions about doctor-patient communication. In this way the respondents we re able to choose for themselves how much or how little personal information they wished to provide. The cultural consensus procedures, which revolve around questions abou t salient elements of sexuality and aging generally, not individually, were also de signed to limit personal information. Additionally, although it may seem counterin tuitive to suggest that face-to-face interviews are an ideal data collec tion strategy for talking with mature women about sexuality, because they are presumed to be more uncomfortable discussing sexual issues, several researchers report th e contrary (Gibble, Rogers, Miller and Turner 1998; Herzog

PAGE 90

78 and Rogers 1988; Levy and Albr echt 1989). Preliminary findings suggest that the utility of self-administered surveys for younger re spondents is not dup licated with older respondents (Gibble, et al. 1998; Herzog a nd Rogers 1988; Levy and Albrecht 1989). Gibble et al. (1998) analyzed data from a large survey experiment investigating the impact of the privacy of self-administered vs interviewer administered survey modes. The experimental results hint that it may be unwise to generalize findings on survey mode effects on younger populations to older popul ations. In-person interviews have the advantage of allowing the researcher to build an empathetic relationship with a respondent in the course of gathering information. The us e of open-ended questions in unstructured or semi-structured format also al lows the researcher to probe for information that may otherwise be neglected, and to clar ify the meaning of responses (Bernard 1995; Levy and Albrecht 1989; Spradley 1979). Sampling Frame As an ethnographic research endeavor that includes cultural consensus modeling this project is an exploration about cultura l patterns that are expected to have high agreement. It therefore does not need a ra ndom sample to serve as a valid measure of cultural pattern. Although each individual asse mbles out of their own experience ways of thinking about those experiences and cons truct an understandi ng of the world around them, their various ideas, emotions, and behaviors intersect in multiple ways with the culture of other individuals that have shared similar e xperiences (Handwerker 2001). This is the point of a sampling frame that encompasses certain life experiencesin this case midlife and older women becoming single again after divorce or widowhoodthat may influence the patterns of social intera ction through which people construct culture that include explicitly measured internal validity confounds (Ha ndwerker 2001: 99). The

PAGE 91

79 number needed to ascertain the cultural domains depends on the degree of cultural consensus in the population considered. Borgat ti notes that "if every informant gives the exact same answers, you only need one," but si nce this is unlikely he suggest computing the frequency counts of free-list data afte r obtaining about 20 or so lists until the frequencies of the top items remain the sa me as informants are added (1999:122-123). The age range, ethnic makeup, marital status and gender specific focus of this project is deemed appropriate based on epid emiological data showing a growing number of women, particularly ethnic minority wo men, in this age range as HIV+ through heterosexual transmission, and due to the lack of prevention programs targeted at people age 50 and older (Rich 2001; Levy 1998; Ory and Mack 1998; Zaboltsky 1998). The age range is also deemed appropriate for cohort comparisons because as women baby boomers enter middle-age, it is predicted that they will be more sexually active and engage in more high risk behaviors than what is occurring in the current older populations (AARP 2003; Levy 199 8; Yates et al. 1999). Five factors combine to make women age 45 and older who join the Displaced Homemakers Program in North Florida an id eal sample and location for investigating patient and physician belie fs about aging and sexu ality, and HIV issues: Florida as a whole is the Bellw ether state for aging issues due to the fact that it is the oldest state in the uni on in terms of demographics, which means that it is an ideal place to study concepts about agi ng and sexuality among physicians and patients; The spread of HIV/AIDS among midlife and older persons, particularly among women and minorities has been increasing over the last decade, and this is particularly true in Florida; The African American population in this area accounts for almost 30 percent of the total population, far exceeding the national average of 12 percent, and the Florida average of 14.6 percent (U.S. Census Bur eau 2001), thus providing a larger sample of minority participants;

PAGE 92

80 Many recently divorced, separated and widow ed women are reentering the arena of sexual activity in an age very different fr om the years before the discovery of HIV and its routes of transmission (Rich 2001) which suggest that their beliefs about aging and sexuality are not only of clinical importance, but that their views may be quite different from that of physicians; and The percentage of persons divorced and sepa rated in this area slightly exceeds that of the state as a whole (United States Census Bureau 2001), although the widowed population is smaller (Table 3-1). Th e number of women in each of these categories exceeds men. All of which suggest s that the issues of this project may be salient for this population. Table 3-1. Disrupted Marital Status of State and Research Community Florida Metropolitan North Florida Divorced % 11.6 13.4 Separated % 2.4 2.9 Widowed % 7.9 6.4 Therefore the ideas about aging and sexuality are relevant and timely for this community, and the findings will serve the clinical communication and HIV prevention needs this underserved population. Additionally, primary care physicians work ing in general practice or family medicine are generally the most frequen tly visited care providers. Drawing physician participants from a hospital-based clinic that serve an older, poor largely minority and mostly female population, was appropriate for addressing the issues in this project. The in-depth approach of this qualitative project made the use of a very large, randomly selected sample prohibitive. Although cultural consensus modeling is an astonishingly powerful method with a small num ber of respondents, the results cannot be generalized to a national cohort of "recentl y single" women between the ages of 45-70, nor to a general population of primary car e physicians. Additionally, relationships among some demographic variables are expected to be anecdotal because of sample size and the lack of applicability of so me standard statistical analysis.

PAGE 93

81 Sampling and Recruitment The DHP women participants for this pr oject were selected through a purposive sampling method aimed at women age 45 to 70 who participated in the state sponsored Displaced Homemakers Program in a large, north Florida city between October 2002 to June 2004. Women were recruited in-p erson at group meetings at the community college where the Displaced Homemakers Program mee ting takes place. I in troduced myself and explained my student status and what my re search was about. I passed around a paper for the women to write their names and contact information if they were interested in participating. I explained that the interviews could occur in their hom es, at the college or an alternative location of th eir choosing. Participants were compensated for their time with a $20 grocery gift certificate as an incentive to participate. The research plan initially proposed to select 60 or more DHP women over the duration of the project (20 African-Ameri can, 20 Latino, and 20 Euro-American women), however the number of Latinas attending the program during the year and a half that I recruited subjects was extremely low, and of those in attendance few met the age (45 or older) and marital status criter ia (divorce, separated, widowed or the equivalent of these statuses for a long-term consensual union). Therefore the final sample of DHP women respondents is 20 self-identifie d Black/African American wo men and 23 self identified White/Euro-American women, and one other. The ethnic group percentages of this sample generally reflect the attendees of this program, except that the Latinas tended to be younger during the research time period. Th e primary demographic difference is that the program caters to women age 35 and olde r whereas this research project included only those age 45 and older. The average ag e of my final sample was 52, whereas the

PAGE 94

82 average for the program tends to be closer to 45 (personal comm unication with program director). (Tables 3-2 and 3-3 show char acteristics of the DHP participants). The research plan initially proposed to sel ect a physician sample of 40 in order to have enough participants to complete the CCM procedures. The final sample ended up at 31 due to the difficulty of recruiting physicia n participants. I ultimately accepted lower than ideal numbers of participants to run the CCM analyses. Thirty-one primary care medical residents and physicians were recruited to assess current physician views and prac tices regarding aging and sexua lity. Almost two-thirds of the physicians participated by completing a short survey, in-per son interview and one of the CCM procedures, and the rest were as ked to participate by completing the survey and a CCM procedures only. For their time th ey received either a $10 or $20 restaurant gift certificate, depending on whether they chose the long or shor t interview format. Research Setting The setting of this research project is not isolated to one small community as is often typical of anthropological research. Rath er the setting of this research occurs in a multitude of locations. The primary entre into the field begins at one of the buildings of the campus of the community college for the DHP women, or at the family practice clinic for the physicians. The community co llege is where the di splaced homemakers program meets for three-week sessions every other month. Table 3-2. Mean Age of DHP Participants, Years in Union and Household Composition. Age at interview Mean (SD) Range Total 52.23 (5.31) 45-68 White/Euro-American 52.48 (5.59) 45-68 Black/African-American 51.95 (5.24) 47-65 Other 52 (NA) NA Years in marital union 18.05 (9.58) 1-34 Household composition 1.97 (0.99) 1-4

PAGE 95

83 Table 3-3. Frequencies and Percenta ges of DHP Women Sample (n=44) Self-Defined Ethnicity Frequency Percentage White/Euro-American 23 52.3 Black/African-American 20 45.5 Other 1 (Middle East Nationality) 2.3 Latino Identity White-Puerto Rican 1 2.3 Non-Latino 43 97.7 Age at interview 45-49 17 38.6 50-55 17 38.6 56-60 7 15.9 61-70 3 6.8 Marital Status Total White Black Total White Black Divorced 22 (1other) 15 6 50 65.2 30 Widowed 11 4 7 25 17.4 35 Separated 9 4 5 20.5 17.4 25 Ended Consensual Union 2 0 2 4.5 0 10 Educational level Total White Black Total White Black Some High school 5 1 4 11.4 4.3 20 High school/GED 8 7 1 18.1 30.4 5 Vocational ED/college 23 (1other) 10 12 52.3 43.5 60 College grad/grad school 8 5 3 18.2 21.7 15 Household composition Total White Black Total White Black Live alone 17 (1other) 10 6 38.6 43.5 30 2 in household 16 10 6 36.4 43.5 30 3 in household 6 1 5 13.6 4.2 25 4 in household 5 2 3 11.4 8.7 15 Self-rated health status Tota l White Black Total White Black Excellent 9 (1other) 6 2 20.5 26.1 10 Good 19 9 10 43.2 39.1 50 Average 8 5 3 18.2 21.7 15 Fair 5 2 3 11.4 8.7 15 Poor 3 1 2 6.8 4.3 10 Employment status Full-time 13 29.5 Part-time 6 13.6 Retired 1 2.3 Unemployed 18 40.9 SSI 5 11.4 Other 1 2.3 Time lived in N.E. FL Entire life 13 29.5 20 yrs. or more 13 29.5 5-15 years 8 18.2 Less than 5 years 10 22.7 Most of the physicians were recruited through a community health and family practice clinic of the universit y-based hospital. The behavi oral science coordinator on

PAGE 96

84 staff there facilitated this recruitment, by introducing me to attending physicians and giving me access to the noon meetings for the re sidents. I was unable to duplicate this useful relationship in any of the other re sidency programs in the area. A few other physicians were recruited through an email announcement that went to primary care physicians in multiple specialties at the hospital, but this so rt of recruitment strategy was of very limited utility and cancellations were much more frequent with those that expressed interest. Interviews with physicians occurred in their office or clinic. The physician sample consisted of 13 attending physicians, and 18 resident physicians. Naturally, attending physicians were on average, older than residents, but they were also more predominately males. Women have begun to predominate in the enrollment in the residency program in recent years, and my sample reflects this trend somewhat with 10 women and 8 men in the re sidency portion of the sample. There has been a trend of increasing numbers of wome n medical doctors since the 1970s and 1980s, but the earlier male dominance in the professi on is seen in the attending physician portion of the sample with only four women attending physicians. Additionally the female participation in my study was low compared to the female-male ratio in the overall residency program, which was just slightly less than 2:1 for 2003 and 2004. (Tables 3-4 and 3-5 show characteristics of the physician participants). Most of the DHP meetings took place in a room located at the end of the hall on the second or third floor, where sl ight ramp leads up to the doo r of the end room. A small vertical rectangular window in the door reveals a large U-shaped table arrangement, and the diversity of women seated at the bottom of the U, who turn to see that me peering through the window in the door. The room is bright and large with windows flanking the

PAGE 97

85 South and East sides of the room. Invited in by the program coordinator, I would sit quietly off to the side of the tabl e waiting for a break in the program. Table 3-4. Frequencies and Percentages of Physician Sample (n=31) Full interview or Survey Frequency Percentage interview/survey 18 58.1 Partial interview/survey 1 3.2 Survey only 12 38.7 Attending/Resident Attending Physician 13 41.9 Resident Physician 18 58.1 Resident Physician Detail 1st year resident 8 25.8 2nd year resident 7 22.6 3rd year resident 3 9.7 Sex/Gender Female 14 45.2 Male 17 54.8 Self-reported Ethnicity White/Euro-American 17 54.8 Black/African-American 2 6.5 Asian/Pacific Islander 7 22.6 Other 5 16.1 Latino Ethnicity Cuban 1 3.2 South American 2 6.4 Non-Latino 28 90.3 Other Ethnicity Guyanese-West Indian 1 3.2 Hispanic (South American) 1 3.2 Venezuelan 1 3.2 Iranian 1 3.2 Pakistani 1 3.2 Table 3-5. Mean Age of Physicians a nd Year of Medical School Completion Age at interview Mean (SD) Range Total (N=31) 39.42 (8.71) 27-57 Attending (N=13) 47.62 (6.37) 39-57 Resident (N=18) 33.50 (4.09) 27-42 Year completed med school 1991.87 (9.64) 1972-2002 Once the program director or coordinator of the class found a pl ace to take a break, she introduced me by name and my UF affiliati on. I passed a one-page description of the project with my contact information around to all of the participants and then proceeded to summarize the information on the handout by te lling them the title of the project, what

PAGE 98

86 they would need to do to participate and how long it will take, the age and marital status criteria and that they would be compensate d with a $20 grocery store gift-certificate. Then I passed around a sheet of paper and as ked that anyone that was interested in participating give me their name and phone number. I always brought a copy of the informed consent with me for anyone who want ed to read it over. I also let them know that if they were uncertain at the moment that my cont act info was on the handout and they could contact me in the future if they wanted to. This only happened once; usually I had to call the participants to find a time to meet. Most of the women attending the displ aced homemaker program were suffering either emotionally, financially, and/or spiritu ally, but they came to this free program to get their lives back on track, or to learn new skills, or to simply improve their selfesteem. I attended several of the class sessions, but each group differs from the one before it, just like any teacher knows that ev ery class has its own co llective personality. All of the women I spoke with said that th ey benefited from the program in some way. Once in a while, when I visited the program to recruit new partic ipants, I would get a very inquisitive group that asked lots of questions, like why I am doing the research, why the specific ages, what will I do with my degree once I have it and so on. Other times, there would be no questions and only one person signed-up. I found it useful when the timing worked to my advantage to ask if anyone would like to schedule a time to meet right then, though this was not always appropriate and I tried very hard not to interfere with the program. The field setting changed every time I scheduled an interview. Scheduling interviews often proved difficult since many were not available when I called and several

PAGE 99

87 never returned my phone calls. This was true for both DHP wo men and physicians. Sometimes I visited the homes for the particip ants, other times we met at the college or some other location. The physicians were all interviewed in their offices. The homes I visited in this research en compassed a range of economic levels and the locations stretched to the outer edges of the service ar ea of the displaced homemaker program. Most of the participants live in urban or urban-suburban areas whereas only a handful lived in more rural areas of the regi on. In some cases I visited homes in up-scale gated communities, others lived in very m odest homes and apartments. Sometimes the homes were spotless, and it was obvious that the occupants had taken great care in decorating their home. Other times the homes we re in total disarray (perhaps a testament to the difficult transition some of the women were undergoing) and I was instructed to climb over boxes to get to a seating area. Howe ver, not all of the in terviews took place at participants homes; I met one woman at her place of work on her lunch break, and about a third of the DHP participants chose to m eet me in the building where the DHP classes took place. This was convenient for both of us since we knew the place and the kind people at the community college helped us find a quiet place to talk. Unlike the interviews with the DHP women, the physicia n interview location was fairly static. I interviewed all of the physicia ns at their place of work, whether this was their clinic office, hospital office, or in the ca se of the residents, in a quiet meeting room at the clinic or the office of the behavioral science coordina tor that was assisting me. As for the physicians that only completed the su rvey and CCM procedure, I met with them when they had a few minutes between meetings or patients at the clinic. I also was given access to 20 minutes of one of the noon meetings and convinced six residents to complete

PAGE 100

88 the survey and the rating procedure. It wa s not uncommon for the physician to delay or cancel our scheduled meetings. I made several tr ips to the clinic and hospital that yielded little more than a chance to reschedule. Most of the physicians in my study were residents or attending physicians in a family practice clinic that is in a larg ely African-American neighborhood. The building consists of two wings, one side is the clinic itself and the other is where the residency program resides. The clinic itself opens w ith the reception staff to the right and with a large waiting area to the right. The waiting ar ea is stocked with countless brochures on every conceivable health issue. There are two TVs and a play area filled with toys for children. On my first visit there, the coordi nator who was assisting me gave me a grand tour and introduced me to several physicians. Most of the physician offices are located on the residency side, and this was my primary point of entre. The interview process typically followe d the same pattern for both DHP women and physicians. For the DHP women, I gave th em the informed consent to read and sign, explaining any elements that they needed clarifie d. I always offered to read it to them if that was their preference, to save them any embarrassment if they had difficulty reading. Once the informed consent was taken care of, I proceeded to the survey, which usually only took a couple of minutes, then I began the vignette-based intervie w. Before reading the first story, I made clear that the stories were fictional, that there were no right or wrong answers to the questions that followed th e stories, and that they could make their responses as personal or impersonal as they wished. All of the in terviews were audio taped with the participants permission, and initialed on the informed consent. The interview section of the encount er usually took from 30 minutes to an hour. One of the

PAGE 101

89 CCM procedures concluded the encounter. The interview process with the physicians followed the same format, more or less, though the focus on the informed consent was not usually needed and they process went qui cker since I only read them one vignette. Almost every woman I interviewed told me that she enjoyed the process much more than she expected. It was not uncomm on for the women to divulge intimate details about themselves and their lives. Many just seemed happy to have someone that was interested in what they had to say. The next chapter begins with the analysis of the data gathered. Chapters 4, 5 and 6 present the research findings. Chapter 4 presen ts part one of the data analysis, focusing on sexuality conceptualizati ons using the Cultural Consen sus Modeling techniques. I provide a description of the types of analysis used and the theory behind them. These data are presented textually and graphically for th eir own merits. In chapters 5 and 6 these data are linked to the data from the in -depth interviews. The data on sexuality conceptualizations are consider ed in terms of their support of the hypotheses, as well as their shortcomings for support of the hypotheses.

PAGE 102

90 CHAPTER 4 CULTURAL CONSENSUS DATA AND ANALYSIS Cognitive scientists believe that the less tang ible aspects of culture, such as beliefs and ideas, can be made more accessible with cognitive models (Borgatti 1997). Borgatti asserts that mathematical models repres enting cultural domains can make accurate predictions about particular cultural groups, but notes th at every representation is uniquely situated in the current concepts and language of the groups. He also notes that representations are just that--representa tions of cultural models not the models themselves. In keeping with Borgatti and other cognitive scientists, Dessler and Bindon (2000) contend that it can be inferred from these representations that individuals are relying on a shared model of the same cultural domain. Free listing is frequently the first step in gathering data about a cultural domain. It is often the prelude to other t ypes of data collection such as piles sorts, triads, and other systematic data collection techniques (Borgatti 1996b, 1999; Weller and Romney 1988). However, the free list itself also provides useful data on its own (Bernard 1996). For example, free listing identifies the parameters of the cultural domain using both the frequency of terms and their salience, determin ed by their rank-order in individual lists. Free lists are gathered from informants to de termine if individuals are using some of the same terms to describe the domain, and are therefore sharing cultu ral knowledge about the domain; they help to determine what be longs in a domain and how it is structured (Weller and Romney 1988). Free listing is an emic task since it draws on the language, ideas, and categories of those be ing studied. It is generally assumed that the more salient

PAGE 103

91 terms are also the ones that are listed most frequently (Bernard 1996; Borgatti 1999). Additionally, salience is also represented by list order, with more salient terms listed closer to the beginning of lists. In the case of this research project, I as ked a set of respondents to "tell me all the words and phrases they could think of to de scribe sexuality for women age 45 and up? This was a systematic way of eliciting the em ic concepts and categories that belong in the domain of sexuality for midlife and older wo men. I asked this of the women themselves and then asked the same question to a group of Family Practice physicia ns to see if they were using some of the same descriptors. Th e free lists provided only the parameters of the domain. To determine how people perceive the similarity among items in a domain, that is, the way that the domain is confi gured, I use piled sorts, and a ratings test. Free listing The initial free-listing procedure with both the DHP women and the physicians yielded a total of 271 terms and short phras es from 32 respondents (183 terms from 18 DHP women and 88 from 14 physicians). The fr ee-list data was entere d into a text file and analyzed for frequency in ANTHROP AC (Borgatti 1996a) using a low soundex setting to highlight similar sounding words th at may be collapsed under one term. This list was reduced to 209 when synonyms and rela ted items were taken into consideration, with 48 items repeated more than once (T able 4-1). For example, several of the respondents stated either the phr ase lack of desire or dec reased libido, both of which I collapsed under the phrase l ack of desire for the 48 repeat ed items later used in the pile-sorts and ratings tests. Although terms and phrases related to decreased sexual activity were common, several respondents al so mentioned that sex-life improved with age with terms and phrases including sex better than when in 20s, sex is better,

PAGE 104

92 hotter, more pleasurable, and so on. Th ere are perhaps many more synonyms within the 48 terms and the remaining terms men tioned by only one indi vidual, however, I preferred to err on the side of caution and limit the combination to only very obvious synonyms. Many of the terms describing some level of improvement in sexuality appear to be more salient in terms of their rank orde r in the individual lists. Smiths Salience (Table 4-1) is automatically calculated in ANTHROPAC (Borgatti 1996a). Smiths Salience calculates two relevant measures of sa lience: 1) is the position of a given item on a list (e.g., first, second, th ird and so on); and 2) the fr equency that the item appears across all of the free lists (Weller and Romney 1988). Repeated frequency is the most commonly used measure of salience, but studi es looking at these measures have shown that terms that appear earlie r on a list are usually more important or better known than items further down the list (Borgatti 1999; Weller and Romney 1988). Smiths formula uses both the order in each i ndividual list and the frequency of the item in the lists to get the gross mean percentile rank of each item across all the free lists. For the results presented here, lack of desire has the hi ghest salience rank, but more pleasurable comfortable and freeing are al so highly salient to these in formants. The salience of these terms and why they are more salient is better considered with the other CCM techniques that follow. I continued to collect informal free list data by asking my informants the same free list question above, but did not urge them to gi ve me this response in list form. Rather some gave me lists of terms, while others de scribed their ideas in sentence form and in this case word list was abstracted from their statements with their assistance. From this

PAGE 105

93Table 4-1. Free List Frequencies with Both Samples. Item Freq Resp PCTAvg Rank Smith S 1 lack of desire 8 25 3.5 0.185 2 comfortable 3 9 2 0.078 3 more pleasurable 3 9 1.333 0.089 4 non-existent 3 9 5.333 0.076 5 more responsible 3 9 5 0.037 6 better 3 9 4 0.069 7 fun 3 9 4 0.058 8 sexy 3 9 5.667 0.044 9 freeing 3 9 2 0.078 10 renewed interest 2 6 8 0.024 11 happy 2 6 5 0.023 12 more mature 2 6 12 0.037 13 intimacy 2 6 4 0.044 14 don't bother 2 6 7 0.031 15 satisfying 2 6 2 0.055 16 menopausal problems 2 6 3.5 0.034 17 monogamous 2 6 6 0.023 18 desirable 2 6 3 0.05 19 menopausal 2 6 2.5 0.056 20 caring 2 6 5 0.043 21 use protection 2 6 3.5 0.045 22 hotter 2 6 10 0.029 23 be left alone 2 6 7 0.023 24 disease risk 2 6 8 0.021 Item Freq Resp PCTAvg Rank Smith S 25on-going 2 6 1 0.063 26ignorance 2 6 15.5 0.028 27talks freely 2 6 3 0.049 28medical effects on sexual function 2 6 2.5 0.047 29less active 2 6 4 0.027 30beautiful 2 6 2 0.058 31body changing rapidly 2 6 9 0.04 32more careful 2 6 9.5 0.017 33been around the block 2 6 7.5 0.029 34kind of lonely 2 6 8 0.026 35comfortable in their own skin 2 6 18 0.009 36dangerous 2 6 3 0.042 37explorative 2 6 8 0.021 38confident 2 6 3 0.052 39sex better than when in 20s 2 6 6 0.039 40got to have somebody 2 6 3.5 0.048 41uncomfortable 2 6 3.5 0.036 42special 2 6 3.5 0.041 43sensitive 2 6 3.5 0.045 44good once a week 2 6 3.5 0.043 45too busy 2 6 4.5 0.026 46enjoyable 2 6 3.5 0.053 47companionship with financial help 2 6 8.5 0.037 48uninhibited 2 6 3 0.042

PAGE 106

94 continued questioning I obtained 27 more lists from both samples and it was heartening to see that several of the terms found in th e first procedure were the same. When these new lists were adjusted for synonyms and combin ed with the original free list data, there were 55 repeated terms and a total was 325 wo rds and phrases. The new repeated terms were in-a-rut, compatibility more important than sex, mature, hot flashes, painful, less sexually active and more sel ective. These were not new terms, rather they were terms mentioned by only one person in the first group. The original shortened list of 48 repeated terms was used to analy ze the domain further by incorporating it into the pile-sort and rating procedures. Pile Sort Procedure Pile sorts are tabulated and in terpreted by measures of sim ilarity, that is each time a respondent puts two items together in the same pile, it is counted as an indication that the respondent believes those two items are similar (Borgatti 1999). Pile sorts are used to elicit the structure of the domain. The mo re respondents that put any given two items together the stronger the interp retation that they are draw ing on one cultural domain and that they perceive it in similar ways. By us ing an item-by-item matrix the pile sort data can be represented spatially with non-metric multidimensional scaling (MDS) and cluster analysis (Borgatti 1999). The MDS serves to reveal the perceptual dimensions that individuals use to distinguish one item from another. The cluster analysis groups items together that share attributes or in some cases constitute subdomains (Borgatti 1999). For the pile sorting procedure, respondent s were asked to organize the 48 terms and phrases--which had been written on indivi dual cards and numbered on the back for coding--in what ever way seemed appropriate to them. This approach is referred to as single unconstrained pile so rts (Weller and Romney 1988). Unfortunately this method

PAGE 107

95 results in the lumper/splitter problem, wh ere, for example, two people have similar views on which items go together, but one takes the time to sort out the finer distinctions between the items and makes numerous piles, and the other makes only a few piles with the broadest distinctions (Bor gatti 1999). This problem can skew the results since the differences between the lumpers and splitters can eclipse the other differences among the pile-sorts (Weller and Romney 1988). Howeve r, efforts to avoid this problem are time consuming and sometimes complicated for th e respondents. Additi onally, some of the more constrained methods can negate the emic nature of the unconstrai ned pile sort since they impose a structure that may be inappropr iate for the domain. I opted for less precise data on this point since I was as king participants to also enga ge in a long interview and to fill-out a survey. I collected 21 pile sorts (15 from DHP wo men and 7 from physicians). This is a small number of pile-sorts since a sample of 30 to 40 is recommended to reach reliabilities above .90 (Weller and Romney 19 88). However, time constraints and the difficulty in obtaining a larger physician samp le limited my options in this regard, particularly since I wanted the opportunity to run ratings tests on the sample as well. This low number of pile sort s, particularly with the phys ician sample, is somewhat ameliorated by the fact that physicians as a group are expected to have more internal consensus as a result of their training and so cialization as health care providers. The analysis below appears to support this contention. When there was time, I asked respondent s to tell me why they grouped certain words and phrases together, however there wa s rarely time for such questions with the physicians. Several of the women mentioned that they put words together that described

PAGE 108

96 themselves or grouped words together that described others. For example, several mentioned that they separated out terms that described themselves (i.e., relates to me, me at this age, etc.), particularly rela ted to relationships, ex plaining that a pile described a new relationship af ter divorce or some other even t. Other pile descriptions included words that described th ings that are bad, wrong or careless; piles that include things that are happy or good; a nd no reason piles. These descriptions were very helpful in deciphering the multidimensional scaling (MDS) maps and the cluster analysis for the DHP women. Multidimensional Scaling of Pile Sort Data Multidimensional Scaling (MDS) shows simila rities by placing items together that are similar and interpre tation relies on looking for patterns in the clusters of items. The pile sort data imported into ANTHROPAC produced an item-by-item aggregate proximity matrix; each cell indicating the number of times a given two items were piled together. The algorithm for multidimensional scaling in ANTHROPAC computes the euclidian distances (geometric distance in multidimensional space) among all the pairs in the matrix and adjusts coordi nates of each point in dimensional space that maximally reduces stress (Borgatti 1996b). MDS measures similarity in stress, which is the measure of how far off the graph is from one that is perfectly proportional (Bernard 1996: 502). MDS is typically run in two dimensions, since it is easier to read and manage the data, however it is usually no t perfect and typically creates more stress than multiple dimensions. Borgatti notes that it is not necessary to have zero stress in an MDS map for it to be useful. The rule of thumb we use is that anyt hing under 0.1 is excellent and anything over 0.15 is unacceptable. Care must be exerci sed in interpreting any map that has nonzero stress since, by defin ition, non-zero stress means th at some or all of the

PAGE 109

97 distances in the map are, to some degr ee, distortions of the input data. In general, however, longer distances tend to be more accurate than shorter distances, so larger patterns are still visible even when stress is high. (Borgatti 1996b: 33). However, Sturrock and Rocha (2000) note that the stress values they generated from their research using random matrices produced mu ch higher levels of stress with only a 1% chance that the objects were arranged randoml y. They showed that as the number of items increased in two dimensions that the st ress values began to level off so that the stress generated for 100 items was not much wo rse than the stress for 40 items. The stress for each of the matrices of this project fall below the 0.15 cut-off noted by Borgatti (1996b). When the analyses were run in three and four dimensions for each sample the stress reduced as expected and each permuta tion reduced the stress by about a one-third. MDS Results The MDS map derived from the aggregate proximity data of the DHP women pilesorts appears in Figure 4-1 and the proximity data of the physicians appears in Figure 4-2. The stress in two dimensions for Figure 4-1 is .129 and .095 for Figure 4-2, below the accepted criterion noted by Borgatti, and refl ects the degree of agreement among the participants. These two representations are strikingly similar in the way that each group organized the words in much the same way, though with a few significant differences. Generally, clusters of items appearing on the map very close together tend to represent words that are more semantically si milar, whereas clusters with words that are more dispersed tend to represent items that are related topically (Stansbury and Sierra 2004). The assumption is that proximity repr esents aggregate psyc hological relatedness, clustered items frequently pointing to near by items (Stansbury and Sierra 2004: 460). For this project the cluster analysis (see ne xt section) served as an ideal companion to the MDS analysis, since the items appeared to be ordered similarly. Additionally, in

PAGE 110

98 order to get a better picture of how and why terms were grou ped together, I dissected the pile sort data into groupings based on what women had told me each pile meant to them. I then ran cluster analyses on topics like new relationship or ca reless behavior and so on, so I could see exactly which terms were grouped together under which category and how closely they were connected. To accomplish this I had to allow items to appear more than once and allow for missing items. This use of clustering proved very helpful in identifying meaningful clusters in the MDS. In the lower left portion of Figure 4-1 are core terms that most participants grouped together to varying degrees of inclusion th at were described as new relationship characteristics, and as terms that described thei r sexuality at this time in their life (or their ideal for it). Terms such as happy enjoyable satisfying etc. were all grouped closely together and share a similar seman tic meaning for describing what the women wanted to find in their relationships after di vorce, or what they had already found with new partners after divorce or widowhood. The inclusion or exclusi on of various terms reflects the intracultural variation in their relationship experiences. Since some women expressed no interest in de veloping a new relationship, thes e terms and phrases reflected their sense of what is ideal in a mature sexual relationship. In interviews, women were not reluctant to express concern about the risk of sexually transmitted infections from new partners, which explains why use protection, more careful and more responsible are in this grouping. It was not uncommon for respondents to describe themselves as beha ving in a safer manner than women in younger cohorts. Also, the combining of such terms as happy, sexy and beautiful with use

PAGE 111

99 protection and more careful points towards the tension between the ideals of intimate relationships and the real ities of the possible risks involved with dating. In the upper right corner of Figure 4-1 are terms that were often grouped together to describe careless behavior. Interestingly, the phrase been around the block was grouped in this area perhaps because for some it de notes promiscuity. However, the women who included it in their free-lists appeared to be expressing experienced without moral connotations. The phrases in this groupi ng were not typically associated with themselves; rather they attributed these to other women, and often to younger women. In the lower left of Figure 4-1 are negative descriptors, and typically terms they referred to as not me or bad. Phrases such as too busy or dont bother were described as excuses other women use when they do not want to engage with a partner or with life in general. Intere stingly, physicians groupe d many of the same words together as the DHP women, although there are not able differences. In the physician MDS map, (Figure 4-2) the stress is lower even though the map is based on only six pile sorts. Since physicians are expected to be more homogenous than are the group of women, this is not surprising. The grouping in the upper ri ght corner of Figur e 4-2 represents patients who are poorly informed about STD risks or behave re cklessly. Interview and survey data with the physicians confirms that they believe that many of their patients are poorly informed in this regard, and that certain subgroups of their patients live dangerous life-styles (Chapter 6). These are the same terms that the DHP women grouped together that some described as careless behavior, sugges ting that the women and the physicians

PAGE 112

100 better special happy satisfying beautiful enjoyable more pleasurable caring more mature talks freely comfortable use protection sexy desirable sensitive fun confident renewed interest more careful more responsible monogamous explorative intimacy sex better than when in 20s hotter body changing rapidly menopausal less active good once a week non-exisitant lack of desire companionship with financial help freeing disease risk got to have somebody dangerous been around the block ignorance uninhibited medical effects on sexual functioning menopausal problems comfortable in their own skin on-going too busy don't bother uncomfortable be left alone kind of lonely CARELESS NOT DESCRIBING THEMSELVES and/or TERMS DESCRIBED AS "WRONG" OR "BAD" DESCRIBING IDEAL NEW RELATIONSHIP --AFTER DIVORCE and/orATTHIS AGE Figure 4-1. MDS for DHP Women Pile Sorts.

PAGE 113

101 intimacy caring comfortable more mature sensitive comfortable in their own skins confident special happy desirable renewed interest fun talks freely sexy beautiful menopausal problems medical effects on sexual functioning lack of desire be left alone non-existant too busy less active disease risk ignorance dangerous got to have somebody been around the block body changing rapidly good once a week use protection more careful menopausal don't bother uncomfortable kind of lonely companionship w/ financial help enjoyable freeing monogamous more pleasurable more responsible satisfying on-going sex better than when in 20s hotter explorative uninhibited OLDER WOMEN PATIENTS IN RELATIONSHIPS COMPARED TO YOUNGER PATIENTS MENOPAUSAL PATIENTS WHOSE SYMPTOMS ARE CAUSING CONCERN and/or NEGATIVELY AFFECTING THEIR SEXUALITY PATIENTS WHO ARE POORLY INFORMED ABOUT DISEASE RISK and/or BEHAVE RECKLESSLY Figure 4-2. MDS for Ph ysician Pile Sorts.

PAGE 114

102 conceptualize this part of the domain in much the same way. The word uninhibited is closely positioned with this grouping in both MDS figures, but is closer to the relationships theme for the physicians and cl oser to careless theme for the women. The lower left area of Figure 4-2 appears to describe relationships for older women patients, and the cluster contains many of th e same terms as those in the DHP womens grouping for new relationships. Again there ar e a few notable differences between the two MDS representations. The physic ians do not appear to be grouping use protection and more careful within this topic and this refl ects their emphasis on younger cohorts when considering STD and HIV risks. When physicians were asked which of their patients they considered at-risk for HIV, they overwhelmingly focused on those in their 20s and 30s, with the perception of risk declining with age. The two groups appear to be furthest apart on their notions regarding menopause and the terms and phrases that combine with it. The physicians have a coherent grouping of terms for menopausal patients who are having difficulties or discomforts, with menopausal, menopausal problems, medica l effects on sexual functioning and body changing rapidly grouped together. The DHP wome n did not develop a coherent grouping on this topic, with the terms listed a bove widely dispersed, suggesting that there may be wide-ranging experiences regardi ng menopause. This may be due to age differences since the average age of those doi ng the pile sort task is 52, ranging from 46 to 61 years of age. With th is age distribution it is likely that some women have not begun to experience the menopause transition, some experiencing changes th at may or may not be uncomfortable, and others are past it. Th ere were no significant differences in how the words were arranged when the respondents were divided into ol der and younger groups

PAGE 115

or when the two major ethnic groups were compared, which may be due to the small sample size. Figures 4-1 and 4-2 are remarkably sim ilar, suggesting that the two groups conceptualize the domain in much the same way. The representations of the cluster analysis below closely resemble those in ea ch MDS, and lend support to the notion that the groups are in agreement. Clustering of Pile Sort Data Cluster analysis is another descriptiv e tool for exploring relationships among matrix items; it shows which items go togeth er and in what order they go together (Bernard 1995). Much like MDS, analyzing a cluster is somewhat subjective since it groups items together, but does not reveal what the clusters mean. Cluster analysis is used as an exploratory tool, and since th e algorithm systematically finds the most significant solution possible, tests of statis tical significance are not necessary (StatSoft 2004). Johnson's Hierarchical Clustering in ANTHROPC computes the distance (or similarity) between all the ite ms (Borgatti 1996b). For this analysis I used average link clustering. Borgatti explains, In average-link clustering, we consider the distance between one cluster and another clus ter to be equal to the average distance from any member of one cluster to any member of the other cluster (Borgatti 1994: 78). That is, the distance between two clusters is calculat ed as the average distance between all pairs of items (StatSoft 2004) that were gr ouped together in the pile sorts. In the hierarchical clustering output fr om ANTHROPAC, the columns represent the items from the pile sorts and the rows indica te the level of relatedness between the items. Figure 4.3 represents the cluster analysis fo r the DHP women, and Figure 4-4 represents

PAGE 116

the clusters analysis for the physicians. Th e 'Xs' mark the place between any given two columns in a given row when corresponding item s are merged together (Borgatti 1994). The lower any two items are c onnected along the rows the fu rther apart they are from each other. In Figure 4-3, there is a marked divisi on about midway along the columns between #47 sensitive and #14 medical effects on sexual functioning This first large cluster in shaded in pink includes all the same terms th at were grouped toge ther in the MDS under new relationship. Clearly some items are gr ouped together in the pi le sorts by most of the women. For example, #34 enjoyable, #12 caring, and # 39 happy we re piled together by most participants. Whereas other terms we re not grouped together with these other terms as frequently, such as #3 more responsible and #10 renewed interest though they are still part of the overall cluster. The next sharp division is between #46 disease risk and #29 less active This last cluster shaded in yellow contains the terms that were described as bad or wrong in the MDS. Less active (#29) is connected with this gr ouping at a very lo w level indicating that it was not always considered similar to the other items in the group. One obvious difference between the MDS and the cluster analysis appears to be the connection between terms related to menopause. On th e MDS map in Figure 4-1 these items did not make a coherent grouping, whereas in the cl uster analysis they are grouped together. However, this cluster shaded in peach links #14 medical effects on sexual functioning and #15 menopausal problems together at a high level, but th e other items are at a lower level of connectedness, which shows that th ey were not grouped together by everyone

PAGE 117

105 Figure 4-3. Hierarchical Clustering of DHP Women

PAGE 118

106 Figure 4-4. Hierarchical Clustering of Physicians

PAGE 119

107 participating, supporting the contention that this issue is experienced differently among this sample. The green cluster represents the carele ss theme that was shown in Figure 4-1, where #36 dangerous and #46 disease risk were not closely positioned but were part of the same group. In the cluster they are linked at a relatively high leve l showing that items that are positioned tightly together in the MDS may not be as important as would be expected in terms of showi ng patterns of agreement. Figure 4-4 represents the cl uster analysis for the physic ian sample. Just like in Figure 4-3, this cluster analysis follows the si milar patterns as those found in the MDS in Figure 4-2. The distinctions that are appare nt in this cluster are the significant sub categories within the clusters and the links wi th peripheral items not grouped together in the MDS. The cluster with the relationship theme is shaded in two shades of pink to show an interesting division between th e first grouping that ends with #43 explorative and the next one that starts with #21 use protection In the physician MDS (Figure 4-2) use protection was near the relationship grouping, but did not appear to be part of it. The cluster analysis here shows that it is part of this grouping, yet it is linked at a very low level, suggesting that it was infrequen tly combined with these other relationship terms. The peach cluster representing part of the menopausal issues grouping presents a similar case as the one above, with #44 good once a week weakly connected to this cluster and not part of the grouping in the MDS (Figure 4-2). It also shows how the physicians were consistently in agreement about the clearly menopausal terms with #14 medical effects on sexual functioning, #37 body changing rapidly #15 menopausal

PAGE 120

108 problems and #33 menopausal all together at a very high level of connectedness. There was much less agreement about the other term s in this cluster as all were weakly connected to the menopausal terms, which was not clear in the MDS. Unlike the DHP women that linked #36 dangerous and #46 disease risk at a high level of connectedness, the physicians dem onstrate much less agreement on these terms with low levels of connectedness. Compared to the DHP women, the physicians appear to be in agreement about which items are si milar since the cluste r analysis shows high levels of connectedness in almost all of th e clusters. Although the MDS and the cluster analysis appear to indicate that the p hysicians and the DHP women share similar concepts about the domain, the ratings test reveals that they do not share the same culture. Ratings Procedure The ratings procedure shows the pattern of agreement, or consensus, among respondents in order to make inferences about their knowledge of the domain. In this case, I asked 24 respondents (14 DHP women an d 10 physicians) to rate the 48 items on a scale of 1 to 4 in terms of relevance for de scribing sexuality for women age 45 and older (1=least relevant, 4= most re levant). Three of the women who were asked to complete the free listing exercise were also as ked to complete the ratings test. The notion of relevance as a critical di mension that respondents were using to organize the items was revealed in the pro cess of conducting the pile sort procedure. Several of the DHP women participants, and a few of the physicians, noted that some of the terms seemed very apt for describing wo mens midlife sexuality, while other terms were considered inappropriate. Some terms were easily organized in to piles relevant to some aspect of womens sexuality, whereas it was a struggled to find a suitable pile for

PAGE 121

109 other terms. The analysis of the data from th is procedure turned out to be useful in many ways. The analysis of these data uses Cultural Consensus Analysis, a type of reliability theory that assesses the patte rns of agreement among individu als. The first step in the analysis was to create a matrix of correla tions between people, based on their responses to the ratings test. In cons ensus theory individuals are used as variables and their responses are used as cases, whereas typically individuals are cases and their responses are the variables (Weller and Romney 1988). Th is is a form of relia bility testing using the Spearman-Brown Prophesy on respondents ra ther than items. The individuals responses are coded as RESPONSE data, that is a 1 if the informant said yes and a 0 if the informant said no rather than as PERFORMANCE data, which measures item reliability, coding the data as 1 for correct and 0 for incorrect (Weller and Romney 1988: 75). In item reliability, it is assumed th at the correct answers are known, whereas in consensus theory, the correct answers are unknown and the responses from the participants are used to determin e the culturally correct answers. The program in ANTHROPAC uses the re spondents input in the form of a correlation matrix to create a key to the culturally correct responses, and factor analyzes the pattern of respondent-by-respondent similarities to show if there is support for the consensus agreement (Borgatti 1996b). Agreement between any two respondents (similarity of their answers) is held to be a function of the extent to which each has knowledge of the culturally appropriate re sponses: those who responded similarly to others in the group were given more weight in the analysis (Weller and Romney 1988).

PAGE 122

110 Cultural consensus is said to exist when the ratio between the eigenvalues of the first factor is three times larger than the sec ond factor or greater, in dicating a single factor solution. The factor loadings are a correla tion of each respondent with the underlying pattern of cultural knowledge about the domain. Therefore, if the factor loading on the first factor is less than th ree times the loading on the s econd factor, then the results indicate the presence of more than one cu lture and the consensus model does not apply. The One Culture assumption is inconsistent with the existence of more than one large eigenvalue. Two large eigenvalues, fo r instance, is strong evidence that (at least) two truths (two syst ematically different patterns of responses) are governing the responses of informants. The program pr ints the ratio of the first eigenvalue to the second. The rule of thumb is that if th e ratio is less than 3 to 1, the assumption of One Culture is indefe nsible. (Borgatti 1996b: 45) Not surprisingly, the consensus analysis resu lts for the sample as a whole (physician sample and DHP women sample combined) did not support the assumption of one culture. The first factor eigenvalue was 2.67 and the s econd was 2.45, suggesting the presence of two cultures. The first factor explained only 45 % of the variance, and the second factor explained 39%. The average respondent knowledge was extremely low at .129 (SD .308). This was not unanticipated sin ce my original hypothesis was that women and physicians would not con ceptualize the domain in the same way. Although the pile sort data appeared to show th at the two groups were organizi ng elements of the domain in a similar fashion, the ratings data dispute that finding. When the samples were considered se parately, the DHP women had sufficient internal consensus, but the physicians did not The consensus for the physicians violated the rule of the 3:1 ratio on th e factor loadings, although they had more internal agreement with each other than when their responses were combined with the DHP women. The factor loading for the physicians was 1.895 on factor one, and .793 on factor two, leaving

PAGE 123

111 a ratio of 2.389 to1.940. This violation of one culture suggests that individual physicians conceptualize the domain differently, indicati ng that there are two or more subdomains among this group. It was suspected that this was due to age or gender differences in the sample, but closer inspection of the result s did not reveal a pattern along these lines. According to the consensus analysis for the DHP women who participated in the ratings procedure, there is th e existence of one culture. Th e ratio of the first factor eigenvalue 2.403 to that of the second f actor 0.797 is 3.016 to 1.203, indicating a single factor solution, but not a very strong one. The first factor accounts for 62.2% of the variance. Fourteen respondents were intervie wed for this procedure and the average level of expertise was .33 (SD .25) on factor one, w ith the pseudo-reliability coefficient of 0.638 for this application. Given that these data support the one cu lture theory, I looked at some of the nuances in the data for clues about how th e domain is structured. There is more agreement among the DHP women on certain po ints than is immediately obvious from the analysis above. Even when the women di d not exactly agree on th e ratings level of a certain term, they tended to consistently rate it at the higher or lower end of the 4-point rating scale. For example, all of the DPH respondents indicated that caring was very relevant to describing sexuality for women ag e 45 and older, with 13 of 14 checking level 4 (highest possible rate) and one checking level 3. Although only nine indicated that more responsible merited level 4, the rest selected level 3. Table 4-2 shows the terms with the most agreement among the women. Table 4-2 shows the overall level of agreement on selected terms, as well as giving clues as to which terms were most salient thought their shared agreement. Comparing th e terms they agree on with those that the

PAGE 124

112 physicians agree on provides clues as to how they are organizing elements of the domain (Table 4-3). Table 4-2. Terms Rating Highest Level of Agreement for DHP Women N=14 Rating level/# selected Term or phrase 1 2 3 4 Caring 0 0 1 13 More careful 0 1 1 12 Sensitive 0 2 2 10 More responsible 0 0 5 9 More mature 0 0 6 8 Special 0 1 4 9 Enjoyable 0 2 3 9 Fun 0 1 8 5 Satisfying 0 1 9 4 Ignorance 9 2 3 0 Another noteworthy finding in the consensus an alysis of the ratings data is that the DPH women appear to be cons idered more decisive. The women were much more likely to rate a term or phrase as a 1 or 4, whereas the physicians were more likely to use the mid-range levels of 2 or 3. Table 4.3 s hows an item-by-item comparison of the DHP Women Answer Key with the Physician Answer Key (labeled Med Key), generated with the consensus analysis in ANTHROPAC. Table 4-3. Comparis on of Ratings Keys Item 1 2 3 4 5 6 7 8 9 10111213141516171819 20 21 22 2324 DHP Key 2 4 4 1 4 4 4 4 2 4 4 4 4 3 3 1 4 3 1 4 4 3 1 4 Med Key 3 3 3 2 3 3 2 2 2 3 3 4 3 3 3 3 3 2 3 3 2 3 2 3 Item 25 26 27 28 29 30 31 323334353637383940414243 44 45 46 4748 DHP Key 1 1 4 4 1 4 4 4 2 4 4 1 3 4 4 1 1 3 3 1 3 4 4 2 Med Key 3 2 3 2 3 3 3 3 4 3 2 2 3 3 3 2 2 3 3 2 3 3 3 2 Perhaps the most significant as pect of the ratings data is what it conveys about the how much DHP women and physicians differ a bout the nature of sexuality for midlife

PAGE 125

113 and older women. Both groups were in complete agreement on 10 of the 48 terms as shown in Table 4-4, which list the terms of agreements and how they were rated. This suggests that physicians are so mewhat aware of what is re levant for aging women and their sexuality. However, the two groups we re in marked disagr eement on another 10 of the 48 terms, shown in Table 4-5. Table 4-4. Terms of Agreement between Physicians and DHP Women Item # Item Rating Level 9 Dont Bother 2 12 Caring 4 14 Medical Effects On Sexual Functioning 3 15 Satisfying 3 22 Uninhibited 3 37 Body Changing Rapidly 3 42 Beautiful 3 43 Explorative 3 45 Companionship With Financial Help 3 48 Too Busy 2 Table 4-5. Terms of Disagreement Between Physicians and DHP Women Item # Item DHP Rating Level Physician Rating Level 7 Sexy 4 2 8 Freeing 4 2 16 Menopausal Problems 1 3 19 Kind Of Lonely 1 3 21 Use Protection 4 2 25 Been Around The Block 1 3 28 Hotter 4 2 29 Less Active 1 3 33 Menopausal 2 4 35 Comfortable In Their Own Skins 4 2 Of interest are the different themes that appear in how each group conceptualized the issue by comparing the terms of agreemen t with the terms of disagreement. For example, both rated terms referring to bodily changes and changes in sexual function were rated as relevant ( medical effects on sexual functioning and body changing rapidly

PAGE 126

114 rated 3 in Table 4-4). However, analysis of the results in Table 4-5 on terms of disagreement, show that physicians are more likely to consider t hose bodily changes as problematic than are the women, even t hough both noted changes as relevant. A particularly noteworthy finding is that the women consistently indicated that terms referring to less sexual activity and menopausal problems were not relevant to describing mature sexuality. To summarize the findings, women appear to describe sexuality in very positive terms, as well as safe and careful with regard to the risks associated with dating. On the other hand, physicians tend to focus on the pr oblematic aspects of sex and aging, with much more emphasis on lack of sexual activ ity due to physical or medical problems. This is likely due to the fact that physicians in the course of their work tend to encounter midlife and older women who are having difficulti es in that area. The physicians do note some positive aspects as well, but they are more likely to disagree with the women on these elements of sexuality. These data provide useful clues as to wh at physicians need to know about midlife and older women. Women believe that matu re women are sexy and beautiful and that sexuality is an important aspect of their identity. They are also concerned with personal safety and their well-being as they ente r new relationships af ter divorce or widowhood. The women in this sample may be describing mo re of their ideal for sexuality in later life than their personal reality, since the intervie w data reveal a few are not interested in finding a new partner, though some conceded that they would be inte rested if the right person came along. This topic will be expl ored in greater detail in chapter six.

PAGE 127

115 Although the ratings data made it clear that the consensus on one culture is apparent for the DHP women, but not for the physicians or the combined samples, the consensus is not very robust. Given that th e consensus appears to exist with the women, I used Property Fitting analysis to assess whethe r or not the dimension tested in the ratings was the dimension motivating the participants perceptions. Property Fitting of Ratings and Pile Sorts Similar to cluster analysis and dimension testing such as ratin gs tests or paired comparisons (Borgatti 1996b), PROFIT (PROpe rty-FITting) is a way of looking at an MDS for groupings of meaning. PROFIT is a multiple regression technique that uses the coordinates from the aggregate proximity matrix of the pile sort data as the independent variables and the attribute, in this case relevan ce of terms used in the ratings test, as the dependent variable (Borgatti 1996b) PROFIT merges the ratings data with the pile sort data and the result indicates if cultural consensus around the dimension used in the ratings test can be supported or refuted. The results of the regression are represented by means of a directional vector in the configuration. The PROFIT output in ANTHROPAC give s coordinates for the regression line with the centroid point (0,0) and another po int (MEAN) from which to form the PROFIT line (regression line). The regression line is l iterally drawn on top of the MDS. In Figure 4-7, the regression line crosses through the Cent roid point, indicated w ith a plus (+), and through the MEAN out to the other side. The arrowhead of the line shows the direction on which the attribute (in this case, relev ance for describing sexuality for women 45+) increases. The relevance increases going s outh and west on the map. Closeness to the line does not indicate anything about the signi ficance of any given item. Rather what is

PAGE 128

116 + lack of desire kind of lonely non-existant comfortable in their own skin renewed interest better satisfying caring special beautiful enjoyable happy more pleasurable more mature comfortable confident ignorance been around the block dangerous be left alone medical effects on sexual funtioning don't bother too busy talks freely uninhibited use protection good once a week more careful sensitive fun sexy on-going exploratative desirable uncomfortable menopausal problems more responsible monogamous body changing rapidly disease risk intimacy menopausal less active got to have somebody MEAN freeing companionship w/ financial help sex better than when in20s hotter Figure 4-5. PROFIT Analysis on DH P Women Pile sorts and Ratings

PAGE 129

117 meaningful is where the perpendicular line connecting the terms and phrases to the PROFIT line, fall in relation to the PROFIT line. For example, got to have somebody is less relevant to respondents than disease risk or too busy. The terms happy, sexy and more responsible are more important than body changing rapidly and so on. The regression line expresses the best pr ediction of the dependent variable ( ratings data ), given the independent variables ( pile sort data ). However, these things are usually not perfectly predictabl e, and there is often substantia l variation of the observed points around the regression line. The r-square of this analysis examines if the arrays in the MDS are representing what was tested for in the rating procedure, th e higher the r-square, the more accurate or real the dimension. Th e r-square indicates how well the model fits the data, such that an r-square closer to 1.0 indicates that the model accounts for almost all of the variability considered with the data. A low r-square value may be caused by there being too many dimensions within th e cultural domain, or a simple lack of dimension, that it, the dimension that appeared to be there, or the patterns in the MDS are not really representing what they seemed to be representing. In this case the r-square is 0.699, signifying that this m odel explains 70% of the original variance and only 30% residual vari ance is unexplained. Naturally, it would be better if the model explained 90% or more of the original variance. However, the significance level ( p -value=.001) calculated for the corr elation is an indicator about the reliability of the correlation, and the si gnificance changes depending on the number of items computed (Borgatti 1996b). Borgatti (199 6b) suggest that for domains with less than 20 items, an r-square of at least .80 is needed to s upport the conclusions that the attribute was an important fact or in determining the similarity among the items. With a

PAGE 130

118 larger number of items, in this case 48 items, the r-square may be slightly lower and still support the hypothesis (StatSoft 2004). Conclusion The Cultural Consensus data presented in this chapter revealed a few unexpected findings. I originally hypothesized that th e DHP women and the physicians would differ substantially in how they conceptualize the domain of mature sexuality for women. The free list and pile sort data s uggested otherwise. Although th ere were differences between the two samples, there was more agreement than expected. It was in teresting to find that the two samples shared more terms and phras es when the samples were combined than when they were analyzed separately. Th e physician free lists had a particularly low frequency of repeated terms, but they had more in common with the women than with each other. I expected that to be the ot her way around, with greater consensus of terms for what was expected to be a homogenous group of physicians, and less agreement from women with differing backgrounds. The MDS and cluster analysis of the pile sorts showed that the two samples organized the items in much the same way, sa ve for with a few notable exceptions. The most significant exception related to risks a ssociated with dating a new partner. The DHP women considered this an important relationship el ement, whereas physician did not. This finding is congruent with the result s on notions about HIV risk that will be presented in Chapter 6. The physicians developed a coherent grouping on menopause in the MDS and cluster analysis, suggestive of their training and experience in clinical practice, but the women did not link all of the menopause re lated terms together perhaps because of intracultural differences with menopause related to age.

PAGE 131

119 Despite the apparent degree of shared cu ltural knowledge between the samples, the ratings data clearly indicated th at the participants in this exercise were drawing on more than one cultural understanding of the topic. The ratings data however, showed more than just their differences. Closer inspecti on of their responses revealed how they were constructing the domain, with women focusi ng on more positive as pects of midlife sexuality and physicians focusing on more negative or medical aspects. Finally, the property fitting analysis supported the conclusion that the women were considering relevance of terms to describe th emselves, and others in their cohort, when they organized the terms and phrases. A lthough the DHP women in this sample come from a range of backgrounds and range in age from 45 to 68, they share certain concepts about aging and sexuality. The next chap ters expand on how they view aging and sexuality in terms of communi cation with physicians about sexual issues (Chapter 5), including age and ethnic divi sions, and their views on HIV and STD risks after divorce and widowhood, and talking about those risks in clinical contex ts (Chapter 6).

PAGE 132

120 CHAPTER 5 AGE, GENDER AND ETHNICITY IN CLINICAL TALK The major focus of this project is ho w mature women and physicians describe sexuality and aging, and how their percepti ons about it may affect physician-patient communication. Questions this research sought to answer included: Are physicians ideas about women's midlife sexuality significan tly different for womens concepts about midlife sexuality? Do physicians hold ageand gender-based assumptions about discussing sexuality that impact their pract ices? Is women's comfort with discussing sexuality issues in clinical contexts impact ed by the age, gender and/or ethnicity of physicians? Hypotheses Based on the literature re views I hypothesized that physicians and DHP women would conceptualize mature female sexuality differently and have different views on how it is addressed in the clinical encounter. P hysicians were expected to view older women as sexually inactive if unpartnered. They would assume them to be in mutually monogamous relationships if they were partnered, and not perceive them to be at risk for HIV/STDs through heterosexual contact. Addi tionally, age and ethni city variables of both patients and physicians were also likely impact the cl inical encounter. Sexually active DHP women were not expected to perceive themselves to be at risk for HIV/STDs because of their age, because their providers had not discussed it, and because they have not been targeted by pub lic health campaigns. Women would not want to raise the topic of sexuality with their physicians and woul d prefer that their physicians

PAGE 133

121 raise sexuality issues. I also expected age, ethnicity, and level of education to be important variables in women's perception of their HIV/STD risk. I anticipated that younger women and minority women would ha ve had more exposure to HIV/AIDS issues. Women were expected to express a preference for female physicians and/or physicians of the same ethnicity. Finally, cultural consensus modeling was expected to show that the domain of sexua lity in later life would differ between DHP women and primary care physicians, but that there woul d be a high degree of internal consensus within each sample. Some of these issues have already been touched on in the previous chapter. The cultural domains revealed through CCM procedur es were different between the samples, though not to the degree expected. The samples shared more terms when combined together than they did separately. However, the way that each sample conceptualized the domain differed in significant ways, and the re sults were instructive as to what physicians need to know to be more competent when ta lking to mature women patients about sexual health issues. This chapter expands on th e findings of chapter four, and addresses many of the hypotheses above that deal with phys ician patient communi cation including how it is impacted by the age, gender and ethnicity of the social actors. Chapter six addresses the hypotheses dealing with HIV/AIDS risk pe rceptions and related issues for clinical communication. Age and Gender Constructs: Reaction s to the Hysterectomy Vignette One of the ways I explored the attitudes of DHP women and physicians was to read them vignettes about a health care encounter and then solicit their opinions about them. The story of a woman who needed a hysterectom y for fibroids was the first health care encounter vignette. The story relates a situ ation of a 64-year-old woman who undergoes

PAGE 134

122 a hysterectomy and suffers painful intercours e after the surgery (s ee Appendices C and D). She is very troubled by the impact on he r sex life and when she sees the doctor about it he tells her she just needs to have sex more often and it will get bette r. This response is very disheartening to her and she fears that sh e may not have a pleasur able sex life again. Most of the DHP women felt that this was a typical story (n=32, 72.7%) in the sense that it depicted something lacking on th e part of the physicia n (Figure 5-1). Just under 10% thought it was unusual. When I as ked what could be done to improve the encounter most mentioned that the doctor shou ld have given her more information before the surgery about what to expect afterwar ds and that he should have shown more compassion. 73% 14% 9% 5% Typical Sort of Typical Unusual Question not asked or not answered Figure 5-1. DHP Womens Views of the Hysterectomy Story A 49-year-old widow felt that the doctor should have been more specific about what to expect. She also pointed out that patients dont necessaril y know what questions to ask to get at the answers they need, sa ying, I mean it sounded like she asked the basic question, you know, will I be normal? And you don't always know these specific

PAGE 135

123 questions because the women haven't been thro ugh this before. Another said, I think he could have asked her more questions to unders tand what is important to her. . Some thought that the physician was hampered by time constraints. A woma n in her early 50s said, The doctor needs to be more compassionate and I know it's a business and they have to watch their time and everything else but sometimes a little suggestion that a patient can do, [for a patient] that is active seeking healing and getting better--it's so important. Another woman in her mid-50s supported this sentiment about the hurried schedule of physicians, and how it impact s the physician-pa tient interaction He doesn't seem to be responsive to her, or concerned, you know. He seems like, doctors nowadays, they're just cut and dr ied, they'll get you in and out, and maybe she needs to find another doctorbecause ha ving more sex when it is hurting like that she's going to be scared to death. . A woman wants comfort; she wants somebody who cares about her situation a nd can tell her up front what can happen afterwards, what she needs to do and I don't think he did that. While most focused on what the physician needed to do differently, about a third focused on what the patient should have done differently. Several thought she should be more assertive and/or seek out another phys ician, and get a second opinion. One 61 year old said, Well number one, she should have g one to two other doctors before having that knife cut on her, because anyway, it' s an incision in your body. That lady should have thought a little better. Another said, She needs to be more open with him, you knowI think she should have communicated w ith him more. And if he didn't respond to her then she needs to go get a second opinion. A few reasoned that the patients lack of initiative was due to her age, and norms for physician-patient interacti on during the era that she was raised. A woman in her mid40s said, I think thatshe didn't push him hard e nough. I think it might be an old world issue that I came across with my mother-inlaw, that the doctor's always right and

PAGE 136

124 [youre] not supposed to question him. It has a lot to do with your age and what you believe, or how you were brought up. . I think she should have questioned a great deal more and the 'you'll be fine don't worry about' is rather condescending, and that would be my first tip off that you need more information and she should have sought it. Others not only suggested that she find a new doctor, but that she should find a woman physician. A woman in her early 50s said, I think she should be able to talk with her doctor comfortably about anything concerning her health. . So I think she should change doctors and get a woman doctor. The vignettes often served to raise topics that I planned to address later in the interview. For example, the hysterectomy vignette led some of the women to comment on age and gender issues in physician-pati ent communication, such as in the quotes above. Gender came up more often than age, and both the physician and patients gender were salient issues. One woman who was very forthcoming with her opinions said, I think it sounds typical and to me it sounds like it's a man doctor that's my opinion. For him to say, well that's not unusual and have sex more of ten, he should have told her there may be ways that might ease your pain or you know, if more sex would be good then you have to get beyond the pain or otherwise what's the point. Others focused on the gender of the patient and how physicians make assumpti ons about their patients based on age and gender stereotypes. One said, [D]octors ha ve a tendency to blow off women and their concerns sometimes. It sounds like he didn't re ally inform her fully of what was going to happen with her sex life. Another said, I th ink a lot of times the doctors will, in their profession, they'll basically classify females a like to a certain extent. Another felt that both the age and gender of the patient shaped the physician approach to patient care, saying, I think he was just thinking at her age, why not just take it all out.

PAGE 137

125 The focus on gender, as well as considering the person as a whole, was brought up a few times. One said, he needs to go to so me more classes and l earn more about, that women are real people, too, and that the issues of sex and, you know, emotionally, physically whatever, and that he needs to pa y more attention to the whole person. A woman in her mid-50s who has suffered numerous gynecological problems in recent years, and interacted with several different types of physicians as a result said, I think it's very poor on the part of the physic ian. I think he's very poorly educated on the complete picture of a patient as a tota l person. My experience is that this is a common phenomenon with docto rs, particularly in th e age of specialization, because they are very att uned to their narrow specialty. And oftentimes their training does not include this wider vi ew of the entire person and how the consequences of whatever the illness is w ill affect that person in all of the other areas of their life. For some this story described a reality analogous to their own. Of the 44 women interviewed 17 (38.6%) volunteered that they had had a hysterectomy and referenced their own experience to respond to the story. Several others mentioned friends or family who had the surgery when replying to my quest ions. One said, Well actually I have had that surgery I had fibroids, so I can definitely relate to th e story. From my perspective, visiting OB/GYNs, I feel more comfortable dealing with the nurse practitioners and talking to them. A widow in her late 60s replied Laughable, when I asked her what she thought of the story. She continued, I e xperienced something similar to this at the age of 34. First of all there is a difference af ter surgery. Also the doc tors are not too much interested to hear about what you're situati on is after this has taken place. A concern that some physicians are not interested in their patients views, and do not show enough compassion, was raised by some of the physicia ns in their reactions to the hysterectomy story.

PAGE 138

126 Physician Views of the Hysterectomy Vignette Physicians shared some of the same c oncerns voiced by the women. Physicians were quick to mention that the doctor in the story did not appear to ask open-ended questions that may have clued him into what was important to the patient. A few said that he lacked or did not show empathy for his patient. Like the DHP women, several physicians pointed out gender and communicatio n issues, but none of them raised the topic of a second opinion like so many of the women did. However, on the whole, they viewed the scenario in much the same way as the women. Only eight physicians heard the hyster ectomy story, seven attending physicians and one resident. The rest heard the HIV story or participated by only doing the survey and one of the CCM procedures. All of the physicia ns who heard this story said it was either typical or sort of typical. Several of the physicians, not surprisingly, focused on the procedural aspects of the story, but most were also attuned to the psychosocial aspects as well. One attending physician said, Just the fact that we tend to be focu sed more on the technical aspects of the procedure and not as focused on the psyc hological or physical long-term effects I've had some women tell me after a hyste rectomy it's very liberating because they don't have to worry about getting pregnant an ymore, and if they were having lots of pain before, the pain is gone. . So it can live up to its billing in other words, that everything does go well and everybody's happy afterwards. [But] I think we have to be aware that doesn't happen in every case. A couple of the female attending physicians me ntioned gender issues in physician-patient communication. One thought that the case was on ly sort of typical, but that it did not sound extreme either, and that it may have mo re to do with the physicians specialty and gender. I would think it might depend on if you're ta lking to a male surgeon specialist, as opposed to a female. I think it probably does make a little bit of difference because they're coming from such a different point of view, as far as their own experience.

PAGE 139

127 We do have an advantage being a female taking care of females because you know what a vagina is and how it feels to be a woman and men just don't Do I think this happens? Yes. Do I think it's real common, I would lik e to think not real common, but does it happen, yes. Several physicians mentioned how they woul d have address the situation, usually mentioning the use of different types of questioning to ascertain that the patient understands what they have told them, and to learn what is im portant to the patient. The physician quoted above continued describing he r own approach to talking to patients, saying It sounds like he covered himself as far as legally what she ca n expect as possible complications. . But then there's this undercurrent of her concern about her human sexuality and potential complications so it sounds like he should have explored that some more--I personally like to try to give people examples of situations that have deve loped and how people handle things. . The follow-up, seemed to me to be rather cold. . It sounds a little terse to say just have intercourse more frequently. Another female attending physic ian was very bothered by the story, saying, I think it's probably pretty common, up until the point wh ere somebody says, oh, you just need to have more sex. She continued, I mean, that just really bugs me that he would say that I think she needs to be reassured and hopefully he, you know, he remembers that she had concerns about her sex life before surg ery and can really take a little more time and explain to her that this is not uncommon and that having more sex is not the answer at this point. Another physician reaffirmed this, and concurred with many of the DHP women who felt the physician in the story lacked co mpassion for his patient, saying it was very inappropriate, inconsiderate, not responding really to her needs, and probably very paternalistic. He felt it was typical from th e patient's perspective, and continued and typical also from the physician's perspective, in that I think physicians probably don't

PAGE 140

128 spend as much time addressing these issues, as you demonstrated in the scenario A couple of physicians mentioned time issues impacting the scenario, saying that time constraints limited his ability to ad dress his patients questions fully. Another noticed that the doctor in the story used some medical jargon in his explanations, saying, As a physic ian, I try to avoid terminol ogy like that because people don't know what you mean and they don't know what questions to ask. Only one physician said that he thought that, except for being more specific about the healing process, that both the patient and phy sician handled the situation well. The fact that most of the physicians that were asked about their opinions of this story were family practice physicians may be indicative of their great er competency with psychosocial issues, since their training incl udes a broader focus on the whole patient and the patient across the life span. However, since these physicians found the story to be typical as well as troubling suggests that they recognize that many primary care physicians are not considering the whol e patient in their practices. The topic of considering the patient as a whole person and not just the ailment that they are seeking treatment for was also addressed with questions about physician sensitivity to the sexual side-e ffects of therapy or drugs. Sensitivity to Treatment Effects on Sex-life. The hysterectomy story was used to highlig ht the issue of post-treatment effects on sexuality. Following the story I asked res pondents if they thought physicians were sensitive to how medical treatments or drug trea tments affected their patients sex life. There is a possibility that this question wa s somewhat leading following a story in which the physician appeared insensitive to his pa tients sexual concerns. However, this

PAGE 141

129 concern is ameliorated by the fact that th e story elicited comments about the physicians lack of compassion from both physician and DHP women before I asked this question. Figure 5-2 shows that more than half of the DHP women felt that physicians were not sensitive (n=15, 34.1%), or most of them were not sensitive (n=12, 27.3%), to how medical treatments or drug treatments affected a patients sex life, and two-thirds of the physicians felt that way (12.5% and 50% respectively). Ho wever, only eight physicians were asked this question, so 50% only repr esents the views of four respondents. Ph y sicians sensitive to how medical & drug treatments effect sex-life? (percents)6.8 0 27.3 12.5 25 37.5 34.1 50 6.8 00 20 40 60 DHP WomenPhysicians Yes No Sometimes Yes Mostly Not Don't know Figure 5-2. Physicians Sensitive to How Me dical and Drug Treatments Effect Sex-life. Time constraints were brought up frequently by both samples on this topic. One physician noted that bringing-up sexuality issues can be very time consuming and therefore physicians try to avoi d the topic. He explained, I think that's an issue; I think many tim es today we--and I'm probably guilty at times--don't spend enough time, don't feel we have enough time with our patients. We kind of circumvent some of these i ssues. Some of them, like sexuality, you're kind of opening a can of worms once you get into it it can be a long discussionit can really kind of set you back in terms of your schedule that afternoon or whatever. So I think many times. . too many times it's avoided.

PAGE 142

130 DHP women considered the issue of time constraints as well, one saying that oftentimes there are too many co mpeting issues, saying, No, I just think they're more or less concerned about your heal th and if you're not going to be sick anymore. They'd really have the time to think about Oh my patient's not having enjoyable sex. Other DHP women felt that women patients in particular were shortchanged on this issue. One in her mid-50s said, Not if it's a woman. If it's a man, they're going to do everythingbut if it's a woman, they think that it's just all supposed to work. Cause you don't have to get it up, you just have to spread your legs. Another felt is was more of a problem related to the patients age. I thi nk it's still a hard subj ect to discuss and if you're not going to they're not going to either, especially with older people. A physician had a similar response, saying, I think ther e's definitely an ageism. We find a young couple, newly married; of course we're going to bring it up every time. When we see an older woman with breast cancer or starting chemotherapy for something, or an older woman past menopause, we wait for her to bring it up. Another of the physicians also raised the s ubject of age, but felt that it had more to do the physician age and maturity, and their background. One of the attending physicians said, It has a lot to do with the maturity level of the indi vidual [physician], maybe the spiritual background or the in terpersonal dynamics. . I don't think many of them, whether they're residents or attendings, talk to the patients about the side effect and how that impacts patient's life. I think there's probably few and far between that actually talk about that or have any idea how much that can impact a couples life. Another attending physician expressed someth ing similar. I think amongst the older residents it's a little bit more relaxed. I think the kids who co me right from high school to college to medical school, I think it does make a difference whether you're married or not, whether you've got kids or not, what your own personal sexual experience is.

PAGE 143

131 A 65-year-old widow raised the issues of time, money, age and gender in how these subjects are addressed when she answered this question, saying, Not all the time, because they're not dea ling in that area, because they mostly thinking about helping you ot herwise, not thinking about sex. You know that's the last thing on their mind when they get bus y, and making money. And they just like give you something quick to fix it, then get onto the next one. They're not thinking about how this might affect me, they're thinking that a woman when she gets over 45 she don't need sex. She's on the change anyway, you see. Not all of the DHP women thought that physicia ns lacked sensitivity on this subject. One said that she is very happy with her physicia n, but said that she knows that if she wants information about side effects she has to as k about them specifically. Another confirmed this sentiment, saying, I feel like doctors more or less expect patients to ask specific questions I don't feel like doctors are thorough enough in explaining things to their patients. I think they more or less leave it up to the patient to be assertive and to ask questions. Clearly, from the responses to this ques tion and to the vignette questions, many of the respondents expect patients to be pro-activ e, even though they realize that some older patients might be reluctant to raise the topic with their physician. As far as physicians are concerned their failure to raise the topic or even to make patie nts aware that it is safe to talk about sexuality is related to whether or not they perceive th eir patients as sexually active. In addition, perceptions of whether a patient is sexuall y active or not is most often centered on gender and age, and on a patients current partner status. However, partner status and level (or type) of sexual activity can change from one day to the next, meaning what may have no relevance for a patient today, may have relevance for the patient tomorrow.

PAGE 144

132 Assuming Older Patients Are Sexually Active I asked both sets of participants if th ey thought physicians should assume that mature patients are sexually active. Sin ce not all patients are sexually active, some respondents found this question difficult to answ er. This was not a survey question, so only the physicians that particip ated in the interviews were asked this question (n=18). Additionally, I neglected to ask one of th e DHP women this question. Physicians overwhelmingly responded, yes to this que stion, whereas DHP wo men were split down the middle. The comparison of the two groups on this topic is statistically significant (Table 5-1). Table 5-1. Cross-Tabulation of DHP Wo men and Physicians and Assuming Older Patients Are Sexually Active Yes, Assume number (%) No, or Dont Assume, Ask number (%) Fishers Exact Test 2-sided exact significance DHP Women 22 (51.2) 21 (48.8) Physicians 15 (83.3) 3 (16.7) .023 The distinction centers on physicians wa nting to err on the side of assuming patients are sexually active so that they do not miss something important to a diagnosis, and half of the women wanting physicians to ask about patients sexu al activity so that they understand their patients as individuals, rather than assuming something that might not be true. One of the DHP women explained her difficulty in answering this question. You know, that is hard to say. Because I know, as I get older, I'm not as sexually active as I was earlier. But you know, I know other people older than me, and they've got a new boyfriend and all. . I don't know how to answer that. Put it like this, I don't think the doctor should assume th at. They should just go along with the doctor patient relationship, find out what's what and go on from there. But they should never assume because you never know when it comes to sex. Some of the DHP women who responded yes to this question pointed out that appearances may be deceiving, and that current situations are not n ecessarily indicative

PAGE 145

133 of what their futures hold. A widow in her late 40s who has a new partner explained this well using her own situation. I would think that any doctor should think that their patient is sexually active unless they're told otherwise. I would think that--all right, take my situation. I'm widowed and a doctor might just auto maticallypeople might automatically assume that I'm not sexually active because of it. But that would be the wrong way to look at it. I think a doctor should assume you're active unless they're told otherwise, that way they're not missing something. One of the attending physi cians supported this poin t of view, and acknowledged that he had made the wrong assumption so metimes and that he catches himself stereotyping older patients as sexually inact ive. I think you should assume that. I don't think they all are, but you shoul d assume that. . I tend to as sume too often that they are not sexually active. I thin k this is kind of a stereotype th at goes on in my mind. Another attending physician concu rred, drawing on her own c linical experiences and misjudgments on this issue, saying, I think we should assume it because . we've had older patients with HIV that people just a ssumed weren't sexually active. And you know, I had a patient with syphilis and I just assu med that she wasn't sexually active. These types of scenarios are what the physicians that supported the pract ice of assuming older patients are sexually active are hoping to avoid. She continued to describe her point of view on the matter, I think it's a dangerous th ing to assume that there's a certain age at which people--sex becomes no longer importa nt or people are no l onger able to do it, because I think it's a real impor tant part of who we are as people. Finally, one of the DHP women raised the issue of on-going change s in an individuals sexual activity. She is separated and does not have a current part ner, but expects to begin dating soon. She said, Yes. We're all human. . Yes, they should assume that they're sexually active. They may not be, but if they're not ri ght now they probably will be later.

PAGE 146

134 Many of the DHP women were bothered w ith the notion of physicians making assumptions, and most felt that physicians s hould ask their patients about it to be sure. One divorced woman in her late 40s was quite adamant, saying, Never assume anything. Period. Nobody should ever assume anything. Another said, No, because like I said, in assuming, you don't really know. . It's bette r to be sure than assume, because with assumptions sometimes you miss out on something that's important. So you just need to know. Several were very clea r that they thought that physicia ns should raise the issue to be sure. One said, Some are and some aren't, so I don't think a doctor should make that assumption. But I think they could bring [ it] up, then the patient could say, well you know, doctor, I don't have any sexual partners . But raising the issues, I think it's important to raise the issues, certainly. Only three of the physicians thought that it was inappropriate to assume that older patients are sexually active, and one resident said that it wa s a matter of asking patients about it. Another resident was more reticent and worried about irritating patients, saying, No. Because you'll really tick off that one person who's not to no end. However an attending physician who thought it was good to assume patients are se xually active, but that it was important to ask to be sure, wa s not concerned about upsetting his patients, saying, Yes, I think that's a good place to st art and you can always laugh. I haven't had anybody ever be offended by me asking. In some cases assumptions about sexual inactivity based on a patients advanced age proved to be erroneous. This type of stereotyping can be dangerous to a patients health, as can other kinds of stereotyping. The physicians and DHP womens comments about the hysterectomy story, and their assump tions about sexual activ ity or inactivity of

PAGE 147

135 older patients, indicate that age and ge nder stereotypes are problematic in many encounters. Physician Gender The literature suggests that women and me n physicians talk to their patients differently. Although not always the case, women physicians are generally thought to be better listeners, to interrupt their patients le ss often, and to be more empathetic (Adelman et al. 1991; Cline and McKenzie 1998; Hall and Roter 2002; Haug 1996; Lorber 1997; Roter and Hall 1998; Street 2002; West 1984). To get a better unde rstanding of how gender impacts the clinical en counter, I asked the DHP wo men if they thought that women physicians and men physicians communi cated to their women patients in the same way. Their comments about the hyster ectomy vignette showed that they have strong opinions about differences between women and men physicians. Two-thirds (n=29, 66%) thought that women physician used different communica tion styles (Figure 5-3). However, their comments on the topic re veal that the issue is much more nuanced than these figures indicate. 14% 66% 16% 5% same different sometimes same/different don't know Figure 5-3. Female and Male Doctors Talk to Their Women Patients in the Same Way.

PAGE 148

136 Six of the respondents (14%) thought that there was no difference between women and men physicians. A woman in her late 40s thought that she had ample experience with physicians to answer the question becau se she has four children and has undergone a hysterectomy for fibroids. She felt that wome n and men physician relate to their women patients in much the same way. I used to think men didn't understa ndI used to think that men were rougher or harder to talk to, but I think that was me being tense. Well. maybe it's a new generation of men too, they'v e grown as I've grown, and they're more open A few women echoed this sentiment, a nd several said that it was more of an individual personality issue th an a gender issue. Just over 15% felt that sometimes it can be helpful to speak with a woman physicia n, and sometimes it makes no difference. A 45-year-old woman said, I've been to both and I think that that is a definite individual thing. You can get an individual male doc tor whos very atten tive and caring, you can get a woman doctor who's very, very professional and all that. . I don't think it matters. A couple pointed out that since they undergo the same training they often behave in much the same way. One woman in her mid-50s related the story of how she felt after an encounter with a male and female physician. Talking about sexuality issues . I mean it's easier for me to open up and start talking when it's a woman than a man, but the woman is still busy and she--you know, and the guy is busy and the woman' s busy, and it seems like they weren't really hearing what I was saying and I went away feeling like, well did I appear to be a fool, you know. . So I come away feeling misunderstood--and maybe she did understand a little bit more, but she still felt a little cold and distant, to the point where I came away I didn't feel totally comfortable about the whole thing. Another in her early 50s seemed to have a similar feeling as the one above saying, I don't know. I've mainly had male doctors. . I would hope the women doctors would be more sensitive to female issues and listening, but I remember years ago going to a female gynecologist and she wasn't any different from the male ones. I think it's just an individual thing, finding the person who will listen to you and make you feel comfortable.

PAGE 149

137 This theme of feeling comfortable wa s brought up often. Many women mentioned that they wanted to be listen ed to, and to feel like their physician cared about them as an individual. Most often this se nse of being listened to and feeling comfortable was linked with interactions with female physicians, but not always. A wi dow that came from a military background said that she felt more comfortable with female doctors, but that in the military she often did not have a choi ce, and was grateful to some of the compassionate male physician she had encountered. There are some male doctors that ar e just as warm and caring--everybody's different, everybody's not the same. I feel more comfortable with a female doctor because I think they understand, although bei ng in the military I didn't have lots of choices like that. . The la st doctor that I did see on the base he was very good and he make me feel comfortable with him. . He looked me straight in the eye. He was just very good and as a military docto r I appreciated that more than they'll know. Most felt that a woman physician could understand or relate to them better by virtue of being a woman. Many repeated so mething similar to the following statement by a 47-year-old, I think a female doctor could probably relate to female health problems more than a man could. A man can go to medi cal school but he's not really experienced what it's like to be a woman. I don't think th at can be explained. Some were concerned with male physicians thinking that all women are alike, as this woman in her late 40s pointed out, You don't want him to think that you're in a category, 'cause everybody doesn't feel the same. Just, I'm different fr om Jane who just came in and Mary that's going to come in after me. . I don't car e what you tell me, a guy doesn't know how you feel. There was a clear preference for female phys icians, particularly for GYN visits. A widow in her late 60s said shed prefer not to go to any physician, but continued, it would depend, if it was a female area, I would go to a female. If it was my heart or my

PAGE 150

138 intestinal whatever, it wouldn't bother me to go to a man. But if it's GYN, then I would prefer a female, only because they would be mo re in tune to what you're trying to say. A few women factored in th e age of the physician. One woman in her mid-40s said, I think that a female doctor can usually understand issues better because they're women themselves. In my personal case, it would be if the doctor's a female instead of male, and be of like age. I don't really think . I'd feel as comfortable speaking to a woman who was just out of medical school in her early 20s I prefer a woman who's at least as ol d as I am or slightly olde r, who's been there and done that. Another said, I haven't really met a doctor that's very sensitive. But then, most of your gynecologists and stuff that you go to are men . I don't think women are comfortable, really, talking to men about any sexual problems or whatever. But then again it could be a question of age in that situation, whether it's male or female. A female would be more likely to understand but if she has not been through menopause . then they're not going to understand as well either. Some felt that women physicians did not seem as rushed as men physicians. One said, I think female doctors are just a little b it more understanding, a nd a little bit more comfortable. Male doctors I thinkand not all of them, some of them--they just get in, examine, diagnose, treat, and move on. A few took the issue a step further, saying that they prefer seeing a female nurse practitioner, saying that they dont seem as hurried as physicians, and since they are usually female, it is more comfortable to interact with them. One explained how she felt about the clinical encounter w ith physicians versus nurse practitioners. Well you're there in the examining room . . I think you're basically at a disadvantage in that you're usually si tting there with no clothes on, basically undressed, waiting to talk to a doctor, so th at's intimidating. And the doctor rushes in and says something to you, then examin es you, so that can be also intimidating and disconcerting. I personally feel more co mfortable with nurse practitioners than I do with male physicians. . The nurse practitioners that I've dealt withI've just been able to communicate and to feel more comfortable, and feel more warmth, and even though . they're as busy as sometimes the doctors are, but I just feel they're more gentle, more caring.

PAGE 151

139 Only three DHP women menti oned nurse practitioners, though it is possible that some women were not aware that th ey were visiting a nurse wh en they were referring to women physicians. I usually asked them if they were sure that they had seen a physician and not a nurse practitioner, and most felt certain that they had seen physicians. These comments about differences be tween women and men physicians are consistent with previous research (Nusbaum et al. 2002) on gender issues in clinical talk about sexuality. This project confirms th at women can feel comfortable discussing sexuality topics with both wo men and men physicians, but that they were more likely to discuss them with women and with physicians that create a comfortable atmosphere, one thats conducive to discussing sensitive issues. Preferences for women physicians. Twenty of the DHP women expressed a preference for women physicians (45.5%), but ju st over a quarter (n=12, 27.3%) said they had no preference. I neglected to ask this que stion to three of the respondents and five were unclear about their prefer ence. African American women were a little less likely to have visited a female physician, but their views about differences between men and women physicians were similar to those of th e Euro-American women. However, four of the African Americans said that they pref erred male physicians, whereas none of the Euro-American women expressed this prefer ence (Figure 5-4). These distinctions, though interesting, may be more a function of the small sample size than of any real ethnic differences. It may also be related to age and SES influences on the number and type of physicians they have seen. The Black women espousing this prefer ence were those who had little to no experience interacting with female physicians. Distinctions by age, educational level and

PAGE 152

140 other variables were unremarkable. One African American widow in her mid-50s described her experiences this way, I've ne ver had a female doctor--I've always had a male doctor, and I find that I get along better with a male physician, 'cause I've never had a female, so I don't know how to accept a female, but I think I could talk to her. However, most African American women e xpressed a preference for a woman physician and reasoned that women could relate better to other women than men. One put it this way, A female doctor tends to be more em pathetic because she's a female and I'm a female so we understand each other. I can relax more with a female doctor when I'm having to discuss personal things, like sexuality and stuff like that. 35% 52% 20% 0 25% 30% 10% 13% 10% 4% 0% 10% 20% 30% 40% 50% 60% Blk/African Amercian Wht/EuroAmercian female male none unclear not asked Figure 5-4. DHP Womens Preferences Re garding Physician Gender by Ethnicity. These perceptions about physicians gender co rrelate with previous research on the topic. Most studies on the influence of the physician's gender on interaction patterns indicate that the gende r of physicians and patients make a difference in medical practice (Adelman et al. 1991; Cline and McKe nzie 1998; Hall and Roter 2002; Haug 1996; Lorber 1997; Roter and Hall 1998; Street 2002; West 1984). Just as in the statements

PAGE 153

141 from most of the DHP women, other research shows that female physicians are thought to be more apt to provide support to their pa tients, to be more egalitarian, and to give more time to consultation than their male counterparts (Adelman et al. 1991; Haug 1996; Lorber 1997; Roter and Hall 1998). According to Lorber (1997) gender differe nces in communication styles are likely related to life-long socializa tion and interaction patterns. Street (2002) writes that physician and patient gender stereotype s and attitudes likely generate a priori assumptions about the capacities of e ach other for both gender concordance and discordance in clinical comm unication. Many of the women in this study appear to support Streets thesis in their assumptions about the na ture and interact ion patterns of women and men physicians. However, some reported that these notions based on gender stereotypes were not always supported in practice. Some of these women also volunteered that they believed that age conc ordance was also an important factor for being able to relate to the woman physician, yet the issue of ethnoracial concordance in physician interactions was not volunteere d. Nevertheless, th e question about the physician gender easily flowed into ques tions about the physicians ethnicity. Ethnicity Concordance in the Clinical Encounter Studies show that for patients with simila r clinical findings there are differences by ethnicity in treatment, amount of inform ation provided, and empathy displayed by physicians (Adler 1998; Cline and McKenzie 1998; Cooper-Patrick et al. 1999; Schneider et al. 2002; Woo et al. 2004). Race-concorda nt visits tend to be longer and are often rated as having more positive affect (Cooper et al. 2003). However, the comments from the women in my study do not tend to support the findings from previous studies. The studies often included preand post surveys as well and audiotape anal ysis of the actual

PAGE 154

142 encounters, whereas my study asked respondents 1) if they thought the ethnicity of either actor impacted the interaction; 2) if th ey thought Blacks and Wh ites were treated the same; and 3) if they thought they were trea ted as well as anyone else (Appendix C). The differences between my study design and the de signs of those in the literature review, as well as race concordant interviews may be an explanation for the different results. However, as I will describe in the following section, some of my research experiences with this project suggest that there may not have been a strong interviewer effect on this topic. The question that phys icians were asked about the race/ethnicity (Do you think the ethnic background (or race) of a patient aff ects how physicians communicate with them?) is not readily comparable to the DHP women data and is therefore not addressed here. This and other similar topics will be discussed under the limitations section in chapter seven. One nationally representative survey found that race/ethnicity is not the only factor affecting the quality of physician-patient in teractions. The survey over-sampled for Black, Hispanic and Asian households and found that race/ethni city discordance contributed minimally to the quality of interactions, and was better explained by differences in physicians cultural sensitivity and in patients hea lth literacy (Saha, Arbelaez, and Cooper 2003). Some of the comments in this study suggest that the physicians demeanor and patient expectations may be as important or more important than race/ethnicity concordance. My study sh ows that cultural and language differences are also important issues in race/ethnicity discordant encounters, and for some more important than racial concordance.

PAGE 155

143 Over a third (n=17) of the DHP women t hought that the race or ethnicity of the physician did not impact physician-patient communication (Figure 5-5), but 18% (n=8) thought that it was an important factor. The question, Does the race or ethnicity of the doctor effect how they communicate with th eir women patients? elicited a lot of comments about physicians from other countries and their English language skills. Many of the women had encountered foreign doctors and expressed a range of opinions about them. This is actually the reason that many of the White women and a few of the Black women mentioned as a reas on for ethnic differences. 18% 39% 25% 18% yes, difference no difference sometimes yes/no don't know Figure 5-5. Physician Race/Ethnicity Af fects Communication with Women Patients Interestingly, Black women were twice as likely as White women to think that race/ethnicity did not matter (55% and 26% re spectively). This finding is contrary to most previous findings on the topic of race concordance. Even though there were some marked ethnic differences between respondents on this question, the differences were not statistically significant due to the small sample size (Figure 5-6).

PAGE 156

144 One Black woman in her late 40s said, W ell the only thing I think in the doctor field [is] if it's a foreign person; sometimes you can't understand what they're saying as far as their words, but I mean you know if you speak plain English you shouldn't have no fear She continued, No, cause if you're goi ng to a doctor and you want to get well you shouldn't care who he is. But then she con ceded, Yeah, I imagine it is in some places, some issues, some people still got that old mentality in them and so, they probably do have a racial issue. This back and fort h between affirmative and negative responses about race/ethnicity discordance was common, and several made the distinction between themselves and others they know on the issue. 55% 10% 17% 30% 26% 15% 20% 26%0% 10% 20% 30% 40% 50% 60% yes, difference no difference sometimes yes/no don't know Blk/African-American Wht/Euro-American Figure 5-6. Physician Ethnicity Effect s Communication by Respondent Ethnicity A White woman in her mid-40s who thought that it made a difference sometimes, explained Well older people--it doesn't bother me about race--but older people it's really an issue with. Because my granny she w ould say, I'm not going to see a Black doctor and I'm like you might have to, this is t oday. Yeah it would bother her real bad.

PAGE 157

145 About a quarter of all the women, mentioned the issue of language or culture in response to this question. Many made the distinction between doc tors with different nationalities and their English language ski lls did not mention differences between African-American and Euro-American doctors. A White woman in her late 40s said, I would say it's probably just because different cultures treat older people differently. Plus, I think, and if you can 't understand the doctor, their English or whatever, that would cause problems as well because I have been to a doctor that I could not understand a thing that he said. Another said, Well I prefer not to have a foreign doctor. I've had a few [and] sometimes it's kind of hard to understand their dialect. A White woman in her early 60s was very direct about it, saying, No, it has nothing to do with it. Except for, sorry if you are, anybody with the last na me of Patel, I will not go to them I don't know I don't like them. This reticence to seeing a fo reign doctor came up a few other times. A White woman in her mid-40s pointed to how cultural issues intertwine with gender issues in the interac tion, saying, I'm only guessing th ere are certain cultures that have a different place in mind for women. Th ey aren't necessary thinking of them as equals and if they hadn't worked through that then maybe they'd be more condescending and less open. As in the case of gender differences, a few women felt that personality was more important than race/ethnicity. A Black wo man in her early 50s, who thought that it depended on the individual physician, not th eir ethnic background, said, I've had them all and some of the make you f eel comfortable and others, it's just a job. So I don't really think it matters. Another said, Not re ally, ifyou really enjoy being a doctor and you know you can help someone it shouldn't matter. A Black woman in her late 50s echoed

PAGE 158

146 this sentiment saying, No, I'm going to sa y no, and their ethnic background shouldn't have anything to do with it. I've had one Bl ack doctor, I've had foreign doctors, I've had White doctors [and] so far they're okay. Even though race/ethnicity concordance in physician visits was not an issue to most of these women, about a third of the Af rican American women related unpleasant experiences to me that had to do with the physicians race or the physicians competence in interracial clinical encount ers. Whether or not an Afri can American interviewer would have heard more of these stories remains an open question. However, the fact that a third of the Black women felt comfortable e nough to tell me about their unpleasant experiences, suggest that my ethnicity did not always unduly influen ce these interviews. One case in point on this is when I interviewed two African American women in the same room together, one right after the other. These women were friends and had shared a ride to the interview. Eac h, if she was interested, could li sten to the responses of the other. On the whole they had very different re sponses to the interview questions, and on the racial topics as well. The woman in her early 50s related this story to me about a White doctor that she felt unfairly judged her based on her color, gender and economic status. I have had experiences in that area where I was eligible for a service, but I really feel like with me being Black and a female--and [I] ended up filing a claim, because of the way the doctor talked to me and he had given me injections and, [saying] you can't feel [it]. So he was saying, well there's no way you can still feel it. It was like he was saying I just wanted something because the service was free I just got up and left because of th e way he was talking to me and I let him know that I was going to take action becau se of the way he had talked to me. Because if you fall eligible for something-then I still have my rights, and I'm not going to allow anyone to talk to me or be little me [just] because I'm in this situation. You're not just going to disres pect me and take my integrity from me.

PAGE 159

147 Her friend, on the other hand, had never had su ch an experience and felt that race and ethnicity were not important factor s in clinical interactions. The story above was not the only one th at demonstrated socioeconomic and race/ethnicity stereotyping. Some told st ories about how physicians did not understand cultural differences in dealing with illness a nd death. Others told of personal account of how they felt they had been treated poorl y, as in the case of the woman above, and reacting proactively to maintain their dignity. A woman in he r late 40s talked about how the interaction pattern s of the physicians exposes their biases. Yeah...you can just feel it from the tone of their voice, if they are being short with you, don't want to touch you, you know then ri ght off the bat that they are not comfortable treating me. . Some pe ople don't speak up... but I just stop the appointment right there. . If I am sp ending my money for a service then I should get what ever, I should feel comfortable... Another woman in her late 40s gave me a de tailed account of a racist White doctor and how the experience made her feel. Well let's just say I've had--the story I was telling you about is. He was a White doctor I don't know if they were using me for a guin ea pig or whatever, but he treated me as though I was somebody just found off the street and my whole body was ridden with a disease. This was what really got me when...he says 'you people' and I've always had a thing, because when I was growing up that we're 'you people' and it always has a stigma with me. What do you mean you people --they're categorizing people. He says you peopl e go out and do things with your body. I'm like gosh, I'm in pain and he was adding pain onto top of pain. He was demoralizing me. . It might not be like th at now, because like I said, that was in the . late 70s, early 80s. It's probab ly not like that now, but when you have one bad incident you kind of carry it with you. You don't forget, you forgive, but now I'm very particular about a doctor that would categorize [me] as you people. Her candor with me, and that of others, in dicates that race concordant interviewing may not always be as crucial as some re searchers have suggest ed. Additionally, the wording of the interview questions, and the de meanor of the interviewer may ameliorate some of the interviewer effects. As a cultural anth ropologist, I endeavored to always take

PAGE 160

148 the approach that I am seeking to le arn from my informantsthey are the teachers and I am the student eager to learn about thei r lives and views of the world around them. This is a time-tested approach to cross-cultura l interactions, which ideally yields emic accounts of the topic under study. However, it is likely that my appr oach did not provide a sufficient comfort zone to all of the wo men in this study, and some women may have told an African American intervie wer very different stories. While a third of the African American women openly told me about unpleasant interracial patient-physician encounters, just over a quarter of White women told me they lacked experience in this area due to thei r race. The White women were much more likely than Black women to say they did not know if there were differences or not with regard to interracial clinical encounters (26% and 10% respec tively), explaining that they had not had any experience with that situation. This issue of feeling like they lacked the requisite experience to answer this questi on also came up when I asked if Blacks and Whites were treated the same. Blacks and Whites treated the same in clinical encounters. The DHP women were relatively evenly split on the topic of whether or not Blacks and Whites are treated the same in clinical settings. Just over a quarter think Blacks and Whites are treated differently (25.6%), 23.3% think they are tr eated the same, and the same amount said they are sometimes treated differently and so metimes the same, and 27.9% said that they did not know. I neglected to ask this ques tion of the first African American woman I interviewed. The relative balance in responses is lo st, however, when the ethnic groups are compared. Cross-tabulation of ethnicity by their opinions on th is issue showed that close

PAGE 161

149 to half of the White women responded don t know, usually based on lack of exposure to the issue. Chi-square analysis showed th at these differences were significant (Table 52). These are interesting findings in that they further refute the pr evious studies on race and physician-patient interactions. Although I have not found any studies that ask this specific question or one similar to it, the stud ies on racial issues in health care suggest that African Americans are of ten less satisfied with their interactions than Whites. Several of the White women said something similar to the following statement by a woman in her mid-40s. I'd like to say no, but I'm honestly not sure of that, and being Caucasian and not faced with a lot of prej udice I don't think I'm really knowledgeable enough to answer that accurately. This sense of not having su fficient experience was common, however others were much less reticent. Table 5-2. Cross-Tabulation of Ethnicity a nd Opinions on Blacks and Whites Being Treated the Same in Clinical Settings. Ethnicity Yes, same Number (%) No, different Number (%) Sometimes yes/no Number (%) Dont know Number (%) Likelihood Ratio Value (2-sided sig.) Blk/AfricanAmerican 8 (42.1) 5 (26.3) 5 (26.3) 1 (5.3) Wht/EuroAmerican 2 (8.7) 6 (26.1) 4 (17.4) 11 (47.8) 13.427 (.004) df =3 Some of the women, regardless of ethnici ty, raised the issue of SES as a more important factor than racial group, or as a factor linked to ethnic ity. One White woman in her early 50s said, If you have somebody who seems to be uneducated, they kind of tend to talk down to you, whereas if you seem ed ucated they get into a more conversation with you. One of the Black women who rela ted one of the storie s about maltreatment by a White doctor in the previous section, al so brought up socioeconomic issues, saying, Patients are intimidated by those who are well educated. It's just possible that they could come from a poverty area and not know how to address the issue. . Financial, insurance, there's a lot of things that contribute to how a physician would

PAGE 162

150 address or help a patient. There are those who are truly dedicated, through the oath that they take, and then ar e those that, for money. Several women, including the woman who selected other for ethnic identity, raised the issue of finances and insurance. So ever ything is money and what kind of insurance you have, they treat you differently depending on that. However, she continued that ethnicity had also been an im portant factor in her treatment. Like I said, I'm an American citizen but I wasn't born in America and I have accent. People show reaction toward this and they treated me differently. In some cases I've been target of prejudice and ignorance and discrimination and it was ve ry frustrating and painful for me. A White woman also mentioned insurance cove rage, saying, Yes, [they are treated the same] if they're have insurance. [If not,] you can forget it, they're not going to talk to you. The fact that 42% (n=8)of the Black wo men thought that Blacks and Whites were treated the same is well qualified by this stat ement about income level and color from a Black woman in her late 40s, Yes, I think they treat them the same, but I think your economic value is different. I mean you're treated different according to your income level. So I don't think it's a color thing, I think it's an income thing. Because I feel that people that are on the poverty level are treated different than th e people are who have a higher income. I don't think it's anything about color. Another woman the same age based her ideas on this issue on her personal experiences. I've actually never seen a case where they we re treated any differently because of their color. I really haven't seen that because we've all got th e same body. So if there are some that have treated me any different because I'm Black--I really haven't gone through that. Another Black woman in her late 50 s had a similar response, saying, I believe that from the experience I've had, yes [they ar e treated the same]. Yeah, I've talked to

PAGE 163

151 people who have been in the offices of my doc tor, but the women all say the same thing. White women, Chinese women; they love him. However not all of the Black women were so positive, explaining that they believed that their were still problematic issues in interracial clinical encounters. A 48 year old Black woman said, I think they take more time to listen to a White person Some of them don't listen to your problem cause I know I had a doctor that...tell me it's in my mind, before I even get it out of my mouth. Another Black woman described an experience she had where she was treated di fferently than a White woman patient. I have noticed different cases where, as far as how they greet you when you come up to the desk for your appointment. Maybe if you arrived a lit tle bit late and you were trying to explain to them your situati on and they may tell me, well you're just going to have to reschedule and make a nother appointment. Then someone else comes in late of a different race and they 'll say, okay, I'm going to see what I can do. And being a single parent 27 years I've noticed a whole lot. It's getting better but it's still there. Several of the DHP women recognized that prejudice and discrimination still exist in physician-patient relations, and some based this on personal e xperiences such as the story above. Many had the notion that socioeconomic factors were more important than racial factors, and others thought th at the issues were some am algam of racial influences, income level, insurance coverage, and edu cational background. On the other hand, some had never experienced the effects of racial stereotypes, leaving them feeling ill equipped to respond to the question. Still others felt it was not a problem, or usually not a problem. One African American woman didnt see this as so much of a di scrimination issue as simply one of cultural differences, saying Di fferent people have been taught different things in their culturesand there are a lot of things in your culture--that you use different terminology for differe nt things and they may not feel comfortable discussing certain things with the doctor. She appeared to feel that racial concordance was an ideal

PAGE 164

152 situation, particularly for discussing sensitiv e issues, noting that pa tients may not raise sensitive topics they are con cerned about with physicians wi th other ethnic backgrounds. This issue of comfort level in discussing sensitive issues is shaped by the clinical atmosphere, SES, gender, age and ethnicity i ssues, as well as by e xpectations about the roles of physicians and patients. Physician/Patient Responsibility to Raise Sexual Health Topics Previous research indicates that most pa tients want their physicians to create an atmosphere favorable to discussing sensitiv e issues, if not outri ght asking them about such issues (Loehr et al. 1997; Metz and Seifert 1990; Nu sbaum et al. 2002; Waterhouse 1993). Ideas regarding whether or not it was the physicians or the patients responsibility to raise sexual health topics yielded some unexpected findings. The DHP women were much more inclined to think that it was the patients re sponsibility to bring up sexual health topics than were physicia ns. A comparison between the two groups was significant (Table 5-3). It was surprising to see almost half of the DHP women place the onus of responsibility on the patient. Howeve r, their comments following their response to this question showed that they place th e responsibility of providing an atmosphere conducive to discuss sensitive topics on the physicians. Table 5-3. Cross-Tabulation of Physician/Patient Responsibility. ID Physicians Responsibility Number (%) Patients Responsibility Number (%) Both Responsible Number (%) Likelihood Ratio Value (2-sided sig.) DHP women 12 (27.9) 19 (44.2) 12 (27.9) Physicians 7 (25.9) 0 (0) 20 (74.1) 26.003 (.000) df =2 Many of the women that said they thought it was the responsibility of the patient to raise sexual health topics also said that physicians needed to open the door to make patients feel it was OK to talk about it. A woman in her late 40s, put it this way,

PAGE 165

153 I think it's the patient's [responsibility], but I wish the doctors would ask more inviting questions. Invite peopl e or show that they're open, even if it is a shy older woman, 'cause he's being paid and he's e ducated and maybe she's not asking or is ignorant, and he won't know if he doesn't ask; it's his job. Another pointed out that patients have to ask questions, because the physician is not going to bring it up, saying If you have a spec ific question, don't wait for them to ask you, you're just going to have to either wr ite it down and ask and open up the dialogue yourself. Although, at first glance, it appe ars that the DHP women are placing more responsibility on patients, closer inspection reveals that this is linked to expectations about physician behavior. Over a quarter (27.9%) of the DHP wo men thought that it was the physicians responsibility to raise the topic, and the same percentage thought it was th e responsibility of both parties. Whereas almost three-qua rters (74.1%)of physicians thought that both were responsible, and over a quarter (25.9%) thought that it was the physicians role to raise the issue. Those that said both usuall y felt that it was either more the physicians role to raise the topic, or th ey placed a little more of the onus on the patient. For example one of the DHP women said, I think it's a little of both. I mean because if the patient is--has any concerns she needs to ask the doctor, but then again it would depend on what they were there for. If it was anything that had to do with sexual stuff I would think the doctor should bring something up, or offer a booklet or at least have them available in his office, so that if you're curious you can pick something up and go home and read it at your leisure. Another woman said, I want to say both, but I think that the patient, where they may be intimidated or embarrassed, that I think the doc tor might need to take a little bit of that into their hands. One woman in her mid-50s who has unfortunately had a lot of recent experiences with physicians, explai ned her response this way.

PAGE 166

154 In a sense, both parties are [responsible ], but we can't do anything about every woman out there that going to come to th e doctor. The population we can actually affect is the physician population; so le t's look at a short list of what we can actually do with these physicians so that they can take the bulk of the responsibility. That may not be fair, but on ce the doctor's prepared to realize that this is an important part of a woman' s life and all that kind of stuff . This suggestion about physician training is of particular interest since most of the physicians in this study are part of a family practice residency program. The vast majority of both resident and attending physicians thought that both parties were responsible for bringing up sexual health concerns, but most place more of the responsibility on the physician. One attending said, It's both. But I think a physician has to be a leader in opening the door for that, and make the patient fe el comfortable with it--not just that but other personal issues. I thi nk it's the physician's responsibilit y to make the setting or the situation conducive to that type of discussion. Otherwise it won't take place. One of the residents was less obliging, saying, The physic ian can't be expected to just dig for everything. It's got to be understood from the beginning that it is a team effort and your physician's there to help you if you give him the opportunity. Another resident said, That's tough. I think the patient will bring it up if there's an issue. . But I think it's the physician's obligation to bring out that subject. I think the physician has the duty to bring this issue, wh enever they talk with the patient. . Sometimes it's difficult to bring that up, busy schedule in the clinic each day, and sometimes you don't think of it And that's the truth-you don't think of it. Several physicians, not surprisingly consider ed time an important factor in shaping which issues they discussed with patients. One attending physician explained that she thinks both parties are responsible and that time issues limit the number of topics that a physician can raise. I think we are very remiss in doing it. A nd again, part of it has to do with one of your other questions, you know, do we have time? . When we have a 15-minute session we have three or f our chronic medical problems that we must address,

PAGE 167

155 sometimes you know, that's just like a l uxury question that you don't have time to cover. . So I think time is a big issue there. She continued, explaining that she believes patients also have an important role to play in the exchange. I do think it is also the patient has responsibil ity, because if it is something that is worrying them or concerning them, and we s ee this all too often where they have a vaginal discharge but it's almost like they'r e waiting for me to by mental telepathy figure out that they have a problem. And so I do think that they have a responsibility particularly if they have que stions or concerns or they're having an issue to bring it to the physician's atten tion and not expect us to be a magician, genie, you know, person to try to figure it out without and that happens quite a bit. However another attending physician agreed th at both were responsible, but considered the physician more obligated to raise the issue, particularly with older patients, saying, I would say definitely the older the patient the more the onus is on the doctor to do that. Again it goes back to cultural and gene rational differences I think are there. So I think it is more important for me to br ing up with my older patients because . the older the patient is the less likely they would be to be comfortable bringing it up. . Overall I would say it's the physician's responsibility. Many of the DHP women would agree with th e statement above, in that in the end it is the role of the doctor to create a sett ing where these issues can be discussed. One said, I think it would be mo re comfortable if the doctor brought it up--at least they would open the door and you would feel more comfortable talking about it. Another said I think it is the doctors [responsibil ity]. He should be trained to do that, and knowing that the average person will not bring that up . he should initiate it. Although close analysis of the comments from both samples suggest that most believe that physicians are obl igated to create an envir onment conducive to discussing sensitive issues, it is clear that time limitations and notions that patien ts will bring it up if they need to, limit the actual discussion of se xual health topics. It is also related to

PAGE 168

156 institutional constraints, in that, clinical practice norms di ctate that physicians focus on only one or two ailments rather than consid ering their patients as individual wholes. Discussion and Conclusion In the process of discussing the vigne tte and responding to my questions about gender and race/ethnicity, several broader to pics were exposed. Several of the DHP women made observations a bout the patient-provider dyna mic, which support the findings in previous literature. One such point was the power dynamic inherent in the dyad, that patients are at a relative disadvantage in the clinical encounter. That is, they may be only partially clothed and feel vulne rable as a result, they are in unfamiliar territory where the language of clinical practice keeps them in the dark, and the provider dictates the rules of engagement. Health care communication scholar s often describe the physician-patient exchange as hi erarchical and intercultural, due to the hegemony of the biomedicine (Baer, Singer and Susser 1997); th e structural power dynamics of modern medicine that places the physician as the pinn acle of the interaction structure (Baer et al. 1997; Cant and Sharma 2000; Henley 1977; Lorber 1997), and due to sociodemographic and structural differences between the participants, creating distance and misunderstandings (Cline and McKenzie 1998). The DHP women and the physicians both point ed to structural aspects of health care interaction that impinge on the effectiveness of a clinical visit. Both raised the issue of time constraints. Physicians pointed out how time issues affected the number of topics they can talk about with patie nts, particularly older patients that are presenting with multiple illnesses, whereas many of the women focused on the hurried disposition physicians, and how this behavior leaves them feeling like the doctor is not interested in them as individuals. The increasing competitiv eness of the health care system is leading

PAGE 169

157 to institutionalization of imp ersonal and routinized care (Coulter and Fitzpatrick 2000), which only compounds the time and financia l pressures on physicians, and leaves patients feeling neglected. Doctors have greater power in the medical en counter by virtue of the fact that they control access to health resources as well as the dynamics of the health care interaction, and that control is institutionally regu lated and sanctioned (Cant and Sharma 2000; Henley 1977; Lorber 1997). Some of the problems associated with physician-patient interactions are related to contemporary ti me and financial constraints imposed by the current health care systems, HMOs and clin ic/hospital administrators (Lazarus 1988; Paget 1993), but also because of the standard way that physicians manipulate the flow of talk and ultimately the meaning of illn ess for their patient s via constrictive communication practices (Clark 1996; Good and Good 2000; Paget 1993). This hierarchical exchange is in part developed in medical schools and residency programs, where students and neophyte physicians learn to reconstruct the pati ents narrative in the context of biomedical pr actice (Good and Good 2000). Despite the inherent asymmetry of th e physician-patient relationship, the DHP women pointed out that patient s are not passive objects in the exchange. They place a heavy emphasis on patient res ponsibility in the clinical in teraction, but tempered this emphasis with what they perceived as the ro le of the physician in terms of providing a safe environment for discussing sensitive i ssues. These women, while recognizing that the system places constraints on how much physicia ns can do in a single visit, assert that doctors need to find meaningful ways to show that they are interested in their patients

PAGE 170

158 concerns. Physicians appear to be aware of what they need to do, but they note that they are sometimes remiss. Some critical medical anthropologists point out that many patients react with what means they have to the existing social structur e of the current health care system (Baer et al.1997). Several of the stories in the page s of this chapter uncover such proactive behaviors or attitudes, such as women seeking second and third opinions, reporting maltreatment due to economic or racial disc rimination, and seeking health information on their own. Physician and patient stereotypes base d on age, gender and ethnicity were extensive. This is consistent with previ ous research showing that both parties are influenced by gender, age and ethnoracial ster eotypes in the clinical encounter (Cline and McKenzie 1998; Cooper-Patrick et al. 1999; Green et al. 1994; Hall and Roter 2002; Haug 1996; Ikels 1998Jenks 1993; Street 2002 ). As hypothesized most DHP women considered the gender of the physician im portant and generally preferred women physicians, particularly to talk to about sexua l health topics. They tended to think that most female physicians were more knowledgeab le about female health issues, and that they were easier to talk to. Several felt st rongly that a patients gender impacted how physicians related to them. The responses from some of the physicians appear to support the hypothesis that many will not consider sexua l health issues relevant for older women patients. However, age and gender stereotype s are intertwined with time constraints and it is difficult to say which is most influential in clinical practice. Their ideas about ethnic st ereotypes were largely inc onsistent with previous research. Except for a small proportion of African American women, few had

PAGE 171

159 experienced racial stereotypi ng, though several thought that it still occurred in clinical practice. On the other hand, the impact of fi nancial issues came up a number of times. Many of these women were very much awar e that educational and economic status affected how physicians interact with patients. This is consistent with previous research showing that if physicians do not have a financial incentive for spending optimal time with a certain class of patient s they are unlikely to do so (Allman et al. 1999). Jackson and George (1998) argue that minorities a nd members of lower socioeconomic classes are frequently assigned lower so cial value by health care prov iders, and that they tend to receive the substandard care. Some of the experiences of th e women in this study support this contention. The physicians in this study, even though they treat a largely low-income population, did not appear cognizant of this issue in their practi ce. Perhaps this is due to the fact that most of their patients are poor and reflections about differences in treatment are moot since there is no frame of comparison. This chapter shows that women and physic ians share some common expectations about the clinical encounter, particularly in terms of the role patients and physicians should take in clinical talk a bout sexual health. Unfortunately structural and institutional factors impede the communication process, pe rhaps as much as age-, gender-, SES-, and ethnicity-based assumptions. Several of the women and physicians men tioned time issues in their comments about physician-patient comm unication. This topic, along with comments about physicians listening to their patients, interrupt ions in discourse, and concerns about the impact of financial issues in health care were all discusse d. There are clues to how the respondents felt about these topi cs in the discussion of physician-patient responsibility,

PAGE 172

160 assumptions about sexual activity of patients, and reactions to the vignette, but there is much more detail contained in these data that there was not ample space or time to explore here. The next chapter addresses the HIV-related data in this project, and links that data with the CCM data and clin ical communication about sexual health.

PAGE 173

161 CHAPTER 6 PERCEPTIONS OF HIV RISK Introduction The last chapter briefly raised the issu e of physician-patient communication about HIV and other STDs, and HIV risk percep tions among the DHP women, under the topic of sexuality and aging. This chapter focuses on the data gathered through this project that specifically deals with HIV risk perceptions among midlife and older women, physicians notions about HIV risk for their aging patients, and further explores the topic of clinical communication about sexual health. As pointed out in Chapter 2, HIV inf ection rates are remaining steady among persons over age 50. However notions about w ho is at-risk for HIV continues to focus on younger age groups, drug users, sex-workers and other high-risk groups. A large percentage of HIV infection in the 50 and over group is fr om heterosexual activity, and that is particularly true for women. Despite the trends, the perception of risk for midlife and older persons continues to be low among health care prov iders. For this project I asked both DHP women and physic ian about their pe rceptions of HIV risk. For the women this included personal risk perceptions and their notions about whether or not other women their age seemed to perceive of the risks associated with developing new sexual relationships after divor ce or widowhood. Physicians we re asked what percentage of their patients they felt were at-risk for HIV and other STDs, and were asked to delineate who was at-risk by age, gender, et hnicity, socioeconomic status (SES), and other categories they deemed relevant.

PAGE 174

162 HIV Vignette The HIV-related vignette introduced the HI V topic to the interviewees (Appendices C and D), except for physician who only comp leted the survey; there are specific HIV related questions on the survey (Appendix B). Briefly, the HIV vignette is the story of a woman who divorces a cheating husband, rema rries, and soon after her new marriage becomes ill. She visits the doctor often, but continues to suffer various problems with yeast infections, urinary tract infections, fevers, and fatigue. She is never tested for HIV because her physician assumes her health probl ems are not a result of sexual exposure to disease due to her age, marital status and reli gious activities. She is ultimately diagnosed with HIV when she is hospitalized for pneum onia. After the story, I ask the respondents to tell me their opinion of the case, along w ith other more detailed questions, and then asked them general questions about HIV risk perceptions. The respondents were less likely to find this story to be a typical physician-patient scenario than the hysterectomy story discusse d in Chapter 5. Only 39% (n=16) of the DHP women thought that this story was typical (Figure 6-1), and another 17.1% (n=7) felt that it was sort of typical, whereas over 80% (n=38) felt that the hysterectomy story was typical or sort of typica l (Chapter 5). Over a third (n=14) thought that this story was unusual and did not think it was re presentative of what typically occurs, particularly in todays health care interactions The nine physicians that heard this story had similar reactions to thos e of the DHP women, and a comparison of the differences between both samples was not statistically sign ificant. However, the physicians were a bit more likely to consider it typical (44.4% n=4) or sort of typical (22.2%, n=2), and less likely to consider it unusual (22.2%, n=2).

PAGE 175

163 16 4 7 2 14 2 1 40 2 4 6 8 10 12 14 16 18 DHP Women (n=41) Physicians (n=9) Typical Sort of Typical Unsual Don't know Figure 6-1. Frequencies of View of the HIV Vignette DHP Womens Views of the HIV Vignette There were no remarkable differences in how women viewed this story based on ethnicity, age, educational level, self-rated health status or mari tal status. The DHP women who thought the story was typical explained that physici ans might not be inclined to test everyone for HIV, particularly in th is case because the character in the story did not seem to belong to a high-risk group. Th e following comments demonstrate that these women are knowledgeable about what constitu tes a high-risk group for HIV/AIDS. A 45year-old woman said, I can almost see where an older woman [wouldnt be] tested for HIV . I can see where they're not going to HIV test ev erybody who walks in the door. I can understand that, especially since she wasn't a real, or didn't appear to be a real sexually active person. But in this day a nd age, there definitely should be a question asked. It should be part of the interview. Another woman in her mid-40s said Maybe if she was living on the str eet or a drug user, maybe they would consider maybe you're at ri sk for AIDS so they'd look for that, but

PAGE 176

164 since she didn't have that type of lifestyle-isnt that just what they go by--she wasn't considered at risk for that. A 60-year-old divorced woman could see how the physician might not think of STDs, but thought that some of his ot her assumptions were erroneous. I think it just sounded like a typical-what a doctor would do. He wouldn't automatically jump to the conclusion that she might have an STD, but church and marriage doesn't have anything to do with STDs. To me it doesn't anyway. There's still the chance. And if they're sexually active, there's always a risk when somebody's not monogamous. The impression that the physician was closed -minded was not uncommon, and the issue of marriage not guaranteeing monogamy and relative safety also came up often. One brought up the misplaced emphasis on religion, sayi ng, First of all I think he pre-judged her based on her environment and her age, because she was a church-going woman, and was married. . But he should of run differe nt tests because the first thing you want to eliminate are any STDs. Several women felt that the physician shoul d have asked more questions, but also that the patient should have volunteered mo re details about her life. There was a consistently strong theme of patient responsibility from many of the women. A woman in her late 40s remarked that the patient c ould have been more pro-active in finding out what was wrong, saying There's so many resour ces out there now that you can do better than just use your own doctor--have more of an idea what's going on with you. Another commented, I think Jenkins should have said something to her doctor. I mean you're as responsible for your health ca re as your doctor is. A woman in the same age group echoed this sentiment, I think the doctor should have asked her about what her previous life was like, and maybe they would have checked sooner and maybe she would have been treated sooner. I think the patient also should have volunteered her information also

PAGE 177

165 because she knows what her life was like . That's her responsibility--the patients are more responsible than they think because the doctors aren't psychic. Many of the women were disturbed that th e physician made assumptions about his patient that caused him to miss an HIV diagnosis, showing their knowledge of the importance of an early diagnosis for HIV tr eatment. Some had intimate knowledge of HIV/AIDS from people they know with the dis ease or who had died from the disease. Three of the African American women describe events or personal issues that the story reminded them of. A widow in her mid-50s rela ted an almost identical story to me that this scenario reminder her of. I had a friend that his wife got sick and he kept taking her back and forth to the doctor and they couldn't find out what wa s wrong with her and she went to the hospital with pneumonia and she died and when they performed the autopsy, that's when he found out that she had HIV. He had never known it, and . he had it. A woman in her late 40s who has undergone treatment for substance abuse, said, You know, I've heard about some things like that through rehab, so yes, it does remind me, but not like that because you know nowadays they do so much more, they've got so much more testing, you know the free HIV testing, and all that. A divorced woman in her late-40s; she seemed to feel that she had been fortunate to avoid the same experience as the woman in the story. That's a hard one. I mean this is a difficult story because any one of us could be in the situation . I c ould be that person because I was married. I got a divorce. And then 15 years--But you know wh at, I have a brother who died from AIDS . I'm just blessed that way, you know, a bout AIDS, but it could be me, you know? Many of the themes that pervade the in terviews with the women were common among the physicians as well. The theme of stereotyping patients based on various characteristics, the problem with the lack of questioning by the doctor, and the sense that

PAGE 178

166 this type of problem is less common now th an it used to be, though still occurring, were all topics that the physicians shared with the DHP women. Physicians Views of the HIV Vignette Only nine physicians heard the HIV vigne tte, two attending and seven resident physicians. There was no analysis by gender since all of the physicians were male, and ethnic distinctions were not possible with such a small number of participants. The two attending physicians thought the story was typical. One attending physician in his late 40s expl ained that he felt the story was fairly typical because physician are so influenced by SES and other so cial characteristics. A resident physician in his mid-30s thought that the type of scenario presented in the story occurs less often these days, but that the practice of judging so meone by how they are dressed or by their gender still happens among some physicians. The resident physicians were a little less incl ined to think that the story was typical. Two felt it was typical, two thought it was s ort of typical, two said it was unusual, and one said he didnt know how to characterize it. Several of the resident physicians noted that the education in their current progr am stresses that anyone can have an STD, and that they are encouraged to avoid st ereotypes that might cause them to miss something important to a diagnosis. A resident in his mid-30s said I think it's unusual; I don't see too many doctors going that far without doing other tests. A few used the phrase treating the whol e person, observing that the physician in the story only focused on one aspect of hi s patient causing him to overlook important clues to an accurate diagnosis. Others sa id that there was an obvious communication problem, since it was clear that the physicia n did not get a comple te sexual history. A

PAGE 179

167 resident in his early 30s felt that the physicia n neglected the possibili ty that his patients problems could be sexually transmitted because of age-related stereotypes, Basically, it sounds that the physician really wasn't open to the notion that someone of that age could be sexually active and/ or with multiple people. . I think it would be related to the age of the physicia n, as well. Because I would assume if you were a physician in your 50s approachi ng someone else in their 50s, you might have a different notion of it, versus a physician in their 20s thinking of grandma and projecting your own personal notions of people having-being sexually active later. I think a younger physic ian would have a tendency to maybe not pursue the sexual route as much as someone who is a physician closer to the age of the patient who realizes that probably sex doesn't stop at age 50. A widow in her late 60s also thought that the age of the physician could be a factor, but in the reverse of what the resident physicia n above suggested. She thought that a younger physician would be more likely to be aware of HIV issues than an older physician. Other than this consideration about the physicians age, the main difference between the physicians and the DHP women was the focus on patient responsibility. As mentioned in Chapter 5, the DHP women were much more likely to place the onus of communicating important health information on the patient. Physicians were likely to consider both parties responsible, but none felt that patie nts bore most of the responsibility. Despite their emphasis on patients being more pro-activ e with their health care interactions, the DHP women overwhelmingly thought that physicians had a responsibility to address the topic of STDs with their older patients (88.6%, n=39). This is likely due to their perception that sexuality is an important part of life regardless of age. Single, But Not for Long The tremendous support for physicians di scussing HIV and STDs with their midlife and older patients is testament to the importance of sexuality in later life for these women. Many were in new sexual relations hips after divorce or widowhood, or they expressed an interest in fi nding a companion. Their comm ents about the HIV vignette

PAGE 180

168 revealed that many are aware of the risks a ssociated with dating, as well as noting that both partners must share a commitment to monogamy to eliminate risk. The AARP study of midlife dating introduced in chapter two reported that over a third of women and men in the US between age 40 and 69 are single, and most are divorced. About a third of single women and men are in exclusive dating relationships, and almost the same percentage (32%) ar e dating non-exclusively. A little over 10 percent of those without a curr ent partner or dating are in terested in finding a date, another 12% or so are what the AARP study calls daters-in waiting, that is they would be interested if the right person happens to come along, while the rest do not have any interest in dating or finding a partner (AARP 2003: 2). Just under a quarter (22.7%, n=10) of the women I interviewed have a steady partner, and 15.9% (n=7) are dating, 6.8% (n=3) have recently ended a relati onship and fall into the AARP category of interested daters, 36.4% (n=16) are daters-in waiting, and only 15.9% (n=7) expressed no interest in findi ng a new partner or dating (Figur e 6-2). It is important to point out that men are include d in the AARP study and that they are more likely than women to be dating. Men also made up a large proportion of int erested daters and daters-in waiting. Some of the women in my study were content with their single status, enjoying the freedom associated with being on your own, but most indicated that they were very interested in finding a part ner, or had recently stopped dating someone. The fact that so many midlife and older singles are dating or interested in dating suggests that sexual health issues are relevant topics fo r clinical communication. Not surprisingly, the younger participants in both the AARP study and my study are the ones most likely to be involved with someone. The difference by age had

PAGE 181

169 marginal statistical significance (Table 6-1) but the fact that women age 45 to 55 out number older women more than 3 to 1, may have affected this and other comparisons by age. 38.6 6.8 36.4 15.9 2.3 0 10 20 30 40 50 Percentages Partner/dating Dated recently dater-in-waiting no interest dating unknown Figure 6-2. Partner Status a nd Dating Interest (Percents) Table 6-1. Cross-Tabulation of Age Range and Partner Status Age range Partner/dating Number (%) No partner Number (%) Fishers Exact Test 2-sided exact significance 45-55 years 16 (47.1) 18 (52.9) 56-68 years 1 (10.0) 9 (90.0) .062 Differences by ethnic group and age group we re not statistically significant in partner status and interest in dating, howeve r some of the differences are interesting, nonetheless (Figure 6-3). Pr oportionately, more women over age 55 indicate that they are not interested in findi ng a partner or dating (30%, n=3) than those 55 and younger (11.8%, n=4), though not numerically. Half of the older women indicated that they are dater-in-waiting, compared to a third of the younger women. Proportionately, more African American women are partnered or dating (45%, n=9) than are Euro-American women (34.8%, n=8), and more Euro American women indicated that they are a dater-

PAGE 182

170 in-waiting than did African American wo men (43.5% and 30% respectively). The woman who selected other for ethnic identity was excluded from this analysis and all other cross-tabulations using ethnicity. 9 8 16 1 0 22 1 6 10 11 5 4 3 4 3 1 0 1 0 0 2 4 6 8 10 12 14 16 18 Blk/AfricanAmerican Wht/Euro-AmericanAge 45-55Age 56-68 Partner/dating Dated recently dater-in-waiting no interest dating unknown Figure 6-3. Frequency of Partner Status by Age Group and Ethnicity The AARP study and its parallels to my study are notable because they clearly indicate that dating and sexual relationships are important to a large proportion of midlife and older singles, yet this c ontinues to not be represen ted in the general cultural perception. As a result per ceptions regarding HIV risks are still relegated to younger cohorts by the general popul ation, including many physicians. However, the DHP women have a relatively high perception of pers onal HIV risks, and this perception seems to correlate with the interest in dating. Dating and HIV/STD Risks Perceptions of personal risk for HIV were quite high for this sample. Almost twothirds (63.6%, n=28) said that they considered themselves to be at-risk. Predictably, partner status had a noteworthy impact on indivi duals risk perception (Table 6-2). Those

PAGE 183

171 with a partner or dating are the most likely to believe that they are at-risk for HIV and STDs. Table 6-2. Cross-Tabulation of Partner St atus and Personal HIV Risk Perception Partner status Yes, at-risk Number (%) No, not at-risk Number (%) Fishers Exact Test 2-sided exact significance Partner/dating 14 (82.4%) 3 (17.6%) No partner 14 (51.9%) 13 (48.1%) .057 This finding also coincides with age; sin ce more women aged 45 to 55 are partnered, and they are also more likely to think of them selves at risk (70.6%, n=24) compared to women aged 55 to 68 (40%, n=4). However, th is is still a much higher risk perception than reported in other studies. From st udies conducted in th e early 1990s, Binson, Pollack and Catania (1997) f ound that 90% of women aged 40 to 75 felt they were not at risk for HIV. Part of this higher perception may be due to changes that have occurred in the demographics in the spread of the dis ease since the early 1990s. It may also be a result of the particular mari tal and social circumstances of the women that I interviewed for my project. Equal numbers of African American (n= 14) and Euro American women (n=14) considered themselves at risk, but a sli ghtly higher proportion of African American women thought themselves as at risk th an Euro-American women (70% and 60.9% respectively). However this high risk per ception among Black women is higher than the perceptions reported by Rogers-Farmer (1999) who found that 98% African Americans (men and women) age 55 and over rate their ri sk as low or no chance. Theall et al. (2003) reported on age differences among African American women and HIV risk perception, noting that 55% of those aged 40-71 thought there was no chance of infection, compared to 29 % of those aged 18-29.

PAGE 184

172 Interestingly, 51.9% of women without a curr ent partner or dating feel that they are at-risk for HIV or other STDs (Table 6-2). The reason that some of the women without current companions mentioned for their sens e of HIV/STD risk was that they did not intend to remain single. A 54-year-old woman who is separated, said Yes, I will be [atrisk], 'cause I'm planning on getting back in th e into the game. A widow in her late 50s without a current partner said, AIDS doesn't pick its victims--anybody's susceptible to it. I don't care if it's the archbishop, I wouldn't sleep with anybody who wasn't wearing a Trojan. Yeah, I would be at risk if I didn't take care of myself, of course. Marriage Doesnt Always Mean Monogamy Those who said that they were at-risk al so talked about how you never really know the status of another person, for sure. A 55-year-old divorced woman, who has dated a bit since her divorce and currently has a st eady partner said, [E] ven if you're in a monogamous relationship, just because you are doesn't mean that the person you think is your partner isn't, you know, going someplace. A 50-year-old divorced woman who does not have a partner said, I know my situation when I found out my husband cheated on me for six months; both of us were tested, because it came out that he was having sex with someone from a foreign country. And because I di d have serious bacterial infections, and ironically enough, now that I'm not with hi m or having sex with him, I don't have any infections. Fancy that. Another comment about unfaithful partners and HIV testing came from a 52-year-old divorced woman. I think the first time I took an HIV test I felt kind of ooh, you know, but after a while I just say whatever, I woul d have to kill him anyway (laughs). One way or another it's gotta be all right cuz he will be dying first!

PAGE 185

173 Interestingly all of the women who are se parated described themselves as at-risk (Table 6-3). The cross-tabulation of HIV risk and marital stat us shows that being separated rather than divorced or widowed may be significant in shaping risk perception. Table 6-3. Cross-Tabulation of Marital St atus and Risk Perception for HIV/STDs Marital status Yes, at-risk Number (%) No, not at-risk Number (%) Likelihood Ratio Value (2-sided sig.) Divorced/ Ended Union 12 (50.0) 12 (50.0) Widowed 7 (63.6) 4 (36.4) Separated 9 (100) 0 (0) 9.991 (.007) df =2 A 53-year-old woman was concerned about the fidelity of her husba nd that she clearly expects to unite with, saying, Right now I'm separated from my husband and he's overseas. I have this fear in my heart th at--you know--I know within myself he can get carried away. So I am concerned [that he co uld put me at risk]. Yes, I'm concerned about that. Several of the women, indicated that they may reconcile and that there may have been sexual risk taking during the separatio n, or they separated due to infidelity and fear they may be infected. E ither way, their circumstances have heightened their sense of risk. Of course, almost half of the women without partners (48.1%, n=13) did not consider themselves to be at risk for HIV/ STDs. Some said this quite matter-of-factly, noting that were not engaging in any sexual activity and they were not planning to do so in the future. This was sometimes due to a l ack of not interest in finding a companion, or because they were too fearful of all the new diseases that one has to be concerned about now. One divorced woman in her late 50s sa id, No ma'am, because it's like I told you, I'm scared and I don't want to be bothere d. And um, I don't like, what do you call it, protection [condoms]? I don't care for that. This sentiment about condoms corresponds with other research that has found low ra tes of condom use among older women (Binson

PAGE 186

174 et al. 1997; Mack and Bland 1999; Theall et al. 2003; Yates et al. 1999; and Zaboltsky 1998). Although a large percentage of the DHP women believe they are at risk for HIV and other STDs, that does not necessarily mean they are practicing safer sex. This is not unusual, as heightened risk perceptions have been shown to have a weak connection to risk behaviors in the case of HIV/AIDS prevention (Poppen and Reisen 1997; Stall and Catania 1994; Theall et al. 2003). The Issue of Practicing Safer Sex Although my research did not focus on act ual sexual activity, several of my informants volunteered information about their sexual activities, both currently and in the past. A 55-year-old divorced woman, who is act ively dating confessed th at she is at risk for HIV and other STDs, because she doesnt always practice safer sex. One 45-year-old woman, who is separated but tr ying to reconcile with her husband, said I've had other partners, during the separation and I think there always is a risk. I think the risk comes down tremendously when you make a commitment to each other. Another separated woman in her mid-50s brought up the newness of practicing safer sex. She does not have a current partner but ta lked about her concerns over condom use for dating in the future. She said, Well, I have never had sex with a condom--never had to . That's going to be a whole new experience for me when that comes to pass. Asking, you know, let's use protection, [or] do you have any?, or going to a store and buying some condoms to keep in my night stand--totally something that I have never done and never expected to do. The issue of practicing safer sex was an interesting finding in the AARP study as well. The report notes that a minority of singles (21%) us es protection regularly when having sex. This is thought to be related the fact that the da nger of pregnancy is past for many of them, and many probably do not think th ey are risk from disease. Women in the

PAGE 187

175 AARP study are more assiduous about protecting themselves than men, with about half (48%) always-using condoms. However since th is is especially true for 40-49 year old women the authors suspect that it may have more to do with pregnancy prevention than HIV/STD prevention. A particul arly disturbing finding is that the [n]on-exclusive daters (28%) are more likely to have unprotected sex than other groups in the study (AARP 2003: 39). The issue of sexual safety is a tricky one Most psychosocial models of health behavior place the onus of responsibility on the individual as t hough other cultural and relational factors are absent. These models ignore the affective relational expectations and gendered power dynamics that impact the be haviors of social act ors the exercise of safer sex (Sobo 1995; Ragsdale 2002; Stansbury and Sierra 2004). The research by these authors on women from disparat e ethnic, socioeconomic and cultural groups showed that the women appeared to be relying on similar cultural models of affective relationships that resulted in risk denial and ultimatel y higher risk sexual re lations. The cultural consensus data presented in chapter four show ed that the issue of safety in new dating relationships is an important element in the domain of sexuality for the DHP women, but also highlighted the ideal aspe cts of affective relations. Cultural Consensus Data Related to HIV/STD Risk The domain of sexuality for midlife a nd older women that are newly single contained several elements that pointed to th eir concerns about HIV and STDs. The freelists included such repeated terms as dangerous, use protection, disease risk, and more careful There were also many idiosyncratic te rms and phrases that focused specifically on the HIV/STD risks associated with dati ng and finding a new sexual partner. The multidimensional scaling (MDS) representation of the pile-sort data revealed that they

PAGE 188

176 linked some of these terms in significant ways. The core terms shown in Figure 4.1 (Chapter 4, p.103), that most participants grouped together to varying degrees of inclusion to describe new relationship characteristics, paired such terms as sexy, beautiful, and desirable with use protection and more careful within the topic. This combination reflects the tension between idea ls and realities of dating and new sexual relationships. The physicians, on the other hand, did not include use protection and more careful within a similar topic on relationships and this reflects their emphasis on younger cohorts when considering STD and HIV risks. The ratings data also support the notion that the physicians and the DHP women differ in their ideas about sexual ity and dating in later life. It is interesting to note the different themes that appear in how thes e samples were conceptualizing the issue by comparing the terms of agreement with the te rms of disagreement. The women describe sexuality in very positive terms, as well as focusing on safety and being careful with regard to the risks associated with dati ng. Whereas physicians tend to focus on the problematic aspects of sex and aging, w ith much more emphasis on lack of sexual activity due to physical or medical problems. These data provide usef ul clues as to what physicians need to know about midlife a nd older women. DHP women believe that mature women are sexy and beautiful and but they are also concerned with personal safety as they enter new relationships afte r divorce or widowhood. The fact that these women rated use protection as highly relevant and that physicians rated it as less relevant, supports the other findings in this research about HIV/STD risk perceptions. Higher Than Expected Perceptions of Risk Personal HIV risk perception is relatively high for this group, and this may have several reasons. First of a ll, I posed the questions about HIV after reading the HIV

PAGE 189

177 vignette. This was a tough call in terms of research design. I was concerned about alienating participants if they felt the main fo cus of the study was HIV. Since this project has a broader focus of clinical communicati on about sexual health, it seemed wise to keep the HIV component low-key. So me of the comments from these women regarding infidelity by husbands and lovers suggest that their perception of risk may not necessarily be tainted by the HIV story, but ra ther by their feeling th at their partner has put them at risk. This perception of risk seem s to also be tied to being separated more so than to the other marital status categories. It is also relate d to their lack of experience and/or reluctance in using condoms. In a ddition, several women had personal knowledge of a friend or family member with HIV/AIDS or who had died of AIDS. Any of these factors or some combination of them could e xplain the high rate of HIV risk awareness, as opposed to the influence of the HIV vignette. Another factor that suggest that the HIV vignette did not unduly affect their perceptions of HIV ri sk is the relatively high proportion of women who re vealed that they had be tested for HIV. Testing for HIV/AIDS Although I did not systematically ask the wo men if they had ever been tested for HIV, several volunteered information that th ey had been tested. When the interview flowed in the direction of personal risk a nd testing, I did ask th e question, but did not include it in my list of questions to ask everyone. Just over 40% (n=18) of the DHP women I interviewed revealed th at they had been tested for HIV. There are also some interesting differences by ethnicity and ag e range (Figure 6-4). Younger women (n=16) and Black women (n=11) were the most likely to reveal that they had been tested. This is consistent with other research findings delineated in Chapter 2 showing that younger

PAGE 190

178 persons and African Americans are generally more likely to have been tested (Mack and Bland 1999; and Florida Depa rtment of Health 2003d). 55 30.4 47.1 20 45 69.6 52.9 80020406080100Black/AfricanAmerican White/Euro-American Age 45-55 Age 56-68 HIV tested No HIV test Figure 6-4. Percentage Test ed for HIV by Ethnicity and Age Range (Self-Reported) Some who revealed that they had been te sted for HIV remarked that they had done so after finding out that a spouse or lover ha d been unfaithful. A woman in her mid50s, said, Oh definitely. I've had my test. A couple of months ago I went, because my husband had an affair with the woman. So I went and had the test done just to make sure. A divorced White woman in her early 60s talked about the te sting, others risks, personal risk, relationships, uns afe sex, and not trusting men. So many people like me are divorced and we date and we're not always careful. We don't like condoms. We hardly ever use them. I mean it's just, we take chances, thinking that well it doesn't happen in my age group or in my class, but it does. But when you have a steady boyfriend, you don't think of yourself, and you don't know what he's doing on the days he's not seeing you, and I don't trust men anymore, so I figure they're going to be out doing something. No matter what they say, they lie. I just flat don't trust th em, because I know them. I know myself, and I've lied to men before about other men, so if I'll do it, then I'm sure they'll do it 'cause I'm not that different. . But I've had myself tested.

PAGE 191

179 A divorced Black woman in her mid-40s, w ho has a current partner, said, Even with me, with my ex-partner, there was a point in our relationship where he did cheat on me. And I don't know if he used protection to this day or not. So therefore, I'm not ashamed to get an HIV test. A couple of Black women mentioned that they get tested for HIV regularly. A separated woman in her ear ly 50s, said, I see Free HIV, I go. This lady was asking me, How many times you going to come in? I say I want to know. The relatively high percentage of women who have been tested may be due to fear of exposure, as in the quotes above. However, in some cases it may also relate to having personal knowledge of someone with HIV or who has died of AIDS. Know or Knew Someone with HIV/AIDS and HIV Testing In addition to revealing they had been tested for HIV, several women mentioned that they knew someone with HIV/AIDS, or knew someone who had died of AIDS. In some cases the person was a family member. Just under a quarter of the DHP women (n=10, 22.7%) know someone with HI V/AIDS or who had died of AIDS. Of those, four had a family member die with the disease. Although I did not ask them if they knew anyone with HIV/AIDS, several of my questions could lead to this topic. For example, after the HIV story, I asked them if the stor y reminded them of any similar events or stories that they knew of from family or friends or even in the media. I also asked them to tell me a little bit about what they knew about HIV/AIDS. These questions were what led some of the women to reveal their personal experiences w ith the disease. Since HIV/AIDS is disproportionately devastating the African American community, it is not unexpected to find more African American women in my study with family or friends who have HIV/AIDS or w ho have died with AIDS Cross-tabulation of

PAGE 192

180 ethnicity (other ethnic identity excluded) and know/knew someone with HIV/AIDS, and a cross-tabulation of Tested for HIV and know/knew someone with HIV/AIDS, revealed statistically sign ificant distinctions between groups (Table 6-4). Table 6-4. Cross-Tabulation of Know/kne w Someone with HIV and Ethnicity and Tested for HIV Knows/knew someone with HIV/AIDS number (%) Does Not know anyone number (%) Fishers Exact Test 2-sided exact significance Black/African-American 8 (40.0) 12 (60.0) White/Euro-American 2 (8.7) 21 (91.7) .028 Tested for HIV/AIDS 9 (50.0) 9 (50.0) No HIV test mentioned 1 (3.8) 25 (96.2) .001 Far fewer White women know someone w ho has or had HIV/AIDS (8.7%, n=2) than Black women (40%, n=8). Th is percentage appears to also be strongly related to whether or not they have been tested for HI V. The relationship here is striking, with most who know of someone with HIV/AIDS having been tested for the disease. These comparisons are significant and appear to signal an important rationale in motivating HIV testing. The high-risk per ception for many of thes e women is motivating some to get tested, however it may not be mo tivating safer sex practices as is evident in some of their comments above about themse lves, and even more so in the comments below about other womens risk perceptions and behaviors. Other Women in Age Group and Perception of HIV/STD Risk The notion that many singles in this cohor t do not regularly practice safer sex is most evident in my study when I asked th em what they thought of other womens perceptions of risk for HIV and other STDs. Although the women in this sample appear to have a relatively high percepti on of risk, they tend to think th at only slightly more than half of women their age, that are divorced, separated or widowed, think they are at-risk. Figure 6-5 shows that a slight majority said yes or yes, most to the question do you

PAGE 193

181 think that other recently single women your age and older think that they are at-risk for HIV? and slightly less than half said no or no, not most to this question. 25.6 27.9 9.3 30.2 6.9 Yes Yes, most No Not most Don't know Figure 6-5. Percentage Who Think Other Recently Single Women Know of HIV Risk Figure 6-6 shows the percentages of t hose who thought others were aware of HIV/STD risk by age group, ethnicity and mar ital status. These percentages are more useful for considering trends within sub-gr oups rather than for intergroup comparisons because of the small sample size. Since the question about o thers perception of risk was framed around other women their age, it is not surprising that women over age 55 were more likely to say no or no, not most since older women were less likely to think of themselves as at risk. Younger women thought that over half (55.9%, n=19) of othe rs in their cohort were aware of the risk, whereas older wo men only thought that 40% (n=4) of women their age were aware of the risk. Separa ted women overwhelmingly thought that other women were cognizant of the risk (88.9%, n=8) and since they were the most likely to think of themselves at risk they may feel that more women share their perspective.

PAGE 194

182 Finally, Black women tended to think there wa s a high perception of the risk in their cohort (60%, n=12), but this was just slight ly higher than the perceptions of White women (47.8, n=11). Slightly higher perc eptions among Black women are possibly related to the fact more of them know someone with HIV/AIDS or simply because HIV is more prevalent in their community. 60 47.8 55.9 40 45.8 36.4 88.9 35 39.1 35.3 50 37.5 63.6 11.1 13 5.9 10 12.5 5 2.9 4.20%20%40%60%80%100%Blk/ African-American Wht/Euro-American Age 45-55 Age 56-68 Divorced/ended union Widowed Separated Yes, or Yes most No or Not most Don't know question not asked Figure 6-6. Others At Risk by Ethnicity, Age Group, and Marital Status (Percents) Most often these respondents qualified thei r ideas about the perceptions of others by saying that they think that others know of the risk, but don t do much about it, or that they are too new to negotiating safer-sex. That is, they often made the distinction between knowledge and actions. They also th ink that other women think they are too old to be at risk or ar e in denial about the risk. One 50 year old divorced White woman said, Probably in the back of their minds [they know of the risk], but they really don't, when it comes right down to it, think that it can happen to them. They don't want to face reality, 'cause they want to live for the moment. You're in that situation and you don' t think about that. Referring to her own

PAGE 195

183 circle of friends sepa rated White woman in her mid-40s said, I think we're a lot more cognizant than any of us would have been 10 years ago. Maybe not as vigilant as we probably should be, but certainly it would never occur to me 10 years ago to worry about that stuff. The issue of age as a factor in risk pe rception also came up a few times. A 52 year old divorced Black woman remarked, Yeah, be cause that is the one thing we probably [say] Oh, that is not going to happen to me cause I'm too old for that. She continued, Most of these old man now are out ther e with the young girls that's given them everything. And here they come back to us and we be thinking that they ain't doing anything. So we're in denial thinking that he ain't bringing anythi ng back to me. On the other hand, there was also a notion that older women are more responsible and knowledgeable as a Black 65 year old wi dow noted, The older they are the safer they are because they've got the experience. A widowed White woman in her early 50s said, Yes, I think they're pre tty well versed in general. However this greater knowledge did not necessarily translate into action. A separated woman in her mid-50s said, I think that most of the wome n in this age group are aware, that I know. But I think they need to be more conscious of having to take action to protect themselves. Another di vorced woman aged 55, echoed this sentiment and even acknowledged her own lack of follo w through. Most of the women I know are a little younger than me and they are very awar e. Now whether they practice safe sex or not I think that varies, and I found the same with myself--I' ve been careless. A separated woman in her mid-40s said, We're more uneducated than the youth. If we are divorced, separated, widowed, it becomes an issue and we have to edu cate ourselves all over again on sexually

PAGE 196

184 transmitted diseases. I know less than my daughter she's in her 20s. Because I felt I wouldn't ever have to know that. But when I talk to my friends and acquaintances that have found a new part ner and are having sex, they're not doing anything [to protect themselves]. A widow in her late 40s who lost her daughter to AIDS, felt that a persons socioeconomic status, and emotional state gr eatly affected sexual risk-taking. Yes, I think we're more at risk than some of th ese young people because we're more settled and sometimes now, the way the world is now. Then especially when you're unemployed, lost a loved one, and depressi on, low self-esteem, sometimes you can just fall into a lurch where you just dont think sometimes. This comment and the ones above point to some of the cultural and psychosocial issues in wo mens negotiation of safer sex, in that there can be significant constraints on sexual pr actices, including how safer sex practices correspond to relationship ideals, gender c onstructs and socioeconomic limitations. The fact that many of these women mentioned the disconnect between knowledge and action with regard to safer sex, did not inhibit th em from urging physicians to educate older patients about the risks. Suggestions for Physicians Another way that I addressed their percep tions about others risk want to ask whether or not they thought physicians s hould address HIV/STDs with their older patients. Most gave short, yes or no responses to this, but some ela borated a bit more. A divorced 60-year-old woman said, I ju st think that when you go for your annual mammogram and Pap smear . it would be a good time to bring it up, because so many husbands are unfaithful, so many people like me are divorced and we date and we're not always careful. Another divorced woman in her early 60s said, Oh no, it's a hidden

PAGE 197

185 issue, nobody talks about it. Doctors are not addressing anything because they don't want to rock the boat they don't say anything. A 55 year old widow, said And I think that HIV should be stressed in older patients more so than the young, because we ju st fail to realize because we have one sexual partner, that that sexual partner may see someone else that's younger than us and they may contact it from someone else, at whic h in turn they give it to you and you never know you have it. The idea that age impacted either their knowledge or actions was frequent, and was cited as a reason for physicia ns to address sexual health risks topics with their older patients. Physicians also c onsidered age an importa nt factor in patient HIV knowledge and risk. However, many did no t think their older pa tients were at risk, and they considered them to have lower le vels of HIV knowledge than younger patients. Physicians Perceptions of HI V Risk for Their Patients I asked the physicians How well informed are the majority of your patients (age 45 and older) about HIV and othe r STDs? They seem to think that the vast majority of their midlife and older patients are poorly or partially informed (Figure 6-7). The differences between attending and resi dent physicians were slight. Attending physicians thought 75% of their midlife a nd older patients are partially or poorly informed compared to 83.3% of resident physic ian. Despite their sens e that many of their patients lack adequate knowledge about HI V risk, many tend to consider their older patients as not being at risk for HIV. The in terviews showed that there is a tendency to think that younger adults are both more knowledgeable and more at-risk. One of the survey questions (Appendix B) and one of the in terview questions (Appendix D) asked the percentage of thei r adult male and female patients they considered at risk. However, because I redesigned the survey mid-way through the

PAGE 198

186 project, three of the physicians were not asked this question. They generally rated their women patients at slightly lower risk than th eir male patients (Figure 6-8). Averages of their responses showed that physicians consid ered 51.43% of their ad ult male patients to be at-risk for HIV, (range 15-100%; SD 25.5), a nd 50.18% of adult female patients to be at-risk, (range 10-100%; SD 24.7). 10% 10% 43% 37% Well Informed Somewhat Informed Partially Informed Poorly Informed Figure 6-7. Patients Aged 45+ Informed about HIV/STD risk. 3.6 0 14.3 17.9 50 42.9 14.3 17.9 17.9 21.40 10 20 30 40 50 60 Females At-RiskMales At-Risk less than 10% 11-25% 26-50% 51-75% 76-100% Figure 6-8. Percentage of Males and Females At Risk.

PAGE 199

187 Several said that they thought that males ty pically had more sexual partners than women, but others remarked that they treat mostly women and that women are more likely than men to present with STD symptoms. Overall Physician Perception for HI V/STD Risk for Patient Population. Doctors consider 52% of th eir adult male patients to be at-risk for HIV/STDs Doctors consider 51% of their adult fema le patients to be at-risk for HIV/STDs The most at-risk groups are those in their late teens and ages 20s-30s Perception of risk declines with advanci ng age, female status, and intact marriage Most risk focused on persons with multiple partners, sex workers/clients, and drug users Since most of these physicians treat low-in come, inner city patients, their focus on higher risk groups such as sex workers and thei r clients, and drug-user s is not surprising. Given that the highest risk populations tend to be younger, their focus on youth is also understandable. However, several acknowle dged that they should do a better job of talking about sexual health issues with thei r older patients. Their focus on high-risk populations also lead some to give more atte ntion to ethnicity in their explanations of who is at-risk. Racial/Ethnic Perceptions Perceptions of patient risk were often linked age, mar ital status, and behavioral issues (i.e., drug use, prostitution, and multiple sex partners), but also to ethnicity and SES. As stated in previously, the physicians in this sample treat a largely lower income, African American population. Se veral of the physicians menti oned this in reference to which patients they perceived as at-risk for HIV and other STDs. A resident acknowledged that the largel y poor, African American co mmunity they serve might affect their views about who is at risk. The majority of our patients are African

PAGE 200

188 American so it might skew a little bit and it's also our community, it's more African American and it's different cultural, ethni city and maybe fina ncial situations. A resident physician in his mid-30s said, In this community it's very high risk living here. Many patients have multiple partners and many patients are using drugs in this community and it's a problem. . For 18 and above Black population, low social economic status, it has a more higher risk of sexuality and promiscuous activity . . An African American resident physician ma de a point of saying that she does not distinguish between ethnic groups. As l ong as you are coming through that door you are getting that question [about sa fer sex]. However, she also noted that the area has a higher rate of HIV than other parts of the area. Apparently in this zip code it's one out of three women are HIV positive. I just thi nk everybody needs to be aware of it--you just can't any chances anymore, it doesn't matter how old or how young you are, I'm bringing it up. An attending physician in his mid-50s agreed that et hnicity was not a factor, but further clarified that younger age groups would be at greater risk than older ones, and that he did not consider ma rried persons at-risk. An attending physician in his late 40s explained that his perception of risk centered on the characteristics of the patients he sees and the trends hes seen in STD rates. Well I see a very diverse population . I do take care of drug addicts, I do have prostitutes in my practice, I do have HIV positive patients in my practice. But I'm just very aware of the fact that there is an awful lot of multiple sex partners, unprotected sex, and actually I think we're not in a particularly high risk area of the country--nothing compared to New York City or Dade County, but we've been seeing increasing STD levels for the last 10 or 15 years I think, maybe even longer. So it's certainly not something to be ignored.

PAGE 201

189 Age and Marital Issues Patient age and marital stat us had a considerable impact on physician ideas about HIV risk. An attending physicia n in her mid-40s suggested that all of her patients are atrisk, but that she feels that younger patients ar e more at-risk, because they are more likely to have multiple sexual partners. However, I've got a number of older people who are widowed or divorced and you know, and some of, most of my older people, most of them would be at risk from sexual contact. And so I do have a, you know, some older people who are playing the field, so to speak. Yeah, for me it 's not [effected by ethnic group or sexual orientation]. It's, you know, it's an age th ing. It's like, you're young, you're healthy you're out there. You know. Whatever, so for me it's probably more age and general physical abilities, you know, although I have had a patient who was in her 80's in a wheelchair with a chronic Foley catheter who was still having intercourse and I just wasn't frankly thinking that th at was in her in her life, but it was. An attending physician in his early 50s di stinguished married persons from single ones, but also factored in age when categorizing their risks. Older people I have in my practice I'm assu ming are sexually active or are sexually active in a monogamous rela tionship, and some of them I've known for probably over 15 years, and I guess I'd be very surprised if they were not in a monogamous relationship In that patient population, I just don't ask [about multiple partners or safer sex practices] at all. . I guess because you'd like to assume that everybody's who's married and in a monogamous relatio nship isn't doing anything outside of the marriage or outside of that relationship, so I'd probably make the assumption there that they're not at risk. Another attending physician in his late 40s looks at it differen tly, based partly on what he has experienced. He gave a re sounding One hundred percent! He explained that he feels that way because he was placed at risk through no fault of his own, saying but I never thought it would be me, so. Before he was placed at risk he says he would have probably responded 50% to the question.

PAGE 202

190 One attending physician felt that younger phys icians would be reluctant to address sexuality issues with their older patients b ecause they assume that they're older and they're not having sex anymore, which is totally ludicrous. He continued, I think the range of normal for sexual desi re and sexual behavior in the geriatric population is amazingly diverse, but it take s a mature physician to sit with an 80 year old and ask her if she is sexually active or engage d in any kind of high-risk activity. . But those risk factors are there and if a person is at risk it's a physician's job to make sure that they're educated as to how they should protect themselves. So I tend the raise the issue w ith most of my geriatric patients, unless I perceive that it's something that they might not be very comfortable with, or unless it's a non-issue because there are elderly patients where they're not sexually active and there's no real issue with AIDS. [L ike for the woman who says] my husband died 20 years ago and I haven't had sex since he died. Then you don't have to worry about STD risk or talking about c ondoms . unless they come back later and say I met this person, and it's amazing how often it happens. Overall, these physicians tend to think that most of thei r married patients and older patients are generally not at ri sk for HIV/STDs or are at lo wer risk compared to younger age groups, but many are aware that they might be missing prevention opportunities because of their assumptions. Medical Education and HIV Prevention Readiness I was also curious about how physicians felt about how their education prepared them for HIV issues. Over half felt that they were well prepared, wh ile others seemed to feel they could be better prepared (Figure 6-9). To get an overall sense of the how much the residents differed from the faculty, I collapse the Likert scale values to put agre e and strongly agree together, but it was not necessary for disagree and strongly disagree because no one selected strongly disagree. Resident physicians were less lik ely than attending physicians to select disagree for this question (11.1% and 38.5 % re spectively). The recentness of education seems a reasonable explanation for that. Ho wever, almost two-thirds (61.5%) of the

PAGE 203

191 attending physicians agree that they have b een well prepared, and this was not affected by age or year of medical education, so continuing medical education, including individual efforts to stay current with the l iterature, is likely a fact or here. Chi-square analysis showed that these differences were significant (Table 6-5). There were no significant differences by gender, though females were a bit more likely to select neutral (28.6%, n=4) compared to males (11.8%, n=2), and they were also less likely to select agree (50%, n=7) compared to males (64.7%, n=11). 12.9 45.2 19.4 22.6 Strongly agree Agree Neutral Disagree Figure 6-9. Medical Educa tion and HIV Preparedness Table 6-5. Cross-Tabulation of P hysician Status and HIV Education Physician status Agree Number (%) Disagree Number (%) Neutral Number (%) Likelihood Ratio Value (2-sided sig.) Attending Physician 8 (61.5) 5 (38.5) 0 (0) Resident Physician 10 (55.6) 2 (11.1) 6 (33.3) 9.059 (.011) df =2 Still most of the physicians, both atte nding and resident, acknowledged in the interviews that participating in this research project has broaden their view of HIV risks and made them aware of their own age-rela ted perceptions about sexuality. However,

PAGE 204

192 this acknowledgement may have little impact on their behaviors in rega rd to talking about sexual health with more of their older patients. Talking About HIV Risk The practice of routinely talking to pa tients about sexual risk was brought up by a few of the physician in response to my ques tions about how they address sexual health issues with a variety of patients. One of th e residents in his earl y 30s felt that it was always appropriate to talk about HIV/STDs with his patients, particularly if he is aware that they are sexually active. I guess in my approach I always ask them how their sex life is first--basically if they've been sexually active or not. . I don't think that I need to spend more time on sexuality or STDs if they don't have sex at all. . Twenty year olds or 30 year olds, single patients; I tend to ask them bluntly as well. Married couples I tend not to. My widows or elderly if they're as king for Viagra or some other estrogen vaginal cream, who are either widowed or separated, I do talk to them about sexually transmitted diseases and I'm pretty much blunt with them. Only the married people... like I said I can see how that can happen where you just lose track and assume that they are monogamous. The comment above is typical of what many of the physicians said. They mostly talk to those they think are at risk, which tends to be younger, unmarried patients, and those they suspect might engage in some high-risk beha vior. A few physicians mentioned that they were concerned about offending older patients and this is consistent with previous research on why physician do not discuss sexua l health issues with older patients. Indeed, the perceptions and prac tices of these physicians were similar to those in other studies in terms of risk perception being re legated to younger age groups and a reluctance to discuss sexual issues with older patients (Epstein, et al 1998; Frank and Harvey 1996; Haas and Coe 1997; Kerr, et al. 1996; Makadon and Silin 1995; Skiest and Keiser 1997; Ross, et al. 2000).

PAGE 205

193 Results from the non-physician sample of th is research suggest that concerns about offending patients might be misplaced. Th e DHP women overwhelmingly agreed that the physicians should talk to their midlife a nd older patients about HIV and STDs. When those that said yes (88.6%) are combined w ith those that qualifie d their answer saying yes, if they think somethings wrong a whopping 94.5% think that the issue should be addressed. One of the DHP women said I would suggest that all women or men, if you're out there during that time, if you're spouse has died, if you're separated, be careful, you go to the doctor, you got to talk about this with yo ur doctor and be tested, and don't be afraid to ask your doctor questions about your sexuality b ecause it's a wonderful thing. And I don't think we should have to give it up because we're getting older. Some of the women made clear that it was not always appropriate to raise sexual health topics, but they were also unambiguous on th e point that physicians should not assume someone was not sexually active simply because they were older or unmarried. Discussion and Conclusion The findings presented in this chapter reve al both consistencies and inconsistencies with previous research on the topics of HIV risk perceptions among women and the factors involved with perceive d risk, physicians notions of pa tient risks, and talking to patients about sexual health issues. DHP wo men and physicians were in agreement about the HIV vignette, with both raising the them e of stereotyping pa tients based on various characteristics, the problem with the lack of questioning by the doc tor in the story, and the notion that this type of s cenario is less common now than it used to be. The one clear difference was that the DHP women put more communication responsibility on the patient. The topic of stereotypi ng is of interest since it was cl early an important factor in shaping these physicians ideas about who is at risk for HIV and other STDs, and appears to influence their practices regarding talking about sexual health issues.

PAGE 206

194 Risk perceptions among the DPH women were quite high, particularly in comparison to previous research. Although th ese findings followed previous trends in age and ethnic differences, there was still a much higher risk perception than reported in other studies. Part of this higher perception ma y be due to changes that have occurred in the demographics in the spread of the dis ease since some of the earlier studies were conducted in the early 1990s. It may also be a result of the particular circumstances of the women interviewed for this project, in that this study focused on recently single women, who might be more likely to have experienced marital infidelity, or partner infidelity, than women in previous studies. Additionally, some re searchers have noted that there are multiple domains of HIV risk perception such as a perceived likelihood of becoming HIV infected as opposed to or co mbined with the level of worrying about being infected, which make it difficult to co mpare findings across studies (Theall et al. 2003). My project did not explore all of the possi ble domains of risk. However, some of the responses about personal risk and the risk of others revealed in teresting distinctions regarding knowledge and behavior, complacen cy in using protection, partner and dating status, denial of risk status due to age cohort, general distrust of men, and a sense of risk related to knowing someone with HIV/AIDS. Their comments also pointed to some of the cultural and psychosocial c onstraints in negotiating safer sex, and risk de nial and how safer sex practices do not necessarily correspond to the ideals of personal relationships. These choices are not always so straight forw ard, and socioeconomic issues, relationship power dynamics and maintaining intimate c onnections are important influences.

PAGE 207

195 The cultural consensus data from chapter 4 also hinted at the tension between relationship ideals and the risks associated wi th finding a new partner. The fact that many women, whether partnered or not, thought that new rela tionships brought with them greater HIV/STD risk is congruent with the cultural consensus findings. The DHP women included risks associated with dating a new partner as an important relationship element in shaping their domain, whereas phys icians did not. The women also showed more awareness to the fact that marriage is not always a guarantee for monogamy. Most physicians, on the other hand, tended to character ize married persons as not at risk for HIV or STDs. Finally, perhaps the most striking finding here is the significant number of women tested for HIV, and the relationship of bei ng tested to knowing someone with HIV/AIDS. This issue was also impacted by ethnicity, with more Black women being tested than White women, and more Black women knowing someone who has HIV or has died of AIDS. Since HIV/AIDS is devastating so many African American communities, these differences between ethnic groups in this regard are not unexpected. However, the importance of knowing someone with HIV/AI DS appears to be significant and may signal an important impetus in motivating HIV testing. These research findings show that age, et hnicity and marital status are significant variables in physicians notions about HIV risk for their patients, influencing discussion of sexual issues in clinical contexts. Their ideas about who is most at-risk appear to limit their perceptions such that they negl ect important primary prevention opportunities. This study supports and enhances on previous research about risk pe rceptions and reveals what physicians need to know and incorporate into their practices to better address HIV

PAGE 208

196 risks for midlife and older women patients. The primary points that physicians should learn from these findings are: 1) that lower overall risk for older pe rsons in general does not equate to no risk for older individuals; 2) that sexuality con tinues to be important for persons as they age, including for thos e who may not have a current steady partner, and; 3) that their fears ab out offending older patients with discussion of sexual health topics are largely unsupported by these empirical results. The next and final chapter summarizes th e project findings and further triangulates the different types of data and analysis used. The limitations of the project are delineated, and it concludes with a discussion of the significance of the data and its utility for improved health care communication re garding sexual health issues and HIV prevention in clinical context.

PAGE 209

197 CHAPTER 7 DISCUSSION AND CONCLUSION This project explored differi ng conceptions of mature female sexuality by recently single women and primary care physicians. Research about clinical communication suggests that patients and physicia ns may differ in their perceptions to such an extent that effective health communication is hindere d (Cline and McKenz ie 1998; Fisher 1993; Hahn 1995). Although the physician and patient ma y share much of the same social and cultural makeup, the specific cultural domain of health care may or may not be shared. My research explored the cu ltural domain of sexuality for mature women that are recently single. I investigated midlife a nd older womens concepts of sexuality and sexual health for themselves and other women their age. I also looked at how primary care physicians conceptualize sexuality and sexu al health for midlife and older patients, and how these perceptions are shaped by ag eand gender-based assumptions and their impact the clinical encounter. The main c oncern for primary prevention in clinical practice was how and if womens sexuality c oncepts, and those of physicians for their patients, related to HIV risk perceptions, and their ideas about discussing sexual health in clinical contexts. This final chapter provides an overview of the research findings a discussion of the limitations of the project, the projects signifi cance and suggestions for future research. The next section reviews the demographic info rmation about the samples in this project and is followed by an overview of the findings presented in the previous chapters.

PAGE 210

198 Synopsis of Demographic Data and Overview of Chapters This project interviewed 44 recently si ngle women attending a state sponsored displaced homemaker program, and 31 primary car e physicians in the same urban area of north Florida. Data collection included de mographic surveys; vigne tte-based interviews; and cultural consensus mode ling techniques from cogniti ve anthropology. The women were recruited in-person at group meetings at the community college where the displaced homemakers program (DHP) meetings took place, from October 2002 to June 2004, and the physicians were recruited at their pl ace of work between April 2003 and June 2004. DHP Sample The final sample of DHP women re spondents included 20 self-identified black/African American women and 23 self-i dentified white/Euro-American women, and one self identified as other. Three also noted another aspect of their iden tity, but out of respect for their primary choice, they were s ituated within their chosen category for all analyses. The women ranged in age from 45 to 68 a nd the average age was 52. Most of the women were divorced (50%), a quarter were separated, 20% were widowed, and those remaining had ended long-term consensual un ions. Differences in educational levels were not remarkable. Half had received vocational training or had attended some college, and 18% were college graduates or had advanced degrees. Black women were more likely to have not completed high school, but they were also more likely to have some college training than were white women. A slight majority of women lived alone (38%) relative to those w ith two to four person house holds, and more black women reported 3-4 person households than white women. Most rated their health as good (43%) or excellent (21%), but more black wome n rated their health as fair (15%) or poor

PAGE 211

199 (10%) than white women (9% and 4% respective ly). A slight majority were unemployed or on SSI (52%), and 43% of them were work ing either partor full-time. Since the primary purpose of the displaced homemaker pr ogram is to assist women with getting their lives back on tr ack after marital disr uption, with a concentration on employment readiness, the low employment rate is not surprising. Several were enrolled in educational programs to further career goa ls, or were planning to enroll soon. Of particular interest for this study was the number of women in new dating relationships after their marital disruption, a nd the large percentage that expressed and interest in dating again. Almo st 39% were in relationships or dating, and 43% expressed an interest in finding a partner. These high pe rcentages suggest that sexual health issues are important topics for health care communication. Physician Sample Most of the 31 physicians were recruite d through a community health and family practice clinic of the universit y-based hospital. Three physicians from other primary care specialties were recruited through an ema il announcement at the area hospital. The physician sample consisted of 13 attending physic ians, and 18 resident physicians. The average age of physicians was 39 years, but attending physicians average age was 48 years and residents average age was 34 year s. Male physicians slightly out numbered female physicians (55% and 45% respectivel y). Women have begun to predominate in the enrollment in family practice residency programs in recent years, and my sample follows this trend with ten women and eight men making up the residency portion of the sample. However, female participation in my study was low compared to the femalemale ratio in the overall residency program, which was just slightly less than 2:1 for

PAGE 212

200 2003/2004 academic year. Just over half of the physicians were white/Euro-American (55%), 23% were Asian, 7% were black/African American and the rest selected other. Of the 31 physician participants, 18 completed the full interview with one of the two vignettes, the demographic/practice char acteristics survey and a CCM procedure. One participated by completing the intervie w without a vignette, and the remaining 13 physicians completed the survey and a CCM pro cedure. In the next section I review the most important findings presented in chapters four, five and six. Synopsis of Findings The primary findings from this project were presented in three separate chapters, each with its own specific focus. This projec t focused on sexuality and aging concepts in and of themselves, as well as on how those co ncepts intertwined with ageand genderbased assumptions in clinical interacti ons. Additionally, through the vignettes and multilayered questioning, social-structural issues were addressed having to do with interracial/interethnic clinical encounters, perceptions of HIV/STD risk and knowledge and other issues. In this ch apter I review the data presen ted in the separate chapters, linking the data from the differe nt chapters together where a ssociations are significant. Cultural Models of Mature Sexuality and HIV The cultural consensus data provided a foundation from wh ich to work with emic concepts about midlife sexuality. The CCM results presented in chapter four uncovered some unanticipated findings. I hypothesized that the DHP women a nd physicians would differ substantially in their concepts of sexuality for midlife and older women. The free list data suggested otherwise, with the two samples sharing more terms and phrases when combined than when they were analyzed separately. The physician free lists had a particularly low frequency of repeated term s, and when their free lists were combined

PAGE 213

201 with those from the women they had more in common with the women than with each other. This is contrary to what I expected to find, which was greater agreement of listed terms from a group of physicians with similar training, and less agreement from women with differing backgrounds. One of the possible explanations for this was the relatively short lists provided by the physicians which gave them less opportunity to find agreement with one another. The lists from the DHP women tended to be longer, and longer lists increase the possibility that more items will be shared. It is also possible that the age range of physicians doing the free list procedure impacted their internal agreement, such that older physicians were more likely to list terms in keeping with the midlife and older women than with younger physicians. This second possi bility was not explored in greater detail, in part, because the pile sort task show ed greater internal agreement among the physicians, and that may mean that the first explanation about free list length is a more reasonable explanation. The multidimensional scaling (MDS) analysis of the pile sort data provided evidence that the two samples were drawing on similar concepts to organize the domain items, with a few very important distinctions The most significant exception related to risks associated with dating a new partne r. Although both samples created groupings along similar dimensions, the inclusion and ex clusion of key items was significant in showing critical differences between how DHP women and physicians constructed the domain. The elements of a new relationshi p for the women included terms associated with the risk ( use protection and more careful ) involved in the dating world today, whereas the physician did not associate terms related to risk avoidance within their

PAGE 214

202 descriptions of later-life relationships for th eir patients. This finding is congruent with the results on notions about HIV risk presen ted in chapter six, showing that DHP women are extremely cognizant of th e risk associated with dating, and that physicians perceptions of HIV/STD risk de clines with patients increasi ng age, female status and intact marriage. Two-thirds of the women thought that they were at-risk for HIV and other STDs, and this was correlated to their current dati ng status. However, their combination of terms like sexy, beautiful and desirable describing idyllic relationships with the terms use protection and more careful reflects the tension between the ideals and realities of dating for these newly single women. The cluster analysis of the pile sort da ta supported the MDS analysis in chapter 4, but provided some additional detail about the nature of the domains for each sample. One point related to the discussion in the paragraph above was that use protection was at a lower level of connectedness than more careful and more mature This indicates that several women did not combine these terms, and this is in keeping with other significant finding in chapter six concerning the disc onnect between knowledge and behavior regarding risk. Analysis of the in-depth interview data showed that although the DHP women were highly cognizant of the risks, many admitted to not always practicing safer sex, and their notions about what other women their age do in dicated the same lack of follow through. This distinction between knowledge and ac tion regarding condom use corresponds to other research with older women (and younger ones, for that matter), pointing to the psychosocial issues in sexual negotiation.

PAGE 215

203 Researchers and prevention specialists may be in danger of becoming overly condom centric in their approach to stemmi ng the tide of HIV infection, as noted scholar, Bryan Page, Ph.D., suggested at th e Society for Applied Anthropology meeting in Santa Fe, NM, April 2005. He encouraged researchers to move beyond an exclusive focus on condom usage and to consider the larg er social forces that contribute to the spread of HIV. Some of those forces, such as labor migration and limited access to health resources in developing countries are not readily applicable he re. But other social and cultural factors are relevant, includi ng gendered power dynamics, and culturally shaped affective relationship expectations. Research on women from different et hnic, socioeconomic and cultural groups illustrates that similar models of romantic relationships influence womens behaviors in sexual negotiation, sometimes resulting in risk denial and higher risk sexual relations (Ragsdale 2002; Sobo 1995; Stansbury and Sier ra 2004). Relationship factors such as duration and perceived seriousness, along w ith the newness of condom use interweave with entrenched gender roles of cultural dom ains or schemas, leaving women with few alternatives that meet their own expectations and those of others about how to negotiate new relationships. Referring back to cultural domain theory as described by Dressler and Bindon (2000) in chapter 3, cultural models are ab stractions of a given domain that help individuals know what to say or do in a give n situation. Yet as ab stractions of ever changing cultural scenarios they are oftentimes necessarily vague and malleable. Thus what may be a very straightforward choi ce for one woman may be a complex social struggle woven with financial or emotional c onsiderations for another. The story from

PAGE 216

204 one woman interviewed for this project is inst ructive on this point. After the death of a loved one, she described the difficulties she faced with depression and low self-esteem, and that despite knowing the risk associated with unprotected sex, she found herself in a renewed relationship with her childrens father that placed her at risk. The long-term relationshipand gender-dynamics of this li aison, and the emotional and financial void that it filled were important issues in her sexual negotiati on. Condom centricity is of limited value when one considers the social complexities and alternative outcomes that are part of such a multifaceted cultural domain. The women in my project appeared to be acutely aware of some of the constraints on practicing safer sex, particularly when they described other women their age. They talked about out-right risk denial, age-based risk denial living for the moment and other emotional responses, self -esteem issues, dislike of c ondoms, newness of negotiating condom use, buying condoms, and so on. Additionally, many talked about HIV testing for themselves and their partners, thus a voiding the issues associated with condom centricity for personal protec tion. The issue of HIV testin g was significant and all the more remarkable that it came up at all since it was not included in the research design. Over a third of the recently single women vol unteered that they had been tested for HIV. Since this was not a specific question that I asked, I do not know if many more had been tested and did not volunteer the informa tion. The rate of testing was significantly correlated with knowing someone with HIV/AI DS. It is not surprising that knowing someone with HIV/AIDS or someone who ha s died of AIDS would motivate testing, however this degree of correlation may signal an important incentive for HIV testing to be considered in other research projects. Ge rmane to the other objectives in this project,

PAGE 217

205 this motivation for HIV testing may be a us eful avenue to open discussion about sexual health issues with certain patients. Issues in the Physician-Patient Dyad Unfortunately, this project confirmed the findings from previous studies regarding physician-initiated discussion of HIV and ot her sexual health topics. Research has repeatedly shown that physicians rarely disc uss sexual health issues, including HIV, and often do so at the patients urging (Makadon and Silin 1995; Gerbert, Maguire, and Coates 1990; Haas and Coe 1997). The vigne ttes were useful for discovering their overall views about physician-patient comm unication on sexual health. Clearly, the vignettes often served to raise topics that I had planned to a ddress later in the interview. Both physicians and DHP women made obs ervations about the patient-provider dynamic in the vignettes that re lated to social, cultural and st ructural issues in clinical communication, several of which support previo us research findings. One such point raised by some of the DHP women was th e power dynamic inherent in the dyad, and relative disadvantage of patients due to being only partially clothed, and in unfamiliar territory where the language and norms of clinical practice sometimes leave them feeling vulnerable. They also felt that time and financial issues we re important determinants in patient care. The DHP women and the physicians talked about how time cons traints interfere with the effectiveness of clinical visits. P hysicians pointed out how time issues affected the number of topics they can talk about with patients, particularly older patients that are presenting with multiple illnesses. Several no ted that sexuality issues are not addressed because they are forgotten in the rush to cove r other pressing health issues. Training that typically focuses physician attention to one or two presenting problems, and a

PAGE 218

206 concomitant emphasis on sexuality in youth w ith the implication that older persons should not be concerned about sexuality (or th e potential health ri sks associated with some sexual behaviors), do not leave older pa tients with much hope that sexual health issues will be addressed in stan dard health care settings. Many of the women focused on the hurried na ture of the clini cal visit, and how physicians often appear disinter ested in them, and always with one foot out the door. Several of the women felt that visiting a woman physician would ameliorate this problem, whereas others thought that women physicians had th e same rushed disposition as men physicians. Consistent with previous research, physic ian and patient stereotypes based on age, and gender were extensive (Cline and McKe nzie 1998; Hall and Roter 2002; Haug 1996; Jenks 1993; Street 2002). Most DHP women preferred women physicians to talk to about sexual health topics, a nd they thought female physicians were more knowledgeable about female health issues, especially wome n physicians close to their age. Several women felt strongly that a patients gender in fluenced how physicians related to them generally as well as regarding sexual health issu es. Physicians appear ed to back this up by noting that either they typically do not c onsider sexual health i ssues relevant (or as relevant) for older women patients, or that they feel that is the pe rspective of many other physicians. Some of the cultural consensus data s upport this contention by revealing the differences in how women conceptualiz e their own sexuality and how physician conceptualize it. For example, the ratings data clearly s howed that the physicians and DHP women had very different cultural unders tandings of mature sexuality for women.

PAGE 219

207 Close inspection revealed that women focused on the more positive aspects of maturing sexuality, whereas physicians focused on more negative or medical aspects. Both rated terms referring to bodily changes and changes in sexual function as relevant to describing midlife female sexuality. However, physicians were much more likely to consider those changes as problematic than were women. Unlike the physicians, the DHP women consistently indicated that terms less active and menopausal problems were not relevant to describing mature sexuality. Rather they emphasized terms like caring, satisfying, sensitive and enjoyable in their choices. The cluster analysis and MDS data furthe r demonstrate how the two samples have very different views about womens mature se xuality in particular areas. These data showed that the physicians developed a coherent grouping of terms for menopausal patients who are having difficulties or discomforts. Whereas there was much less agreement about menopause-related terms fo r the women, suggesting that there may be wide-ranging experiences regard ing menopause. These differen ces also point to the fact that physicians in the course of their work are more often faced with biomedical problems of midlife aging and sexuality, whereas th e women have a much broader range of experiences to draw on, many of which may be very positive. These differences in perceptions are informative as to what physic ians need to understand about midlife and older women to better serve their patients. Despite the clear differences between the samples on seve ral important points, the data also showed that women and physicians share some common expectations about the clinical encounter, mainly in terms of the role patients and physici ans should take in clinical talk about sexual h ealth. Although several women talked about patients being

PAGE 220

208 more pro-active in clinical se ttings, this mainly took the form of seeking second and third opinions, reporting maltreatment due to econom ic or racial discrimination, and seeking health information on their own, rather than confronting the inhere nt asymmetry of the physician-patient relationship. The DHP women put a lot of emphasis on pa tient responsibility in the clinical interaction, however this was assuaged by their ideas about the role of the physician to provide a safe environment for discussing sens itive issues. Physicians largely agreed with this perspective, but noted that th ey were often remiss in providing such an atmosphere simply through the neglect of rais ing sensitive topics so that patients know that it is safe to discuss them. Even though the women and physicians have similar perspectives about the ideal clinical encount er, regrettably, structural and institutional factors appear to hamper the communication process, possibly as often as age-, gender-, SES-, and ethnicity-based assumptions do. The DHP womens perceptions about race and ethnic stereotypes were mostly not in agreement with previous research. Excep t for a fraction of African American women, few had experienced racial stereotyping, t hough several believed that it was still a common occurrence. However, those who had experienced such maltreatment, told devastating stories about events in clinical settings that had left long lasting emotional scars. When the topic of race/et hnicity was brought up, severa l women said they thought the impact of socioeconomic status was more important than ethnicity. Many DHP women felt that educational and economic status significantly affected the way physicians react to patients. Previous studi es have shown that many physicians need a

PAGE 221

209 financial incentive to spend optimal time with patients, thus leaving poorer patients inadequately attended to (Allman et al. 1999; Jackson and George 1998, Lazarus 1988). Some of the DHP women expressed the sentimen t that if one does not have money or the right type of insurance, they can forget getting fair, let alone optimum, treatment. The physicians in this study, even though they tr eat a largely low-inco me population, did not appear very mindful of this issue and this co uld be because most of their patients are poor and therefore they have no frame of compar ison for reflections about differences in treatment. However, more fine-grained anal ysis of the physician data may show greater cognizance of this issue than was presented here. This dissertation touched on several impor tant topics in clinical communication generally, and about sexual he alth in particular. However, not all of the data that collected for this project were fully explore d. Several of the wome n and physicians made comments about physicians listening to their patients, interruptions in discourse, and recommendations for improving pati ent care, but there is more to be investigated on these and other topics within the data collected. Limitations The limitations of this research project ar e discussed on several levels: 1) some of the data collection methods limit the generalizeab lity of the results; 2) some of the project goals were not achieved due sampling issues; 3) not all of the hypot heses were addressed with the interview instrument, and 4) time and space limitations on analysis of all the data gathered. This project yielded a rich dataset despite its limitati ons; so rich th at not all of the data could be analyzed at this time.

PAGE 222

210 Generalizations to Larger Groups The results from this project have several important caveats, particularly in terms of generalizations to larger popul ations. This is a largely ethnographic endeavor, and as such it focuses on a group of people that may or may not be representative of larger groups of people. In order to generalize to populations beyond those that were directly studied requires the use of re presentative samples (Bernard 1995). This study used what Bernard calls purposive or judgment sa mpling, where the researcher decides the purpose that the respondents will serve. Judgment samples are ideal for qualitative research on special populations. In this case, recently single midlife and older women were an ideal such population, so that I could understand how they conceptualize sexuality and HIV/STD risk for themselves and others their age and in similar circumstances. Women in such a situati on have rarely been considered for HIV prevention research, yet their numbers are c ontinuing to grow among those infected with the disease. Although non-random sampling is regularly us ed in qualitative studies (Bernard, 1995; Handwerker, 2001), care is needed when generalizing results to other recently single women and other primary care physic ians. Given that the recently single participants were seeking guidance through a state sponsor ed program, these findings may not be representative of other re cently single women who do not choose to participate in such programs. Although the study participants represented the predominant ethnic groups in the area, and ha d varied backgrounds in terms of education and living arrangements, women from different ethnic groups and in other areas of the state or country may have very differe nt views from the ones presented here.

PAGE 223

211 Handwerker argues that good et hnography blends numerical analysis with textual analysis to look for interconnections betw een respondents, so that ethnographers can interpret the linkages to make inferences that reach beyond the few people they had the chance to learn from (2001). The work of Chavez and colleagues (2001), show that ethnographic results can be successfully generalized to larger groups with similar characteristic, and should not be discounted as isolated or idiosync ratic musings. Thus, while this study may not adequately char acterize the universe of recently single women, I would argue the findings are instructive as to what topics are important for physician training and future prevention efforts directed at this population. Sampling Issues The research plan proposed to select 60 women from the displaced homemaker program over the duration of the project ( 20 African-American, 20 Latino, and 20 EuroAmerican women). Although I anticipated that the Latina sample might be smaller than this based on my communication with the progr am director and the fact that the Latino population in the area is about 5%, I hoped to ha ve a sufficient number to include them in this project. Unfortunately the numbers of Latinas attending the program during the year and a half that I recruited subj ects was lower than usual, and of those attending most were younger than the 45 years of age minimum or they were still married. Recruitment into the displaced homemaker program in this area does not actively targ et Latinos with its limited budget, a fact I became aware of only mid-way through the project. This was disappointing, and leaves a voi d in terms of how this la rger segment of Floridas population would have responded to this research. A related sampling issue with the DHP women was that I was not able to determine refusal rates. My method of recruitment rare ly left me the opportunity to inquire why a

PAGE 224

212 given potential participant chose not to participate. Most often women in the meetings I attended, who did not sign up, were thos e that did not meet the demographic requirements (age, marital status). The program caters to women age 35 and older, whereas my project focused on women age 45 and older. According to conversations with the program director the average age of program participants is 45 and some years closer to 40. Women who met the age crit eria of any given meeting tend to average about a quarter or a thir d of the participants. On the occasions that I was able to ask the older women who di d not sign-up about why they had not signed-up, I was simply told they were not intere sted with no further explanation or that they did not have tim e to participate. The only other refusal information that I gathered was from those th at had signed-up, but that I was never able to meet with. Reasons ranged from movi ng out of the area wrong telephone numbers, but most commonly they did not return my calls. The project also fell short on the number of physicians. The research plan proposed a physician sample of 40 in order to have enough particip ants to complete the CCM procedures. The final sample ended up at 31 due to the difficulty of recruiting physician participants. Physicians are notoriously difficult to recruit for research projects, particularly projects that ta ke more than a few minutes to complete. Several email announcements that went to a hospital-wide primary care listser v, yielded only three additional physicians over th e 28 recruited with the help of the behavioral medicine educator at the family practice clinic. I had to accept lower than ideal numbers of participants to run the CCM analyses. Ho wever, since the reliability from cultural consensus is not just a matter of sample si ze, but also a function of agreement among

PAGE 225

213 respondents (Weller and Romney 1988) the resu lts are defensible. The difficulties I faced with recruiting a physician sample are testamen t to the importance of having the help of an on-staff go-between. Objectives and Interview Instrument Limitations A few of the project objectives were not adequately assessed through the interview instrument. Although I asked women if they t hought that blacks and whites were treated the same in clinical encounters, I neglected to ask this of the physicians. Race/ethnicity focused questions were too distinct betw een the two samples to provide adequate comparisons. Although the topic of physician age came up in many interviews with both women and physicians, mostly in response to the vignettes, and from the open-ended nature of the questioning, I did not specifica lly ask about the impact of physician age to address this objective as I did with physician ethnicity and gender. I was also interested in how these issues relate to womens comfor t level in talking to physicians. Although I managed to get a great deal about physician gender, I was less successful in gathering information about the impact of ethnicity and age on comfort levels. So me of this relates to my expectations regarding the amount of information that would be elicited from the vignettes. Explicit descriptions of these char acteristics for the actors in the stories would have been prudent on this matter. Limitations on Analysis and the Problem of Too Much Data Finally, this project with in-depth interv iews with 62 participants yielded a very large amount of data. There are numerous i ssues that could be further developed with more analysis. Some of these include: more on time issues in clinical talk; comments about physicians listening to their patients ; more on concerns about the impact of financial issues and health insurance in health care; and more on HIV/AIDS related

PAGE 226

214 knowledge such that it could be compared to ot her research more dire ctly. In addition, I asked all of the respondents what they recommended for improving physician-patient communication, and an analysis of this da ta and the data collected from medical educators is needed. There was simply not ti me to review all of this data with the attention to detail that it deserves for this report. Significance and Future Research Considerations This is an important project for im proving health care communication about sexuality issues, as well as clinical HIV/ST D prevention efforts with midlife and older women that are newly single. These research findings have clinical and educational utility. Physicians and other health care prac titioners need practical skills for addressing their mature patients' sexuality concerns, as well as practical solutions for minimizing the hierarchical dynamics of c linical interactions. Sin ce sexuality is simultaneously individual and social, greater knowledge about how it is culturally framed will be useful for implementing culturally competent care. Id eally these results will be used to develop physician-training modules on communicating ab out mature sexuality. These modules will address the real-life concerns of midlife and older women regarding sexuality issues. These data showed that although physicians and recently single mature women share many concepts about mature sexuality, the li ved reality that the women brought to the study showed that there were also significant differences in how mature sexuality is conceptualized. The most significant differen ces highlight the need to steer physicians away from age-, gender-, ethni cityand marita l status-based assumptions about sexuality. Based on the analysis of the data contained in this report there are several concrete suggestions to improve clinical communica tion about mature sexua lity: (a) clinical settings need to display visu al aids (pamphlets, posters, et c.) about sexual ity and sexual

PAGE 227

215 health so that patients will f eel like it is OK to ask thei r doctor about these topics; (b) physicians should use open-ended questions a bout relationships and sexual activity; (c) physicians should not make assumptions about patients sexual ac tivity based on age, marital status, or health status; and (d) phys icians should not assume that patients will bring up sexual health topics w ith their doctors if they have concerns. Time and resource constraints sometimes limit the number of t opics that physicians can address in an encounter, it is therefore important to make ot her resources available to patients, such as: recommending other office staff that can be helpful in addressing patient concerns; and directing literate patients to reading material s on various topics. Fu rther analysis of the data gathered for this project is expect ed to add to this list of suggestions. This projects focus on an understudied sample of divorced, separated, and widowed midlife and older women, provided uniq ue information about an often forgotten segment of the population, particularly with re gards to sexuality and sexual health risk. This is instructive to programs, such as the one used for sample recruitment, that sexual health issues are not only im portant to the women, but also relevant for their continued health and well-being. In a ddition, the findings may also be useful for public health approaches to HIV/STD prevention because some of the views sexuality issues for mature women can be extended beyond the clinic, and into targeted prevention campaigns. The project builds on the body of cultural consensus theory work regarding health beliefs. With further analysis of the HIV/AI DS related knowledge in this data, I can add to and compare to the existing data by Trotter and colleagues (1999), and Weller and

PAGE 228

216 Baer (2001, 2002), which explore intraand intercultural va riation in HIV/AIDS risk, knowledge and health beliefs. Furthermore, future research could use th e ethnographic data from this project to develop a study with a larger, more representa tive sample of matu re recently single women, and a larger, more representative sa mple of physicians. A questionnaire that focuses on a few major findings presented here would lend greater credibility to physician training suggestions. Finally, this project reached beyond the interpersonal dynamics of the physicianpatient dyad to examine the cultural and structural factors impacting clinical communication, situating a micr o-level exploration within macro-level phenomena. Many of the structural and institutional issu es raised here are beyond the scope of this investigation. However, the research reveal s the endurance of long held beliefs about the hierarchical nature of the medical encounter and how it is impacted by gender, age and ethnicity. Future research could seek avenue s of training and instit utional structure that could minimize these impacts, such as including more physic ian education on the problem of stereotyping patients based on age, gender and ethnicity; and improving history taking skills and techni ques to enhance physicians un derstanding of their patients needs and concerns. The incorporation of patient viewpoints and greater patient involvement in the therapeutic process results in higher quality of car e, and better health outcomes.

PAGE 229

217 APPENDIX A SHORT DEMOGRAPHIC SURVEY FOR DHP WOMEN 1. Age ___ Birth date ___/___/19___ 2. Which race or ethnic group do you identify with? ___ White/Euro-American ___ American Indian ___ Black/African-American ___ Asian/Pacific Islander ___ Other (write-in) _______________ 3. Are you Spanish, Latino, or Hispanic ___ yes: please specify heritage _______________________________________ ___ no 4. What is the highest level of education you have completed? ___ 8th grade or less ___ Trade School ___ Some High School ___ Some College ___ High School Graduate ___ College Graduate ___ GED ___ Advanced Graduate training 5. Marital Status/Partner Status: ___ Divorced (months/years married ____) ___ Widowed (months/years married ____) ___ Separated (months/years married ____) ___ Living together with a boyfrie nd/girlfriend (months/years ____) 6. Do you currently have a dating pa rtner (boyfriend/girlfriend)? Yes __ No__, or dating w/out commitment___ 7. Do you have any children? Yes___ No___, If yes, how many still live with you ___ 8. How many persons live in your household ______ 9. Religious/Spiritual Affiliation: ___________________________________________ 10. Employment status ___ working full-time (40 hours or more) ___ retired ___ working part-time (30 hours or less) ___ unemployed ___ other (specify ____________________) ___ disability income (SSI) 11. How long have you lived in the J acksonville area (north FL/south GA) ___ entire life ___ 20 yrs or more ___10-15 yrs ___5-10 yrs ___ <5 yrs

PAGE 230

218 12. Please rate your curre nt health status. ___ excellent ___ good ___ average ___ fair ___ poor 13. When did you last visit a doctor? ___ within the last month ___ within the last 2 years ___ within the last six months ___ have not seen a doctor in over 2 years ___ within the last year 14. Type of doctor you visited ______________________________________________ 15. Do you have any chronic health conditi ons that require periodic or continuous medical care (such as heart problems, high bl ood pressure, diabetes, arthritis, etc.)? ___ yes ___ no ___ not sure 16. Do you consider yourself disabled? ___ yes, totally ___ yes, partially ___ no

PAGE 231

219 APPENDIX B PHYSICIAN SURVEY Circle one: attending resident ( Circle appropriate year : 1, 2 or 3 ) 1. Age ___ Birth date ___/___/19___ 2. Which race or ethnic group do you identify with? ___ White/Euro-American ___ American Indian ___ Black/African-American ___ Asian ___ Other (write-in) ____________________________________________________ 3. Are you Spanish, Latino, or Hispanic? ___ yes: please specify heritage ___________________________________________ ___ no 4. Do you speak any languages ot her than English, fluently? ___ yes: please list _____________________________________________________ ___ no 5. Do you have bilingual (trilingual, etc.) st aff in your office (clinic)? ___ yes ___ no 6. What year did you comp lete medical school _______ 7. Please provide a rough estimate of the percen tage of persons you treat annually in each age range: ____ under age 10 ____ age 35-44 ____ age 10-19 ____ age 45-59 ____ age 20-34 ____ age 60-74 ____ age 75+ 8. What percentage (estimate) of your patients are women? ___ less than 10 % ___ 50-70% ___ 10-30% ___ more than 70% ___ 30-50% 9. What percentage (estimate) of your patients are: ____ White/Euro-American ____ American Indian ____ Black/African-American ____ Asian ____ Other (write-in) __________________________ 10. What percentage (estimate) of your patients are Hispanic/Latino? ___ less than 10 % ___ 50-70%

PAGE 232

220 ___ 10-30% ___ more than 70% ___ 30-50% 11. How often do time constraints affect your ab ility to take a thorough history with your patients? (circle one) Always Almost always Half the time Almost never Never Comments ____________________________________________________________________ ____________________________________________________________________ ____________________________________________________________________ 12. Do you feel that your medical educati on adequately prepared you for addressing current sexuality issues (including STDs /HIV) with a wide range of patients? Strongly Agree Agree Neutra l Disagree Strongly Disagree Comments ____________________________________________________________________ ____________________________________________________________________ ____________________________________________________________________ 13. Where do you access updated information on aging and sexual health? ________ ____________________________________________________________________ ____________________________________________________________________ 14. How well informed are the ma jority of your adult patients (age 45 and older) about HIV and other STDs? (circle one) Well-informed --1 2 3 4 5 --not at all informed 15. Generally, what percentage of your adul t patients do you consider at-risk for HIV or other STDs? (Note: This question was not includ ed until the survey was revised and was only administered to half of the physician participants) ____% males ____% females By age group: (Note: This question was not included until the survey was revised and was only administered to half of the physician participants) ____% males age 18-30 ____% females age 18-30 ____% males age 31-45 ____% females age 31-45 ____% males age 46-64 ____% females age 46-64

PAGE 233

221 ____% males age 65+ ____% females age 65+ Comments _____________________________________________________________________ _____________________________________________________________________ _____________________________________________________________________ 16. Do you think it is primarily the patient's or the physician's responsibility to bring up sex life issues in a clinical encounter? ___ patient ___ physician ___both ___ neither ___dont know Please explain WHY you made th e choice you did for question 16. ________________ ______________________________________________________________________ ______________________________________________________________________ (Note: This question was not included until the survey was revised and was only administered to half of the physician participants) Other comments: ______________________________________________________________________ ______________________________________________________________________ ______________________________________________________________________

PAGE 234

222 APPENDIX C VIGNETTE INTERVIEW FOR DHP WOMEN Vignettes and Interview Questions I'm going to read you [or play a recording of] two stories about docto r-patient interaction. The people in the stories are not real people, but you may find their situation to be familiar. After I read [play] each story, I'll ask you some questions on your opinions about their situation. There are no right or wrong answers. Your responses to my questions do not need to be based on your personal experience, but instead may just reflect your general knowledge and opinions about doctor-pa tient communication. Please do not hesitate to ask me to repeat any of the questions or the stories. Story one (read aloud) Cynthia Cooper, has back pain because of growths (fibroids) in her uterus, but at age 64 she is in good health otherwise. When Ms. Cooper's GYN diagnosed her with fi broids 6 months ago, he suggested that she may need to have a hysterectomy to remove them to relieve the discomfort and to avoid future cancer risk. No w that the pain is getting worse Ms. Cooper has made an appointment to visit with her doctor to talk about the Hysterectomy. She recently read a magazine article about a woman who experienced sexual problems after a Hysterectomy and is worried, but she f eels uncomfortable bringing up this issue to her doctor. Her doctor suggests removing her uterus for the fibr oids, as well as her cervix and ovaries to prevent future cancers, and si nce she is well past menopause she does not need these organs anymore. She asks, "Are there any complications I should be concerned about?" The doctor explains, It is major surgery, but it is also very common. There is the remote possibility that you could have post-surgical difficulti es with incontinence and other urogenital problems, but the risk is very low." He conti nues, "It does have the risks associated with any major surgery and then spends anothe r 15 minutes explaining the procedure and all of the possible risks to Mrs. Cooper, in detail. Afterwards she still has concer ns and asks, "But won't the re moval of these organs (affect my sex driveormake me less feminineor affect my sex-life)?" "No," the doctor responds, "though some women may have difficu lties because they believe that they will." Ms. Cooper is reassured, and feels that the doctor knows best about her health and cancer risks.

PAGE 235

223 She has the surgery and is out of the hospita l in a few days, and is told she should be "back to normal in about 6 weeks." On he r one-month follow-up visit the doctor says she is healing just fine, and on the two-month follow-up he says she has healed, and can resume her life as normal. When she tries to have sex for the first time after the surgery, it is extremely painful and she has some bleed ing. She calls the doctor's office and makes an appointment. When she explains what ha ppened during sex the doctor says, "Oh, that's not unusual. You just need to have sex more often and it will get better." Ms. Cooper is very troubled as she remembers the pain and wonders if she will ever be able to enjoy making love again. What is your opinion about this doctor-patient communication? Probes: Was it typical, unusual, normal, good, bad, positive or negative? How do you think this doctor-patie nt exchange could be improved? Do you think doctors are usually sensitive to how medical treatments (or drug treatments) may affect a patient's sex life? Does this story remind you of any of your e xperiences, or perhaps experiences of your friends that you have been told about? Story two When Wanda Johnson began dating in the mid1990s, she was a little rusty, having gone through a divorce in 1992 after 31 years of ma rriage to an unfaithful husband. She did not frequent singles ba rs. She dated a few men she met th rough friends and at church. But Carl Jenkins, a long time friend from her c hurch, was different. He was handsome, funny and someone she thought she could get seri ous with. They began spending more and more time together; Wanda felt she had found love like she had never known before, and after a year they were married in the church where they met. A year later, Wanda Jenkins began feeling very sick. She had had an annoying cough for weeks, and was now running a fever and felt ach y all over. She went to see the primary care doctor, Dr. Lewis, on her health insura nce plan. After taking her temperature and discussing her symptoms, Dr. Lewis said, "You appear to have a b acterial infection, but antibiotics should clear it up in a week or two." Mrs. Jenkins did improve over the next few weeks, but began feeling poorly again soon after. She began having yeast infections, which were not clearing up with the over-thecounter medicines. Dr. Lewis prescribed some thing stronger (fluconazole). He explained that the reason she was having difficulty curi ng the infection was because of her earlier dose of antibiotics. She improved, only to get a urinary track inf ection a few weeks later. Carl was worried and the constant medical bills and Wanda's poor health was putting a strain on both of them. Over a 10-month period, Wanda was pr escribed various an tibiotics and antifungals. Although Mrs. Jenkins' chronic difficu lties were typical of immune disorders

PAGE 236

224 like HIV/AIDS, Dr. Lewis knew that his pa tient was a married, church-going woman, and he did not think her problems were sexually transmitted. Dr. Lewis attributed Mrs. Jenkins' fevers and night sweats to menopause and to a rough bout with chronic fungal and b acterial infections. He put her on hormone therapy, but there was no improvement. She was very fatigued and ached all over, and Dr. Lewis suggested, "You may have chr onic fatigue syndrome, Lupus, or Rheumatoid Arthritis, and we can start running some tests for thes e." After a year of ch ronic infections and general malaise, Mrs. Jenkins was rushed to the hospital with pneumonia, where she was finally tested for HIV/AIDS. No one ever thought she was at risk. What is your opinion about this doctor-patient communication? Probes: Was it typical, unusual, normal, good, bad, positive or negative? How do you think this doctor-patie nt exchange could be improved? Do you think doctors ought to address the topi c sexually transmitted in fections with their mature patients? Probe: When? Why? Does this story remind you of any of your e xperiences, or perhaps experiences of your friends that you have been told about? Do you think that other recently single women your age and older think that they are atrisk for HIV or other STDs? Probe: Why? Why not? What about doctors? Do you thi nk that they think of women your age and older as being at-risk for HIV or other STDs? Probe: Why? Why not? Tell me a little bit about wh at you know about AIDS/HIV? Pe rhaps things that you have picked-up from TV, ma gazines, etc. Do you think that you are at-risk for HIV or other STDs? Probe: Why? Why not? General questions for your opinions ab out patient-doctor communications: What are the most common communication issu es that women face when they talk to their doctors? What makes it easier to discuss important sexuality (or sex life) issues with doctors? What makes it more difficult to discuss impor tant sexuality (or sex life) issues with doctors? Do you think that male doctors and female doc tor talk with their mature women patients in the same way? Probe: What about when they discuss sexuality issues?

PAGE 237

225 Do you think that the ethnic background (race) of a doctor affect how they talk with mature women patients? Do you think that the ethnic background (race) of a patient affect how the doctor talks with mature women patients? Do you think that it is the patient' or the docto r's responsibility to bring up sex life issues? How do you think sexuality issues should be addressed between doctor and patient? Do you think that doctors should assume thei r mature patients are sexually active? Probes: Even if they are divorced or widowed? How do you define "sex lif e" or "sexuality" for women aged 45 and older? (use only for pile-sort and ratings test participants) Has your definition of sexuality changed over time?

PAGE 238

226 APPENDIX D VIGNETTES AND INTERVIEW FOR PHYSICIANS Vignette and Interview Questions for First Group of Physicians Story (read aloud): When Wanda Johnson began dating in the mid-1990s, she was a little rusty, having gone through a divorce in 1992 af ter 31 years of marriage to an unfaithful husband. She did not frequent singles bars. She dated a few men she met through friends and at church. But Carl Jenkins, a long tim e friend from her church, was different. He was handsome, funny and someone she thought she could get serious with. They began spending more and more time together; Wanda felt she had found love like she had never known before, and after a year they were ma rried in the church where they met. A year later, Wanda Jenkins began feeling very sick. She had had an annoying cough for weeks, and was now running a fever and felt ach y all over. She went to see the primary care doctor, Dr. Lewis, on her health insura nce plan. After taking her temperature and discussing her symptoms, Dr. Lewis said, "You appear to have a b acterial infection, but antibiotics should clear it up in a week or two." Mrs. Jenkins did improve over the next few weeks, but began feeling poorly again soon after. She began having yeast infections, which were not clearing up with the over-thecounter medicines. Dr. Lewis prescribed some thing stronger (fluconazole). He explained that the reason she was having difficulty curi ng the infection was because of her earlier dose of antibiotics. She improved, only to get a urin ary track infection a few weeks later. Carl was worried and the constant medical bills and Wanda's poor health was putting a strain on both of them. Over a 10-month period, Wanda was pr escribed various an tibiotics and antifungals. Although Mrs. Jenkins' chronic difficu lties were typical of immune disorders like HIV/AIDS, Dr. Lewis knew that his pa tient was a married, church-going woman, and he did not think her problems were sexually transmitted. Dr. Lewis attributed Mrs. Jenkins' fevers and night sweats to menopause and to a rough bout with chronic fungal and b acterial infections. He put her on hormone therapy, but there was no improvement. After a year of ch ronic infections and ge neral malaise, Mrs. Jenkins was rushed to the hospital with pne umonia, where she was finally tested for HIV/AIDS. No one ever thought she was at risk. Interview Questions: What is you opinion about this physician-patient exchange? Probes: Was it typical, unusual, normal, good, bad, positive or negative?

PAGE 239

227 How do you think this physician-pati ent exchange could be improved? When do you think sexuality issues should be addressed between physician and patient? Should physicians assume that all of th eir adult patients ar e sexually active? Probes: Even if they are divorced or widowed? Older Patients How should physicians address the topic sexually transmitted diseases with their mature patients? Do you think the sex (male or female) of a patient affect how physicians communicate with them? Probes: Does this affect how they discuss sexuality issues? Do you think the ethnic background (race) of a patient affects how physicians communicate with them? Does the age of a patient affect how the physician communicates with them? General Interview Questions: What are the most difficult communication issues that physicians face when they talk to their mature female patients about sensitive topics? What makes it easier to discuss important se xuality (or sex life) issues with patients? What makes it more difficult to discuss impor tant sexuality (or sex life) issues with patients? ** (only for those not doing free-list) How do you define "sex life" or "sexuality" for women as they age? Do you think it the patient's or the physician' s responsibility to bring up sex life issues? What percentage of your patient do you c onsider at-risk for HIV or other STDs? Probes: Why so many? Why so few? Do es their age/ethnicity/sex/sexual orientation influence risk status? How? Why? Where do you get updated ethnic sensitive healthcare information? Probe: How often? What are your suggestions for improvements in clinical communication with mature female patients?

PAGE 240

228 Vignette and Interview Questions for Second Group of Physicians Story (read aloud): Cynthia Cooper, has back pain because of fi broids in her uterus, but at age 64 she is in good health otherwise. When Ms. Cooper's GYN diagnosed her with fi broids 6 months ago, he suggested that she may need to have a hysterectomy to remove them to relieve the discomfort and to avoid future cancer risk. No w that the pain is getting worse Ms. Cooper has made an appointment to visit with her doctor to talk about the Hysterectomy. She recently read a magazine article about a woman who experienced sexual problems after a Hysterectomy and is worried, but she f eels uncomfortable bringing up this issue to her doctor. Her doctor suggests removing her uterus for the fibr oids, as well as her cervix and ovaries to prevent future cancers, and si nce she is well past menopause she does not need these organs anymore. She asks, "Are there any complications I should be concerned about?" The doctor explains, It is major surgery, but it is also very common. There is the remote possibility that you could have post-surgical difficulti es with incontinence and other urogenital problems, but the risk is very low." He conti nues, "It does have the risks associated with any major surgery and then spends anothe r 10 minutes explaining the procedure and all of the possible risks to Mrs. Cooper, in detail. Afterwards she still has concer ns and asks, "But won't the re moval of these organs (affect my sex driveormake me less feminineor affect my sex-life)?" "No," the doctor responds, "though some women may have difficu lties because they believe that they will." Ms. Cooper is reassured, and feels that the doctor knows best about her health and cancer risks. She has the surgery and is out of the hospita l in a few days, and is told she should be "back to normal in about 6 weeks." On he r one-month follow-up visit the doctor says she is healing just fine, and on the two-month follow-up he says she has healed, and can resume her life as normal. When she tries to have sex for the first time after the surgery, it is extremely painful and she has some bleedi ng. She calls the doctor's office and makes an appointment. When she explains what ha ppened during sex the doctor says, "Oh, that's not unusual. You just need to have sex more often and it will get better." Ms. Cooper is very troubled as she remembers the pain and wonders if she will ever be able to enjoy making love again. Interview Questions: What is you opinion about this physician-patient exchange? Probes: Was it typical, unusual, normal, good, bad, positive or negative? How do you think this physician-pati ent exchange could be improved?

PAGE 241

229 Do you think physicians are usually sensit ive to how medical treatments (or drug treatments) may affect a patient's sex life? Do you think the sex (male or female) of a patient affect how physicians communicate with them? Probes: Does this affect how they discuss sexuality issues? Do you think the ethnic background (race) of a patient affects how physicians communicate with their patients? Does the age of a patient affect how the physician communicates with them? When do you think sexuality issues should be addressed between physician and patient? Do you think physicians should assume that all of their adult patients are sexually active? Probes: Even if they are divorced or widowed? Ol der patients? How should physicians address the topic sexually transmitted diseases with their mature patients? General Interview Questions: What are the most difficult communication issues that physicians face when they talk to their mid-life and older female patients about sensitive topics? What makes it easier (and more difficult) to discuss important sexuality (or sex life) issues with patients? ** (only for those not doing free-list procedure) How do you define "sex life" or "sexuality" for women as they age? Probe: Does this definition change over time (marital/partner status)? Do you think it is the patient's or the physician' s responsibility to bring up sex life issues? What percentage of your patient do you c onsider at-risk for HIV or other STDs? Probe: Why so many? Why so few? Does their age/ethnicity/sex/sexual orientation influence risk status? How? Why? Where do you get updated ethnic sensitive healthcare information? Probe: How often? What are your suggestions for improvements in clinical communication with mature female patients?

PAGE 242

230 LIST OF REFERENCES Adelman, R. D., M. G. Greene, and R. Charon 1991 Issues in Physician-Elderly Patien t Interaction. Agei ng and Society 11:127148. Adler, S. 1998 Patient Assertiveness in Ethnically Diverse Older Women With Breast Cancer: Challenging Stereotypes. Jo urnal of Aging Studies 12(4):331-342. American Association of Retired Persons (AARP) 1999 Sexuality Study, AARP/Modern Matu rity, August. Washington DC: AARP. 2003 Lifestyles, Dating & Romance: A St udy of Midlife Singles. September, Washington DC: AARP. Association of Reproductive Hea lth Professionals (ARHP) 2000 ARHP Clinical Proceedings. Mature Sexuality: Patient Realities and Provider Challenges. Washington, DC: ARHP. Reprints available: Association of Reproductive Health Professionals, 2401 Pennsylvania Avenue, NW, Suite 350, Washington, DC 20037-1718. Baer, H. A., M. Singer, and I. Susser 1997 Medical Anthropology and the World System: A Critical Perspective. Westport, CT: Bergin and Garvey. Bernard, R. H. 1995 Research Methods in Anthropology: Qu alitative and Quantitative Approaches. Walnut Creek, CA: AltaMira. Binson, D., L. Pollack, and J. Catania 1997 AIDS-Related Risk Behaviors and Sa fer Sex Practices of Women in Midlife and Older in the United States: 1990 to 1992. Health Care for Women International 18:343-354. Borgatti, S. P. 1994 How to Explain Hierarchical Clustering, Connections 17(2): 78-80. 1996a ANTHROPAC 4.0. Natick, MA: Analytic Technologies. 1996b ANTHROPAC 4.0 Methods Guide. Natick, MA: Analytic Technologies

PAGE 243

231 1999. Elicitation Techniques for Cultural Domain Analysis. In Enhanced Ethnographic Methods, Vol. 3, Ethnographer's Toolkit. J. J. Schensul, M. D. LeCompte, B.K. Nastasi, and S. Borgat ti, eds. Pp.115-151. Walnut Creek, CA: AltaMira. Cant, S., and U. Sharma 2000 Alternative Health Practices and Systems. In Handbook of Social Studies in Health and Medicine. G.L. Albrecht, R. Fitzpatrick, and S.C. Scrimshaw, eds. Pp. 426-439. London: Sage Publications. Centers for Disease Control and Prevention (CDC) 2001a HIV/AIDS Surveillance Supplemen tal Report; 7 (No. 1). Washington, DC. Electronic document, http://www.cdc .gov/hiv/stats/hasrsupp71.htm, accessed May 2002. 2001b HIV/AIDS Surveillance volumn13 (No.1). Washington, DC. Electronic document, http://www.cdc.gov/hiv/stat s/hasrlink.htm, accessed June 2002. 2002 HIV/AIDS Surveillance Report 200 2; Number 14. Electronic document, http://www.cdc.gov/hiv/stats/ha sr1402.htm, accessed November 2004. 2003 Cases of HIV infection and AIDS in the United States, by race/ethnicity, 1998. HIV/AIDS Surveillance Supplemen tal Report;10 (No.1). Electronic document, http://www.cdc.gov/hiv/st ats/hasrsuppVol10No1.htm, accessed November 2004. Chavez, L. R., F. A. Hubbell, J. M. McMu llin, R. G. Martinez, and S. I. Mishra 1995 Structure and Meaning in Models of Br east and Cervical Cancer Risk Factors: A Comparison of Perceptions among La tinas, Anglo Women and Physicians. Medical Anthropology Quarterly 9:40-74. Chavez, L. R., J. M. McMullin, R. G. Martinez, S. I. Mishra, and F. A. Hubbell. 2001 Beliefs Matter: Cultural Beliefs and the Use of Ce rvical Cancer-Screening Tests. American Anthropologist 103 (4): 1114-1129. Chiang, L. 1998 The Geriatrics Imperative: Meeting the Need for Physicians Trained in Geriatric Medicine. Journal of the Am erican Medical Association, Pulse-Report 279:1036-1037. Clark, P. G. 1996 Communication Between Provider and Patient: Values, Biography, and Empowerment in Clinical Practice. Ageing and Society 16(6):747-773.

PAGE 244

232 Cline, R. J. and N. McKenzie 1998 The Many Cultures of Health Care: Di fference, Dominance, and Distance in Physician-Patient Communication. In Health communication research: A guide to developments and direction. L.D. Jackson and B.K. Duffy, eds. Pp. 57-74. Westport, CN: Greenwood Press. Coates, T. J. 1999 HIV Prevention in Clinical Practic e. AIDS Clinical Care 11(1):1-4. Cooper, L. A., D. L. Roter, R. L. Johnson, D. E. Ford, D. M. Steinwachs, and N. R. Powe 2003 Patient-Centered Communication, Rati ngs of Care And Concordance of Patient And Physician Race. Annals of Internal Medicine 139:907-915. Cooper-Patrick, L., J. J. Gallo, J. J.Gonzales, H. T. Ve, N. R. Powe, C. Nelson, and D. E. Ford 1999 Race, Gender, and Partnership in the Physician-Patient Relationship. Journal of the American Medical Association 282(6):583-589. Coulter, A., and R. Fitzpatrick 2000 The Patient's Perspective Rega rding Appropriate Health Care. In Handbook of Social Studies in Health and Medicine. G.L. Albrecht, R. Fitzpatrick, and S.C. Scrimshaw, eds. Pp. 454-464. London: Sage Publications. Coupland, N., and J. Coupland 1999 Ageing, Ageism and Anti-Ageism: Moral Stance in Geriatric Medical Discourse. In Language and Communication in Old Age, Multidisciplinary Perspectives. Heidi E. Hamilton, ed. Pp. 177-208. New York: Garland Publishing. Damron-Rodriguez, J. A., 1998 Respecting Ethnic Elders: A Perspec tive For Care Providers. Journal of Gerontological Social Work 29(2-3):53-71. Dancy, B. 1996 What African American Women Know, Do, and Feel About AIDS: A Function of Age and Education. AIDS Education and Prevention 8:26-36. Davidhizar, R., and J. Giger-Newman 1996 Reflections On Minority Elderly In Health Care. Hospital Topics 74(3):20-24. Dressler, W. W. 1996 Culture and Blood Pressure: Using Consensus Analysis to Create a Measurement. Cultural Anthropology Methods 8:6-8.

PAGE 245

233 Dressler, W. W., and J. R. Bindon 2000 The Health Consequences of Cultura l Consonance: Cultural Dimensions of Lifestyle, Social Support, and Arterial Blood Pressure in an African American Community. American An thropologist 102(2):244-260. Dustow, C. 1995 Sex, Sexuality, Safer Sex and HIV. AIDS Clinical Care 7(7):53-59. Elgin, S. H. 1999 Language In Emergency Medicine: A Verbal Self-Defense Handbook. Philadelphia, PA: Xlibris Corporation. Epstein, R. M., D. S. Morse, R. M. Frankel, L. Frarey, K. Anderson, and H. B. Beckman 1998 Awkward Moments in Patient-Physic ian Communication about HIV Risk. Annals of Internal Medicine 128: 435-442. Fishbein, M., and M. Guinan 1996 Behavioral science and public he alth: a necessary partnership for HIV prevention. Public Health Reports 111(SU PPL. Special Issue: Behavioral Science in HIV Prevention):5-11. Fisher, S. 1993 Doctor talk/patient talk: How decisi ons are negotiated in doctor-patient communication. In A. D. Todd and S. Fish er (Eds.) The Social Organization of Doctor-Patient Communication, pp. 161-182. Norwood, NJ: Ablex. Fields, J., and Casper, L. M. 2001 Americas Families and Living Arrangements: March 2000. Current Population Reports, P20-537. U.S. Census Bureau, Washington, DC. Florida Department of Health 2003a HIV/AIDS Among Persons Age 50 and Ov er, Florida. Florida Department of Health, Bureau of HIV/AIDS. Electronic document, http://www.doh.state.fl.us/disease_ctrl/ aids/updates/facts/2003SeniorsFS.pdf, accessed May 2004. 2003b United States and Florida HIV/AIDS Fact Sheet. Florida Department of Health, Bureau of HIV/AIDS. Electronic document, http://www.doh.state.fl.us/disease_ctrl/ aids/updates/facts/2003USvsFLFS.pdf, accessed May 2004. 2003c HIV/AIDS among women fact sheet, Florida 2003. Florida Department of Health, Bureau of HIV/AIDS. Electronic document, http://www.doh.state.fl.us/disease_ctrl/ aids/updates/facts/2003WomenFS.pdf, accessed May 2004.

PAGE 246

234 2003d Florida Behavioral Risk F actor Data. Electronic document http://www.floridacharts.com/cha rts/brfss.aspx, accessed May 2004. Garro, L. C. 1986 Intracultural Variation in Folk Medical Knowledge: A Comparison Between Curers and Noncurers. Amer ican Anthropologist 88:351-370. Garro, L. C. 2000 Remembering What One Knows and the Construction of the Past: A Comparison of Cultural Consensus Theory and Cultural Schema Theory. Ethos 28(3): 275-319. Gerbert, B., B. T. Maguire, and T. J. Coates 1990 Are Patients Talking To Their P hysicians About AIDS? The American Journal of Public Hea lth; April 80(4):467-470. Gibble, J. N., S. M. Rogers, H. G. Miller, and C. F. Turner 1998 Measuring AIDS-Related Behaviors in Older Populations. Research on Aging 20 (6): 798-821. Good, M.D., and B. J. Good 2000 Clinical Narratives and the Study of the Contemporary Doctor-Patient Relationship. In Handbook of Social Studies in Health and Medicine. G.L. Albrecht, R. Fitzpatrick, and S.C. Sc rimshaw, eds. Pp. 243-258. London: Sage Publications. Greene, M. G., R. D. Adelman, E. Friedmann, and R. Charon. 1994 Older Patient Satisfaction with Co mmunication During an Initial Medical Encounter. Social Scien ce and Medicine 38(9):1279-1288. Haas, D. M., and R. M. Coe 1997 Do Physicians discuss HIV and AIDS with patients? A Survey of Physician Practices. Journal of Community Health, Dec.; 22(6): 401-16. Hall J. A., and D. L. Roter 2002 Do patients talk differently to male and female physicians? A meta-analytic review. Patient Educatio n and Counseling 48: 217 Handwerker, W. P. 2001 Quick Ethnography. Walnut Creek, CA: AltaMira. 2002 The Construct Validity of Cultures: Cultural Diversity, Culture Theory, and a Method for Ethnography. Americ an Anthropologist 104(1):106-122.

PAGE 247

235 Handwerker, W. P., and D. F.Wozniak 1997 Sampling Strategies for the Collection of Cultural Data: An Extension of Boas's Answer to Galton's Problem Current Anthr opology 38(5):869-875. Haug, M.R. 1996 Elements in Physician/Patient Intera ctions in Later Life. Research on Aging 18(1):32-51. Henley, N. 1977 The other dimension. In Body Politics. Pp. 1-26. Englewood Cliffs, NJ: Prentice Hall. Herzog, A. R., and W. L. Rogers 1988 Interviewing Older Adults: Mode comparison using data from a face-to-face and telephone survey. Pub lic Opinion Quarterly 54:84-99 Hotvedt, M. 1983 Sexuality in the Later Ye ars: The cross-cultural and historical context. In R. B. Weg (Ed.) Sexuality in the Later Y ears: Roles and Behavior, pp. 13-37. New York: Academic Press. Jackson, P. B., and L. K. George 1998 Racial differences in satisfaction with physicians: A study of older adults. Research on Aging 20(3):298-316. Jacoby, S. 1999 Great Sex: What's Age Got to Do With It? AARP/Modern Maturity, Sept/Oct. Washington DC: AARP. Janus, S. S., and C. L. Janus 1993 The Janus Report on Sexual Behavior. New York: John Wiley & Sons. Jenks, E. B. 1993 Communication in health care settings. In Women and Men Communicating. L. P. Arlis and D.J. Borisoff, eds. Pp. 173-192. Fort Worth, TX: Harcourt. Kerr, S. H., R. O. Valdiserri, J. Loft, L. Bresolin, D. Holtgrave, M. Moore, R. MacGowan, W. Marder, and R. Rinaldi 1996 Primary care physicians and their HI V prevention practices. AIDS Patient Care, Aug; 10 (4): 222-35. Kreps, G. L. 1990 A systematic analysis of hea lth communication w ith the aged. In Communication, Health and the Elderly. H. Giles, N. Coupland and J.M. Wiedmann, eds. Pp. 135-154. Manchester England: Manchester University Press.

PAGE 248

236 Lazarus, E. 1988 Theoretical Considerations for the Study of Doctor-Patient Relationship: Implications of a Perinatal Study. Medical Anthropology Quarterly 2:34-58. Levy, J. A. 1994 Sexuality and Aging. In Principles of Geriatric Medicine and Gerontology, Third Edition. W. R. Hazzard, W. R. H azzard, Edwin L. Bierman, John P. Blass, Walter H. Ettinger, and Jeffery B. Halter, eds. Pp. 115-122. New York: McGraw-Hill, Inc. Levy, J. A., and G. L. Albrecht 1989 Methodological Considerations in Research on Sexual Behavior and AIDS Among Older People. In AIDS in an Aging Society: What we need to know. M.W. Riley, M.G. Ory, and D. Zaboltsky, eds. Pp.96-123. New York: Springer Publishing Company, Inc. Linsk, N. L., J. P. Fowler, and S. Klein 2003 HIV/AIDS Prevention and Care Services and Services for the Aging: Bridging the Gap Between Service Systems to Assi st Older People. Journal of Acquired Immune Deficiency Syndromes 33:S243-S250. Loehr, J, S. Verma, and R. Seguin 1997. Issues of sexuality in older women. Journal of Womens Health 6(4):451-7 Lorber, J. 1997 Gender and the Social Construction of Illness. Thousand Oaks, CA: Sage Publications, Inc. Mack, K. A. and S. D. Bland 1999 HIV Testing Behaviors and Attitudes Regarding HIV/AIDS of Adults Aged 50-64. The Gerontologist 39(6): 687-694. Makadon, H. J., and J. G. Silin 1995 Prevention of HIV Infection in Prim ary Care: Current Practices, Future Possibilities. Annals of In ternal Medicine 123(9):715-719. Marwick, C. 1999 Survey says patients expect little physician help on sex. Journal of the American Medical Association 281: 21732174. Metz, M.E., and M. H. Seifert, Jr. 1990 Men's expectations of physicians in sexual health concerns. Journal of Sex and Marital Therapy 16(2):79-88.

PAGE 249

237 Miller, M, K. Mayer, and H. Makadon 1999 Talking About Safer Sex With Your Patients. Boston, MA: The Fenway Institute/Fenway Community Health. National Council on Aging (NCOA) 1998 Healthy Sexuality and Vita l Aging. Washi ngton DC: NCOA. Nusbaum, M.R.H., G. Gamble, B. Skinner, and J. Heiman 2000 The High Prevalence of Sexual C oncerns Among Women Seeking Routine Gynecological Care. Journal of Family Practice 49(3):229-232. Nusbaum, M.R.H., G. R.Gamble, and D. E. Pathman 2002 Seeking Medical Help for Sexual Concerns: Frequency, Barriers, and Missed Opportunities. The Journal of Family Practice 51 (8):706-710 Nussbaum, J. F., T. Thompson, and J.D. Robinson 1989 Death and Dying. In Communication and Aging. Pp. 214-252. New York, NY: Harper and Row. Ory, M. G., and K. A. Mack 1998 Middle-Aged and Older People with AIDS: Trends in National Surveillance Rates, Transmission Routes, and Risk Factors. Research on Aging 20(6):653664. Ory, M.G., D.L. Zaboltsky, and S. Crystal 1998 HIV/AIDS and Aging: Identifying a prevention research and care agenda [Guest Editorial]. Research on Aging 20(6):637-652. Paget, M. A. 1993 On the Work of Talk: St udies in Misunderstandings. In The Social Organization of Doctor-Patient Communica tion (second edition). A. D. Todd and S. Fisher, eds. Pp.107-126. Norwood, NJ: Ablex. Penson, R. T., J. Gallagher, M. E. Gioiella, M. Wallace, K. Borden, L. A.Duska, J A. Talcott, F. J. McGovern, L J. Appleman, B. A. Chabner, B. A., and T. J. Lynch, Jr. 2000 Sexuality and Cancer: Conversation Co mfort Zone. The Oncologist 5(4):33644. Poppen P. J., and C. A. Reisen, 1997 Perception of Risk and Sexual Self -Protective Behavior : A Methodological Critique. AIDS Education and Prevention 9 (4):373-390. Ragsdale, A. K. 2002 Tourist Women Balancing Temptation and HIV-risks in Costa Rica. Ph.D. dissertation, Department of Anthr opology, University of Florida.

PAGE 250

238 Rich, E. R. 2001 Negotiation of HIV Preventive Behavi ors in Divorced and Separated Women Reentering the Sexual Arena. Journal of the Association of Nurses in AIDS Care 12(4):25-35. Romney, A. K., S. C.Weller, and W. H. Batchelder. 1986 Culture as Consensus: A Theory of Culture and Informant Accuracy. American Anthropologist 88:313-338. Ross, M. W., L. D. Channon-Little, and B. R. S. Rosser 2000 Sexual Health Concerns: Intervie wing and History Taking for Health Practitioners, Second Edition. Philad elphia, PA: F. A. Davis Company. Roter, D. L., and J. A. Hall 1998 Why Physician Gender Matters in Shaping the PhysicianPatient Relationship. Journal of Womens Health 7:1093. Ryan, E.G., and R. N. Butler 1996 Communication, aging, and health: To ward understanding health provider relationships with older clients. Health Communication 8:191-197 QSR International Pty Ltd. 2002 N6 (NUD*IST, v 6.0). Doncaster, Vi ctoria, Australia: QS R International Pty Ltd. Sadovsky, R. 2000 Integrating erectile dysfunction treatment into primary care practice. American Journal of Medicine, Dec 18; 109 Suppl 9A:22S-8S. Saha, S., J. J. Arbelaez, and L. A. Cooper 2003 Patient-Physician Relationships and R acial Disparities in the Quality of Health Care. American Journal of Public Health 93 (10): 1713-1720. Schneider, E.C., A. M. Zaslavsky, and A. M. Epstein 2002 Racial Disparities in the Quality of Care for Enrollees in Medicare Managed Care. Journal of the American Medical Association 282:583-589. Singer, M. 1998 Beyond the Ivory Tower: Critical praxis in Medical Anthropology. In Understanding and Applying Medical Anth ropology. P. Brown, ed. Pp. 225-239. Mountain View, CA: Mayfield. Reprinte d from Medical Anthropology Quarterly 9(1), March 1995.

PAGE 251

239 Skiest, D.J., and P. P. Keiser 1997 Human Immunodeficiency Virus Infecti on in Patients Older Years: A Survey of Primary Care Physicians' Beliefs, Pr actices and Knowledge. Archives of Family Medicine 6:289-294. Sobo, E. J. 1995 Choosing Unsafe Sex: AIDS-Risk Denial Among Disadvantaged Women. Philadelphia, PA: Pennsylvania Press. Spradley, J. P. 1979 The Ethnographic Interview. Or lando, FL: Harcourt Brace Jovanovich College Publishers. SPSS (Statistical Package for the Social Sciences) 2003 SPSS v.12.0 for Windows. Chicago, IL: SPSS, Inc. Stall, R., and J. Catania 1994 AIDS Risk Behaviors Among Late Middle-Aged and Elderly Americans. Archives of Internal Medicine 154:57-63. Stansbury, J. P, and M. Sierra 2004 Risks, Stigma and Honduran Garifuna Conceptions of HIV/AIDS. Social Science & Medicine 59:457-471. Stansbury, J. P., M. Mathewson-Chapman, and K. E. Grant. 2003 Gender Schema and Prostate Ca ncer: Veterans Cu ltural Model of Masculinity. Medica l Anthropology 22:175-204. State of Florida 2001 Three-Year State Plan Under a nd In Compliance with The Displaced Homemaker Program, Chapter 446.50, Flor ida Statutes STATE OF FLORIDA January 1, 2001 through June 30, 2004. Electronic document, http://www.floridajobs.org/pdg/Disp lacedHomemaker/DHPStatePlan021405.rtf, accessed October 2004. StatSoft, Inc. 2004 Electronic Statistics Textbook. Tulsa, OK: StatSoft. Electronic document, http://www.statsoft.com/textbook/ stathome.html, accessed April 2005. Street, R. 2002 Gender Differences in Health Care ProviderPatient Communication: Are They Due to Style, Stereotypes, or Accommodation? Patient Education and Counseling 48:201.

PAGE 252

240 Strombeck, R., and J. Levy 1998 Educational Strategies and Interven tions Targeting Adults Age 50 and Older for HIV/AIDS Prevention. Research on Aging 20 (6): 912-936. Sturrock, K., and J. Rocha 2000 A Multidimensional Scaling Stress Evaluation Table. Field Methods 12(1):49-60. Theall, K. P., K. W. Elifson, C. E. Sterk, and H. Klein 2003 Perceived Susceptibility to HIV Among Women: Differences According to Age. Research on Aging 25 (4): 405-432. Trotter, R. T., S. C.Weller, R. D. Baer, L. M. Pachter, M. Glazer, J. E. Garcia de Alba Garcia, and R. E. Klein 1999 Consensus Theory Model of AIDS/SIDA Beliefs in Four Latino Populations. AIDS Education and Prevention 11(5):414-426. United States Census Bureau 2001 Duval County Quick Facts. Electronic document, http://quickfacts.census.gov/qfd/st ates/12/12031.html, accessed February 2005 Waterhouse, J. 1993. Discussing Sexual Concerns with Health-care Professionals: Positive Attitudes in Healthy Subjects. Journal of Holistic Nursing 11(2):125-134. Weller, S. C., and A. K. Romney 1988 Systematic Data Collection. Newbe rry Park, CA: Sage Publications, Inc. Weller, S. C., and R. D. Baer 2001 Intraand Intercultural Variation in the Definition of Five Illnesses: AIDS, Diabetes, the Common Cold, Empacho, and Mal de Ojo. Cross-Cultural Research 35(2):201-226. 2002. Measuring Withinand Between -Group Agreement: Identifying the Proportion of Shared and Unique Beli efs Across Samples. Field Methods 14(1):6-25. Wilson, I. B., and S. Kaplan 2000 Physician-Patient Communication in HI V Disease: The Importance of Patient, Physician and Visit Characteristics. J ournal of Acquired Immune Deficiency Syndromes 25(5):417-425. Wilson, M. G., and R. Castillo. 2004 Discussing Sexual Issues with Older Pa tients. Geriatric Times (July/August): 4-7.

PAGE 253

241 Woo, J. K. H., S. H. Ghorayeb, C. K. Lee, H. Sangha, and S. Richter. 2004 Effect of patient socioeconomic status on perceptions of firstand second-year medical students. Canadian Medical Association Journal 170(3):1915-1919. Yates, M. E., R. K. Stellato, C. B. Johannes, and N. E. Avis 1999 The importance of AIDS-related knowledge for mid-life and older women. AIDS Education and Prevention 11(3):224-31. Zaboltsky, D.L. 1998 Overlooked, Ignored, and Forgotten: Older women at risk for HIV infection and AIDS. Research on Aging 20(6):760-775. Zweifler, J., A. Padila, and S. Schafer 1998 Barriers to Recognition of Erectil e Dysfunction Among Diabetic Mexican American men. Journal of the American Board of Family Practice 11(4):259-263.

PAGE 254

242 BIOGRAPHICAL SKETCH Kathryn Elaine Grant received he r Bachelor of Arts degree in sociology/anthropology from Florida Intern ational University in 1992, where she graduated with honors. She received her Ma ster of Arts degree in Latin American Studies with a concentration in anthropology from the Univer sity of Florida in 1995. Her research focused on generational and ethni c identity issues among first and second generation Cuban American women in Miami. After completing her M.A. she worked as an instructor of anthropology, sociology and gende r studies at Flagler College, and taught anthropology courses at the University of No rth Florida. She also worked on a research project assessing older adults attitudes about home health services, cultivating her interest in the health and well-being of older persons. Ms. Grant retuned to the University of Florida in 1998 to pursue a doctorate in medical anthropology with a health and agi ng focus. Through her educational and professional experiences she explored the inte rplay between biomedic ine and the cultural context of modern medical care in the course of her work wi th prostate cancer patients, medical students, and as a research associat e on a project about mature sexuality. Her current work focuses on physician-patients co mmunication issues for midlife women. She has presented academic papers and posters on her current and previous research, and has several research publications. Ms. Grant is currently the Ce nter Coordinator for the UF Women's Health Research Center, where she plans local and national educational activities focused on women's

PAGE 255

243 health issues and conducts womens health research. She resides in St. Augustine Florida, with her husband, Pedro Montes de Oca.


Permanent Link: http://ufdc.ufl.edu/UFE0010808/00001

Material Information

Title: Sex and the Recently Single: Cultural Models of Mature Sexuality and HIV Risk by Physicians and Midlife Women
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010808:00001

Permanent Link: http://ufdc.ufl.edu/UFE0010808/00001

Material Information

Title: Sex and the Recently Single: Cultural Models of Mature Sexuality and HIV Risk by Physicians and Midlife Women
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010808:00001


This item has the following downloads:


Full Text












SEX AND THE RECENTLY SINGLE:
CULTURAL MODELS OF MATURE SEXUALITY AND HIV RISK BY
PHYSICIANS AND MIDLIFE WOMEN














By

KATHRYN ELAINE GRANT


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005





























Copyright 2005

by

Kathryn Elaine Grant















ACKNOWLEDGMENTS

There are many people whom I would like to thank for their assistance and

patience. I am profoundly grateful to my parents, Joe and Jean Grant, who provided the

emotional and financial support that allowed me to pursue my educational goals. To my

husband, Pedro Montes de Oca, I want to express my appreciations for his unending

encouragement, unwavering faith in my abilities, and the patience to see me through this

long, arduous process.

I am grateful to Dr. Virginia Swartz, of the family practice clinic for her enthusiasm

for my research project. Without her help in recruiting physicians to participate, I may

have never completed the project as planned. I greatly appreciate the directors and

coordinators of the state-sponsored displaced homemaker programs in north Florida for

allowing me to visit the program meetings and to recruit my sample from among the

participants. Special appreciation is reserved for the women participants from the

displaced homemaker program, who allowed me into their homes and lives to complete

this project. Their interest in what I was doing was continually inspirational over the

long duration of the project. I am also very thankful to all of the physicians who made

time in their busy schedules to be interviewed for this project.

I offer my sincerest thanks to Drs. James P. Stansbury, Leslie Sue Lieberman,

Sharleen Simpson, and Barbara Zsembik of my supervisory committee. Their timely

responses and instructive criticism kept me on track for completing the degree. I thank

Dr. Lieberman for including me in a research project on mature sexuality, which









cultivated my career and research interest. I am grateful to Dr. Zsembik for her help with

statistical analysis and for nurturing an interest in mid-life women, so that my focus

expanded on traditional aging research. Many thanks go to Dr. Simpson for her very

timely responses and feedback during the dissertation-writing process. Finally, my most

profound gratitude is reserved for Dr. Stansbury, my supervisory committee chair, who

gave me instructional advice, practical research opportunities, and unwavering

encouragement.

I am most fortunate that Dr. Bryan Page took the time to assist me early on with my

proposal. I credit his guidance and that of Dr. Stansbury for my ultimate funding from

the National Science Foundation and the Woodrow Wilson Foundation.

Many friends and colleagues helped me during the research and writing phases of

this project. I want to express my appreciation to Dr. Kathleen Ragsdale, for her

encouraging comments at difficult moments, and practical suggestions. I am also deeply

appreciative to my co-worker and friend, Elspeth Keller, who picked up the slack when I

needed to be away from work for research. I thank my friend and supervisor (Dr. Mary

Ann Burg, director of the Women's Health Research Center at the University of Florida)

for giving me research opportunities that enhanced the skills I needed for the dissertation,

for letting me take time off to work on the doctorate and all of her helpful advice.

I want to remember my dearly departed grandmother, Ethylene Dixon Grant, who

helped to embolden my resolve to complete this process. Finally, I want to thank my

friend Susan Buhler, who passed away recently. In the concluding phases of my project I

missed her uplifting sense of humor and her direct way of encouraging me to always

consider the utility of my research.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iii

LIST OF TABLES ........................................... .............................. ix

L IST O F FIG U R E S .............. ......................... ........................... ....................... .. .. .... .x

A B S T R A C T ...................................................................................................... ............ x i

CHAPTER

1 IN T R O D U C T IO N .................................................................. .. ... .... ............... 1

O objectives and R research Q uestions......................................................... ...............3...
Genesis of the Project ................................ .............................4
Contributions of the Project................... ..... ...............9...
Obtaining Institutional Review Board Approval.............. .................................... 12
F u n d in g .................................................................................................... ....... .. 12
O verview of C hapters .... ................................................................... .............. 12

2 LITER A TU R E REV IEW ................................................................... ............... 17

In tro du ctio n .................... ..... ... .................. ... .. .......... ........ ................ 17
Part One: Issues in Clinical Communication for Midlife and Older Patients............. 19
A geism and H health C are........... ............. ... ................ ......... .......... ..... ........... 20
Gender Issues in Physician-Older Patient Communication...............................22
W om en Physicians/M en Physicians............................................. .................. 23
Gender, Ethnicity and Social Class in Medical Encounters..............................25
Minority Elders and Health Communication: ..........................................26
Im proving H health Com m unication.......................................... ....................... 29
Part Two: Issues in Physician-Patient Communication About Sexual Health ...........30
Patient Perspectives on Physician Communication About Sex and HIV
Prevention ....................... .............. .... ........... ............... 32
What Physicians Do: Discussion of sexuality and Other HIV Exposure
Possibilities in C clinical C ontexts .............................................. .................. 36
Part Three: Epidemiology of HIV/AIDS Among Persons Aged 50+ and Age,
Gender and Ethnicity in HIV/AIDS-Related Knowledge and Risk..................... 41
Surveillance Reports on HIV/AIDS Among Persons Age 50 and Over .............42
E pidem biology in Florida.................................... ....................... ................ 45









Intersection of Age, Gender And Ethnicity in HIV/AIDS Risk & HIV/AIDS-
R elated K n ow ledg e ................................................................ ..................... 4 6
H IV /A ID S in older persons ............................................................... ............... 47
Midlife and Older Women and HIV/AIDS ...................................................... 49
Midlife and Older African-Americans and HIV/AIDS .............. ..................... 52
Perception of Risk: The perils of the "at-risk" label ................ ..................... 56
Testing the Behavioral Change Theory for Older Adults ..............................58
Targeting Older Persons with HIV/AIDS Prevention Messages .....................60

3 DATA COLLECTION: THEORIES AND METHODS ................. ..................... 64

T heoretical C on sideration s .........................................................................................64
In tro d u ctio n ......................................................................................................... 6 4
E th n o g ra p h y ........................................................................................................ 6 5
C ognitive Theory and M ethods...................................................... ................ 66
Consensus Theory and Cultural Schemas ......................................................68
M ethods and Sam pling .. ..................................................................... ................ 72
In tro d u ctio n ......................................................................................................... 7 2
V ignette-based interview s ....................................................... ................ 73
Analysis of the interview and survey data .............................................74
Cultural consensus modeling techniques ................................................75
A analysis of cultural consensus data ........................................ ................ 75
Suitability of M ethodology ............................................................. ................ 77
Sam pling F ram e ... ..................................................................... ............... 78
Sam pling and R ecruitm ent ..................................... ..................... .................. 81
R research Setting .............................. ............................................. 82

4 CULTURAL CONSENSUS DATA AND ANALYSIS ......................................90

F re e listing g .................................................................................................................. 9 1
Pile Sort Procedure .............................................................. ..... ............... ........ ..... 94
Multidimensional Scaling of Pile Sort Data...................................................96
M D S R e su lts ........................................................................................................ 9 7
Clustering of Pile Sort Data...... ......................... ......................103
R atings Procedure .................................................................. ...... ...... .. ........ .... 108
Property Fitting of Ratings and Pile Sorts...................................... 115
C o n c lu sio n ............................................................................................................... 1 1 8

5 AGE, GENDER AND ETHNICITY IN CLINICAL TALK ................................... 120

H y p o th e se s ............... .. ........ ........ .............................. ...... .............................. 12 0
Age and Gender Constructs: Reactions to the Hysterectomy Vignette................. 121
Physician Views of the Hysterectomy Vignette...... ...................................... 126
Sensitivity to Treatment Effects on Sex-life ...................................................... 128
Physician Gender .................. ...... ... ........ ................ 135
Ethnicity Concordance in the Clinical Encounter ....... ................. ...................141









Physician/Patient Responsibility to Raise Sexual Health Topics.......................... 152
D discussion and C conclusion ..................................... ....................... ............... 156

6 PERCEPTIONS OF HIV RISK...... ........... .......... ...................... 161

In tro d u ctio n .............................................................................................................. 16 1
HIV Vignette ........ ... ........ .................. ................... 162
DHP Women's Views of the HIV Vignette ......................... ...................163
Physicians' Views of the HIV Vignette ....... ... ................. ................... 166
Single, B ut N ot for Long .................. ............................................................ 167
D eating and H IV/STD Risks ........................................................... 170
Marriage Doesn't Always Mean Monogamy ........................ ...................172
The Issue of Practicing Safer Sex........................ ........................................ 174
Cultural Consensus Data Related to HIV/STD Risk..................... ...................175
Higher Than Expected Perceptions of Risk ........................................ ............... 176
T testing for H IV /A ID S ............................................................................ ............... 177
Know or Knew Someone with HIV/AIDS and HIV Testing ............................. 179
Other Women in Age Group and Perception of HIV/STD Risk ........................... 180
Suggestions for P hy sicians ............................................................. ..... ............... 184
Physicians' Perceptions of HIV Risk for Their Patients ............... ...................185
R acial/Ethnic Perceptions....................................................... ............... 187
A ge and M arital Issues ................................................................. ............... 189
Medical Education and HIV Prevention Readiness...... ................. .................. 190
Talking A bout H IV Risk ................. ............................................................. 192
D discussion and C conclusion ..................................... ....................... ............... 193

7 DISCUSSION AND CONCLUSION .......... ..........................197

Synopsis of Demographic Data and Overview of Chapters................................198
D H P S am p le ...................................................................................................... 19 8
Physician Sample ... ............................................................... 199
Synopsis of Findings .......................................... ..................... ... .. ............. 200
Cultural Models of Mature Sexuality and HIV................ ...................200
Issues in the Physician-Patient Dyad ....... ... .................................... 205
L im itatio n s .................................................................. .............................................. 2 0 9
Generalizations to Larger Groups ....... ... ......................... 210
Sam pling Issues ...................... ... .... .... ............. .................. 211
Objectives and Interview Instrument Limitations.................. ...................213
Limitations on Analysis and the Problem of Too Much Data .........................213
Significance and Future Research Considerations ....................... ...................214

APPENDIX

A SHORT DEMOGRAPHIC SURVEY FOR DHP WOMEN................................217

B PH Y SICIA N SU R V EY ....................................... .......................... ................ 219









C VIGNETTE INTERVIEW FOR DHP WOMEN.................................................. 222

D VIGNETTES AND INTERVIEW FOR PHYSICIANS...................226

L IST O F R E F E R E N C E S .................................................................................................230

BIOGRAPHICAL SKETCH ................................................................................ 242















LIST OF TABLES


Table page

3-1 Disrupted Marital Status of State and Research Community...............................80

3-2 Mean Age of DHP Participants, Years in Union and Household Composition....... 82

3-3 Frequencies and Percentages of DHP Women Sample (n=44)...............................83

3-4 Frequencies and Percentages of Physician Sample (n=31) .....................................85

3-5 Mean Age of Physicians and Year of Medical School Completion......................85

4-1 Free List Frequencies w ith Both Sam ples........................................... ................ 93

4-2 Terms Rating Highest Level of Agreement for DHP Women .............................112

4-3 Com prison of R ratings K eys....................................................... ............... 112

4-4 Terms of Agreement between Physicians and DHP Women.............................. 113

4-5 Terms of Disagreement Between Physicians and DHP Women..........................113

5-1 Cross-Tabulation of DHP Women and Physicians and Assuming Older Patients
A re Sexually A ctive .............. .............. ............................................. 132

5-2 Cross-Tabulation of Ethnicity and Opinions on Blacks and Whites Being
Treated the Same in Clinical Settings .............. .......................... 149

5-3 Cross-Tabulation of Physician/Patient Responsibility................ .................. 152

6-1 Cross-Tabulation of Age Range and Partner Status.................... ...................169

6-2 Cross-Tabulation of Partner Status and Personal HIV Risk Perception ..............171

6-3 Cross-Tabulation of Marital Status and Risk Perception for HIV/STDs............ 173

6-4 Cross-Tabulation of Know/knew Someone with HIV and Ethnicity and Tested
fo r H IV .............................................................................................................. .. 1 8 0

6-5 Cross-Tabulation of Physician Status and HIV Education ...............................191















LIST OF FIGURES


Figure page

4-1 M DS for DHP W omen Pile Sorts. ...... ......... ........ ...................... 100

4-2 M D S for Physician Pile Sorts...... ............ ............ ..................... 101

4-3 Hierarchical Clustering of DHP Women...................................... 105

4-4 Hierarchical Clustering of Physicians....... ... ........................ 106

4-5 PROFIT Analysis on DHP Women Pile sorts and Ratings..............................1...16

5-1 DHP W omen's Views of the Hysterectomy Story...................... .................. 122

5-2 Physicians Sensitive to How Medical and Drug Treatments Effect Sex-life......... 129

5-3 Female and Male Doctors Talk to Their Women Patients in the Same Way......... 135

5-4 DHP Women's Preferences Regarding Physician Gender by Ethnicity ..............140

5-5 Physician Race/Ethnicity Affects Communication with Women Patients...........143

5-6 Physician Ethnicity Effects Communication by Respondent Ethnicity ...............144

6-1 Frequencies of View of the HIV Vignette ....... ... ...................................... 163

6-2 Partner Status and Dating Interest (Percents)...... .... ................ .................. 169

6-3 Frequency of Partner Status by Age Group and Ethnicity ................................170

6-4 Percentage Tested for HIV by Ethnicity and Age Range (Self-Reported).......... 178

6-5 Percentage Who Think Other Recently Single Women Know of HIV Risk ......... 181

6-6 Others At Risk by Ethnicity, Age Group, and Marital Status (Percents)............ 182

6-7 Patients Aged 45+ Informed about HIV/STD risk...................... ................... 186

6-8 Percentage of M ales and Females At Risk...... .... ..................................... 186

6-9 M medical Education and HIV Preparedness ....................................... ................ 191















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

SEX AND THE RECENTLY SINGLE:
CULTURAL MODELS OF MATURE SEXUALITY AND HIV RISK BY
PHYSICIANS AND MIDLIFE WOMEN

By

Kathryn Elaine Grant

August 2005

Chair: James P. Stansbury
Cochair: Leslie Sue Lieberman
Major Department: Anthropology

This dissertation explored differing cultural models of midlife female sexuality by

recently single women from a displaced homemaker program in north Florida versus

primary care physicians in the same community. Research about clinical communication

suggests that patients and physicians may differ in their perceptions to such a degree that

effective health communication is hampered.

As baby boomers age, senior sexuality is becoming more socially and clinically

important than it was previously. Additionally, throughout the 1990s and 2000s, the

national HIV/AIDS infection rate has remained around 11% for those aged 50 and over,

and in Florida that percentage is higher. The proportion of women infected continues to

rise, particularly among African Americans. These statistical trends are of concern for

women who are re-entering the dating scene after divorce, separation, or widowhood.









My study considered how 44 midlife women conceptualize sexuality and sexual

health for themselves and other women their age, and how 31 physicians conceptualize

sexuality and sexual health for older women patients. These sexual health

conceptualizations included a focus on HIV/AIDS knowledge and risk perceptions, and

their ideas about discussing sexual health in clinical contexts.

African American and Euro-American women between the ages of 45 and 68

completed demographic surveys, healthcare encounter vignette-based interviews, and one

of three cultural consensus modeling techniques: free list, pile sort, and rating procedures.

Attending and resident physicians, mostly from a family practice clinic completed a

similar interview format.

Results showed that the recently single women and the physicians shared

many of the same concepts about midlife sexuality for women, with a few notable

exceptions. Most of the women believed that they were at-risk for HIV; whereas

most physicians considered their younger patients to be at-risk and thought that

risk declined with patients' advancing age and with female status. Women

believed that physicians should discuss sexual health issues with older patients,

and create a clinical atmosphere conducive to discussing sensitive topics.

Physicians felt that they should do a better job of talking about sexual health

issues with older patients, but thought that time constraints and their own biases

about aging and sexuality hindered their progress.














CHAPTER 1
INTRODUCTION

I argue that, in a basic sense, all medicine is cross-cultural -- that patient and healer
inevitably conceive of the world, communicate, and behave in ways that cannot be
reasonably or safely assumed to be similar or readily compatible. Use of a
common language, concepts, and behavior between patient and physician gives the
appearance of shared understandings and effective communication; but appearances
may be deceptive. --Robert Hahn, 1995, p. 265

My study deals with the tension among differing constructions of clinical talk about

sexual health, and the impact of that talk on health and well-being. The quote above

suggest that patients and physicians may differ in their perceptions in such a way that

effective health communication is hampered and effective health care is illusive. That is,

although the physician and patient may share a common language, or for that matter

share any other number of characteristics, they could understand and interpret the domain

of health care in ways different enough to have an ineffectual interaction. For these

actors, the cultural domain of health care may only be partially shared, and for patient

health and well-being, these differences can be minor or critical.

I explored the cultural domain of sexuality for midlife and older women and if that

domain is culturally shared between the women themselves and with primary care

physicians. I considered how midlife and older women conceptualize sexuality and

sexual health for themselves and other women their age, how primary care physicians

conceptualize sexuality and sexual health for midlife and older women, and how these

perceptions impact the clinical encounter. These sexual health conceptualizations include









a focus on Human Immunodeficiency Virus (HIV) knowledge and risk perceptions, and

their ideas about discussing sexual health in clinical contexts.

In the wake of the HIV/AIDS (Acquired Immunodeficiency Syndrome) pandemic,

research on provider-patient communication about sexuality issues has taken on a new

significance. There are numerous studies with women and men of various ages,

ethnicities, and sexual orientation, and with their health care providers, to determine how

and if sexual health issues are addressed (Coates 1999; Epstein, Morse, Frankel, Frarey,

Anderson, and Beckman 1998; Haas and Coe 1997; Makadon and Silin 1995; Metz and

Seifert 1990; Kerr, Valdiserri, Loft, Bresolin, Holtgrave, Moore, MacGowan, Marder,

and Rinaldi, 1996; Nusbaum, Gamble, Skinner, and Heiman 2000; Nusbaum, Gamble

and Pathman 2002; Rich 2001; Sadovsky 2000; Skiest and Keiser 1997). Many studies

have explored what physicians do in their practices with regard to HIV prevention as well

as more general discussions of sexuality, sexual function, and sexual health.

Unfortunately, despite a broad range of participants, many of these studies do not

delineate their finding by age, class, ethnicity or sexual orientation; and large-scale

studies on HIV and/or sexuality communication in clinical contexts usually focus on the

younger, higher-risk groups, excluding midlife and older persons altogether.

Health care providers would be an ideal source of HIV prevention information

since adults in the U.S. average almost three visits a year to a physician, and most people

believe physicians are credible sources of health information (Makadon and Silin 1995).

However, this resource usually goes untapped partly because of physicians' narrow

notions of what constitutes appropriate medical care, their discomfort with discussing

sensitive issues such as sexuality and drug use, limited conceptions about who is at risk









for HIV/AIDS, and time and resource constraints (Strombeck and Levy 1998). Many

providers were trained before HIV/AIDS became widespread, and/or the

sociodemographic shift in risk groups took hold, compounding the issues mentioned

above. Additionally, as is true of the general population, some physicians have ageist

attitudes about older persons in terms of sexuality and other risk behaviors and do not

perceive them as at risk for HIV infection (Strombeck and Levy 1998). However, the

continuing rise of cases among those typically not considered "high risk" indicates that

many physicians are (whether they know it or not) interacting "with HIV-positive

patients and other patients who are at risk for infection that may not fall into the

'traditional' high-risk categories" (Haas and Coe 1997: 403).

Objectives and Research Questions

My study focused on midlife and older women's concepts of sexuality and aging,

and their notions about HIV and STD risks, comparing those concepts with those of

primary care physicians in an urban north Florida community. The primary objective

was to describe cultural models of mature sexuality by analyzing semantic domains

emerging from in-depth interviews with "recently single" (separated, divorced or

widowed) midlife and older women (African American and Euro-American), and from

interviews with primary-care physicians. The central research question is how two

cultural models of mature sexuality might differ in such a way as to inhibit effective

patient-physician communication. Do physicians hold age- and gender-based

assumptions about discussing sexuality and/or HIV/STD-related topics with midlife and

older women? Are physicians' concepts about sexuality and aging different from

women's own sexuality and aging concepts? What do women think about the impact of

the age, gender, and ethnic characteristics of physicians on clinical communication? Are









there differences in women's perceptions about discussing sexuality issues in clinical

contexts based on age, marital status and ethnic groups?

Genesis of the Project

My search for a research topic was not a search at all; rather it was the culmination

of several research experiences, personal experiences, and a long-term interest in

women's issues. I became interested in health care communication about sexuality

through several experiences and activities. First, the health care encounter stories from

family and friends brought the 'struggle to be heard over the noise of medicine' to my

attention. Not all of the stories were negative; the positive encounters were a relief and

surprise me and to the participants.

Secondly, experience with several research projects shaped my interest in

gerontology and health care communication. After completing my master's degree, I

worked as an interviewer on a project (through the University of Florida (UF), Formal

Services Project of NE Florida in 1996) assessing older adults' attitudes toward home

health services. I gained experience using "vignettes" to explore respondents' feelings

and knowledge about available services, and what they believed to be the best path in

solving difficult situations afflicting the elderly. Working on that project cultivated my

interest in the health and well-being of older persons.

After a few years of teaching undergraduate courses in anthropology, sociology,

and gender studies, I returned to UF to pursue my Ph.D. in applied medical anthropology

with a focus on gerontology. As I pursued my coursework, I had the opportunity to work

as an assistant on a nursing anthropological research project with prostate cancer

survivors (Gender Ideology and Prostate Cancer: Veteran's Cultural Model of

Masculinity) that raised my concern about the lack of attention to quality-of-life issues in









healthcare, particularly sexuality issues for older persons. That project aimed to

determine the impact of male gender roles on Veterans' coping skills, health outcomes,

and seeking of medical services.

My interest in clinical communication about sexuality was further developed when

I served as a research associate on a project for the Association of Reproductive Health

Professionals (Mature Sexuality: Patient Realities and Provider Challenges), reviewing

the literature on physician-patient communication about sexuality. During that time I

also assisted with the training of first-year medical students, and served as a standardized

patient on local and national examinations through the Harrell Professional Development

and Assessment Center, UF College of Medicine. That experience gave me a glimpse of

the rigors of medical education and the ways that it shapes the perspectives of

participants.

Finally, through my work and studies, I also came across the story of Jane P.

Fowler, an average Midwestern woman who contracted HIV in her mid-50s while dating

a few years after her divorce. She never thought of herself as at-risk, but once she had

started dating again, she asked her doctor if she should be tested for HIV and other STDs;

her doctor said "not you, you don't need that test" (Linsk, Fowler, and Klein 2003: S249;

personal communication with Ms. Fowler). Two years later, an exam for health

insurance revealed that she was HIV-positive. After reading this story and corresponding

with Ms. Fowler, I searched for all I could find on HIV prevention (clinical and public

health) for midlife and older women who were "recently single" due to divorce or

widowhood. It was soon clear that this population was not considered in most research

on HIV prevention, or in clinical communication research, particularly about sexuality.









Despite their growing numbers, midlife recently single women seemed to be largely

ignored. Thus the research topic of clinical communication about sexuality and

HIV/STD risk for recently single midlife women finally melded into focus. The next

question was where to find a sample of women to work with.

Preliminary Research and the Research Setting. Because of my interests in

gerontology, I had expected to stay in Florida for my dissertation research, since it is

demographically the "oldest" state in the union. Although there are other areas with

higher HIV infection rates, Florida ranks third highest according to the CDC. As it

turned out, choosing a locale close to home for me was not a stretch theoretically, since

Florida has higher rates of HIV/AIDS among women, minorities, and persons over 50

than the national average (Chapter 2).

In 2002, I had the good fortune, through my work at the UF Women's Health

Research Center in Gainesville, to meet the director of the local Displaced Homemakers

Program. I told her about my research interests and she explained that most of the

program participants were midlife and older women who were divorced or widowed.

This was my window of opportunity, and provided my first entree into the field. To gain

a better understanding of both the Displaced Homemakers Program and its participants, I

interviewed the director and her assistant about the program, and researched the program

through the state website.

According to State of Florida documents,

The Displaced Homemaker Program was established by the Florida Legislature in
1976 .... The Displaced Homemaker Program is a state funded program designed
to address the job training and educational needs of Florida citizens who: are thirty-
five years of age or older; have worked in the home providing unpaid household
services for family members; are not adequately employed; likely will have
difficulty securing adequate employment; and, have been dependent on the income









of another family member but is no longer supported by such income, or have been
dependent on federal assistance. (State of Florida 2003)

The program's goal is to help participants attain independence and economic security.

The program is administered through the Division of Community Colleges. There are

rarely any male participants in the program, and most of the women are divorced,

separated, or widowed.

In order to be sure that this was a reasonable sample for what I wanted to do, I

conducted a preliminary pilot project with 18 women in a Displaced Homemakers

Program. I gathered free-lists from several women in the Gainesville program to test the

free list elicitation procedure and to explore the initial parameters of the cultural domain

of aging and sexuality. For this pilot project, a very short anonymous survey was used to

gather demographic information and free list data. The survey was handed out at a

support group meeting and later returned to me. Participants were told that the survey

was voluntary and that the data would never be published in any way. Using these

surveys, I obtained information about the age, marital status and living arrangements,

health status, educational level, and recentness of doctor visits of the respondents. The

free lists on the surveys consisted of 4 to 10 words or phrases (and in few cases, a

paragraph or two) about how they define aging and sexuality for midlife and older

women.

Although there was not enough complete data for formal analysis, the results of

that small pilot study showed the continued importance of sexuality for most of the

women, particularly their desire to be considered sexual beings regardless of their age

and marital status. Several women were sexually active or had recently been sexually

active, though many lamented the difficulty of finding someone to share intimacy with.









This pilot study also indicated that there was a cultural domain that needed to be explored

further.

The director of the program in Gainesville introduced me to a nearby program and

the people involved there. This other north Florida program served a larger population

with more minority women, which was ideal in providing a more diverse sample. The

director of the nearby program was not as interested in my research topic as the other

director who had helped me initially, but she finally agreed to allow me to visit the

program from which I eventually recruited the recently single women of my sample

(Chapter 3). Once I had the opportunity to interview the director and her assistant from

this program, I understood a bit more about her reluctance to participate. She explained

that the program used to focus more on the whole person, but that the state now wanted a

stronger employment focus. She also found that in years past, the health-related topics

were not well received by the participants and had been removed from the program's

content. She seemed to feel that the women would not be interested in my health related

project, but would allow me to give it a try. Fortunately, many women in the program

found the topic of interest and wanted to participate.

The director interviews generated some basic information about how the displaced

homemaker programs operate and some informal demographic information about the

participants. Although demographics variy from year to year, most of the women are in

their 40s and 50s with a range from 35 to about 75 years old; and about 65-70% identify

as White, 20-25% identify as Black, under 10% identify as being a member of a Latino

subgroup, and the remaining 1 to 3% identify as one of the Asian nationalities, American

Indian, and other. Most are either divorced or widowed; fewer are separated or married









with a recently disabled spouse. Both directors defined their participants similarly in

terms of why the women choose to join the program and the benefits of the program. The

vast majority of the women come because they need to work and need to develop the

skills and support to help them find work. Just under half continue in the support-group

programs year after year or sporadically.

Once the site had been established for the women in my sample, I contacted the

Research Coordinator in the department of a Family Practice Residency program (a

friend of the director of the center where I worked) in the area to determine the protocols

for working with the residents and the attending physicians in the department. We met

for lunch and I explained the project to her, and we discussed how she could help. The

research coordinator was extremely helpful, and eagerly introduced me to all of the

physicians she could, each time I came by. I was very fortunate to find someone as

helpful as she, and her utility became even clearer as I attempted to contact physicians in

other departments. My success with recruiting other physicians was poor, and I believe

that the lack of such a contact person as I had with the family medicine program was the

main reason. Although the few physicians that I managed to recruit outside of the Family

Practice program told me of the importance of the issue and their delight at seeing

someone doing this kind of research, the lack of a contact person to serve as a

go-between for the physicians and myself significantly hindered my ability to recruit

doctors from other departments.

Contributions of the Project

My research project has applied contributions that can benefit physician-patient

communication about sexual health concerns, and public health initiatives directed at

midlife and older women concerning sexual health with culturally appropriate









educational campaigns. This project is important because it will inform improved health

care communication about sexuality issues, as well as clinical HIV/STD prevention

efforts with midlife and older women by dissemination of results to the research and

clinical communities through medical and social science publications and presentations.

Additionally, in the future I plan to use the results to develop physician-training modules

on communicating about mature sexuality. My study explicitly focused on

underrepresented groups because of its research/sampling location at a community

college, its understudied sample of divorced, separated, and widowed women age 45 and

over, and its balanced sampling of majority and minority groups. As a means of "giving

something back" to the research participants (and to those who participate in the

Displaced Homemakers Program, but who did not participate in the project) I will give

post-project presentations of the results. I also plan to develop tip-cards that women can

keep in their wallets with respondent recommendations on how to get the most out of

their doctor-office visits.

My project also contributes to the larger body of scientific knowledge. An

ethnographic investigation with a small sample provides an opportunity to explore the

nuances alluded to in larger studies on physician-patient communication about sexual

health topics and HIV risk perceptions. My project builds on the work of Weller and

Baer (2001), Chavez and colleagues (2001), and Trotter and colleagues (1999) by

exploring between-group and within-group variation in health beliefs. In addition to

specifically comparing mature women with physicians in terms of their ideas about aging

and sexuality, and HIV/STD risks using cultural consensus methods, the systematic

cultural consensus data is woven with the interview and survey data to explore the









cultural and structural factors, and the issues related to interpersonal dynamics of the

physician-patient dyad. This situates my micro-level exploration within macro-level

phenomena, including the hierarchical nature of the medical interview and the hegemony

of the biomedical model of disease that often structures healthcare interactions (Baer,

Singer and Susser 1997).

Applied medical anthropology and other social science endeavors in the medical

arena can serve to highlight social and cultural factors that affect health care outcomes

and to incorporate better understandings of these factors in disease prevention. The

anthropological tradition of collecting detailed data is well suited to situating micro-level

events within their larger macro-level determinants. Epidemiologists and social scientists

have shown repeatedly that the biological and social aspects of health are interrelated

(Hahn 1995; McElroy and Townsend 1989). Unfortunately, most medical education

minimizes the importance of social and cultural factors in health care (Good and Good

2000; Hahn 1995).

The body of research that specifically addresses physician-patient communication

about sexual health concerns, including HIV risk, has been steadily growing, yet many

questions remain unanswered. My study examined how physicians conceptualize

sexuality for midlife and older women relative to how women conceptualize it for

themselves, and how those concepts may affect health care interaction with regard to

HIV/STD prevention. Despite the growing evidence that midlife and older persons are

continuing to engage sexually (some in a high-risk manner), many physicians do not

address sexual health issues with these patients. My study considered this issue and









related issues in an effort to raise awareness about physician biases and the social and

cultural factors that contribute to them.

Obtaining Institutional Review Board Approval

Instruments and forms for my project were approved by the UF Institutional

Review Board (Protocol# 2002-U-721) September 6, 2002. In compliance with standard

Institutional Review Board (IRB) regulations, these materials were resubmitted on

August 15, 2003 to extend the approval time period for data collection until August 25,

2004. Participation was voluntary and respondents were compensated for their time

(Chapter 3).

Funding

I am very grateful for the funding to conduct the fieldwork for my research project.

In February 2003, I received notification that I had been awarded two grants. Funding

was provided by a dissertation improvement grant from the National Science Foundation

(BCS-0241675) and a Woodrow Wilson/Johnson & Johnson Dissertation Grant in

Women's Health.

Overview of Chapters

The following pages provide an outline of the chapters in this dissertation. Chapter

2 is a literature review that focuses on issues in physician-patient communication about

sexual health. The first section highlights some of the structural, cultural, and economic

issues that broadly effect health care communication, focusing on the research studies

and reviews that are most relevant for this dissertation research project. Following this

introduction to health care communication issues, I focus on issues in physician-patient

communication about sexual health. This includes a description of research findings on

the role of patient-physician communication in HIV prevention. I reveal the disconnect









between patient and physician perspectives on clinical communication about sex, aging

and sexuality, and HIV/STD prevention, particularly as it relates to midlife and older

persons. Next, the relevance of this research project is clarified by situating it in the

epidemiology of HIV/AIDS among persons aged 50 and older in the U.S., in Florida, and

finally in Northeast Florida where the research takes place. I further show the project's

relevance by exploring age, gender and ethnic trends in HIV/AIDS transmission and

HIV/AIDS-related knowledge. Finally, chapter two concludes with a brief discussion of

the primary prevention strategies that shape the nation's public health initiatives and how

primary prevention models relate to older persons. My premise is that ethnographic

research that looks at intra- and intercultural variation in health-related perspectives can

contribute to the development of more effective health care communication and HIV

prevention strategies.

Chapter 3 describes the methods used in this research. A brief introduction outlines

the methods used and is followed by a more detailed discussion. The explanation of

methods continually links back to the previous chapter by showing how the goals and

design of the research comes out of the questions raised by previous research and

epidemiological findings. I give a description of Cultural Consensus Modeling (CCM)

and health beliefs and then show support for using CCM in combination with traditional

qualitative data collection. The next section describes the surveys and interviews

instruments and their strengths and weaknesses in terms of addressing the research

questions in chapter one. I give a synopsis of the vignette-based interview and refer the

reader to the appendices for more detail, and then discuss the CCM procedures and how

they link with the other data collection methods. The sampling and the interview settings









are explained in this chapter. I explain how participants were recruited for the project,

and how the selection of this sample is supported by the epidemiological data. Finally, I

give a brief introduction to the final samples and their key characteristics.

Chapters 4, 5 and 6 present the research findings. Chapter 4 presents part one of

the data analysis, focusing on sexuality conceptualizations. I provide a description of the

types of analysis used, including software programs, and an explanation of the theoretical

considerations for using these techniques. This chapter focuses on the Cultural

Consensus Modeling data (CCM), presenting the data from the free lists, pile-sorts and

the rating test with the women in the Displaced Homemaker Program (DHP) and the

primary care physicians. These data are presented textually and graphically for their own

merits and are linked to the data from the in-depth interviews in the following chapters.

The interview data contextualizes the CCM data and shows how it is linked to the other

issues in the research project, discussed in chapters five and six. The data on sexuality

conceptualizations are considered in terms of their support of the hypotheses, as well as

their shortcomings for support of the hypotheses.

Chapter 5 continues with part two of the data analysis, focusing on physician-

patient communication regarding sexual health issues. After a brief introduction

describing the various issues addressed and hypotheses, I introduce the data gathered on

physician-patient communication. This chapter contrasts the views of women in the DHP

program and the views of primary care physicians. This section starts with a focus on the

responses to the first vignette and point out the differences and similarities between the

two groups. This follow by gender-, age- and ethnicity related issues in physician-patient









communication about sexual health:. I conclude with an overview of the findings

presented in this chapter, their support of the hypotheses or lack of support.

Chapter 6 concludes the data analysis chapters with part three, which focuses on

HIV risk and prevention issues. After an introduction to the topic of HIV risks, I present

the data on HIV related knowledge from women in DHP program. I introduce some of

the issues raised with the HIV vignette, and then follow it with their comments on

personal risk perceptions and their perceptions of the risk for other women in their age

group and partner status. Next I present the data on HIV risk perceptions from primary

care physicians regarding their patient population. The risk for patients is further

delineated by age, sex, ethnicity and then other various possible variables suggested by

me or the physicians, such as drug use, sexual orientation, sex work, marital status,

number of partners, and so on. This chapter concludes with a comparison of the views

the women in the DHP program and the physicians, linking these findings with the CCM

data.

Chapter 7 concludes the dissertation with an overview of the findings discussed in

previous chapters and a discussion of how the findings square with the literature in

chapter 2. The next section unites the CCM findings with the vignette-based interview,

showing the utility of combined methods, and links the findings about HIV risk

perceptions to other issues clinical communication about sexual health. Limitations of

the research project are discussed on several levels: 1) some of the project goals were not

achieved due sampling issues; 2) some of the data collection methods limit the

generalizeablity of the results; 3) not all of the objective were addressed with the

interview instrument, and 4) time and space limitations on analysis of all the data






16


gathered. This project yielded a rich dataset despite its limitations; so rich that not all of

the data could be analyzed at this time. This chapter concludes with a discussion of the

significance of the data and its utility for improved health care communication regarding

sexual health issues and HIV prevention in clinical context. Finally, the utility of the data

is situated within the framework of anthropological methods and theory and how they are

relevant to health care and health care communication.














CHAPTER 2
LITERATURE REVIEW

Introduction

The end of women's reproductive lives and facile assumptions about their

difficulties in finding partners frequently lead some health care providers to the mistaken

conclusion that sexual issues are unimportant for midlife and older female patients,

(particularly for those who are no longer married due to divorce, separation, or

widowhood). However, evidence suggests that many mature women and men remain

sexually active, some into their 80s and beyond, that some have multiple partners, and

practice a full range of sexual behaviors (American Association of Retired Persons

(AARP) 1999, 2003; Association of Reproductive Health Professionals (ARHP) 2000;

Jacoby 1999; Janus and Janus 1993; Levy 1994; Levy and Albrecht 1989; National

Council on Aging (NCOA) 1998; Stall and Catania 1994). As baby boomers age, senior

sexuality is expected to become socially and clinically more important than it was

previously perceived to be (ARHP 2000; Janus and Janus 1993; Yates, Stellato, Johannes

and Avis 1999). Despite this projected trend, in American society sexuality is often

viewed as a concern of the young. This cultural emphasis on sexuality in youth, and the

concomitant implication that older persons should not be concerned about sexuality or the

potential health risks associated with some sexual behaviors, is poor preparation for age-

related changes in sexuality (Hotvedt 1983). Furthermore, it does not prepare older

persons or their health care providers for clinical communication about sexual issues.









This literature review focuses on issues in physician-patient communication about

sexual health. The first section highlights some of the structural, cultural, and economic

issues that broadly effect health care communication, focusing on the research studies

and reviews that are most relevant for this dissertation research project. Following this

introduction to health care communication issues, I focus on studies that reveal the

disconnect between patient and physician perspectives on clinical communication about

sex, aging and sexuality, and HIV/STD prevention, particularly as it relates to midlife and

older persons. Next, the relevance of this research project is clarified by situating it in the

epidemiology of HIV/AIDS among persons aged 50 and older in the U.S., in Florida, and

finally in Northeast Florida where the research takes place. I further show the project's

relevance by exploring age, gender and ethnic trends in HIV/AIDS transmission and

HIV/AIDS-related knowledge. Finally, this chapter concludes with a brief discussion of

the primary prevention strategies that shape the nation's public health initiatives and how

primary prevention models relate to older persons. My premise is that ethnographic

research that looks at intra- and intercultural variation in health-related perspectives can

contribute to the development of more effective health care communication and HIV

prevention strategies.

My study focuses on midlife and older women, but it should be noted that the

literature concerning aging and health care, aging and HIV related issues, and aging and

sexuality issues is not consistent in how it frames the categories of "midlife", "older", and

"elder." Age ranges for what is considered midlife and what is considered older or

elderly vary widely from one study to the next, making comparison between studies









difficult. In this literature review, whenever possible, I clarify the age groups considered

in a given study.

Part One: Issues in Clinical Communication for Midlife and Older Patients

Physician-patient interaction is shaped and influenced by cultural, structural and

institutional factors and the interpersonal dynamics of the physician-patient dyad.

Physicians communicate with and react to patients differently on the basis of the patient's

age, gender, ethnicity, educational level, and social class. By the same token, patients

also communicate with physicians differently based on a similar set of factors.

Additionally, many of the difficulties patients face in their experiences of health care are

related to some aspect of health care systems, which are becoming increasingly

competitive and resorting to impersonal and routinized care to remain institutionally

viable (Coulter and Fitzpatrick 2000).

Interactions between physicians and patients are often hierarchical as well as

intercultural. These interactions may be construed as hierarchical because of the

hegemony of the biomedical model of disease (Baer, Singer and Susser 1997) and the

structural power dynamics of modern medicine (Baer et al. 1997; Cant and Sharma 2000;

Henley 1977; Lorber 1997); and as intercultural because of sociodemographic and

structural differences affecting the lived experiences of the participants (Cline and

McKenzie 1998). The hierarchical and intercultural nature of the interaction is carried

out in the medical discourse where patients struggle to be heard over the noise of

medicine (Elgin 1999). "Patients enter medical interactions from a position of relative

weakness They enter unfamiliar surroundings in which all the participants seem to

share a common language .. Medical practitioners, on the other hand, are in their 'home

court' in the medical setting" (Fisher 1993: 168).









Doctors have greater power in the medical encounter by virtue of the fact that they

control access to health resources, as well as the dynamics of the health care interaction,

and that control is institutionally regulated and sanctioned (Baer et al. 1997; Cant and

Sharma 2000; Henley 1977; Lorber 1997). However, recognizing the asymmetry of the

relationship does not mean that patients are passive objects in the exchange; rather they

respond within the possibilities of the existing social structure by reacting to or resisting

medical directives and procedures (Baer et al. 1997; Lazarus 1988). Additionally, the

increasing medical pluralism of U.S. society offers alternatives to the dominant

biomedical paradigm that sometimes serve to empower patients who may rely on

multiple medical systems at one time (Cant and Sharma 2000).

There are a host of communication challenges in the interaction between midlife

and older patients and their health care providers. Ageist, sexist, and racist attitudes and

behaviors on the part of providers and patients, age-associated physiologic

communication difficulties, generational, ethnic and gendered differences in

communication patterns, and the influence of a third party all relate to the various layers

of difference that may influence physician-patient interactions (Adler 1998; Cline and

McKenzie 1998; Damron-Rodriguez 1998; Jackson and George 1998; Ikels 1998; Lorber

1997; Ryan and Butler 1996). The complexity of the interaction is often greater for older

than for younger patients because they frequently present multiple medical problems that

are mostly chronic, incurable, and thus contrary to the treatment paradigm in modern

medical training (Adelman et al. 1991; Hahn 1995).

Ageism and Health Care

Ageism results in stereotypes of older individuals as a homogenous group with

declining productivity, health, independence, sexuality, memory, and communication









skills (Ryan and Butler 1996). The ageist attitudes of care providers are complex and

reflect a societal preoccupation with productivity and youth that effectively stigmatizes

aging and illness (Kreps 1990). Ageist attitudes arise from fears about aging and death

(Ryan and Butler 1996; Nussbaum, Thompson and Robinson 1989) and even the

patient's own ageist assumptions and behaviors (Adelman et al. 1991; Coupland and

Coupland 1999; Ryan and Butler 1996). Ageist behaviors arise in clinical settings partly

because providers there are more likely to encounter older adults who are ill, frail,

confused, and hospitalized rather than vigorous, healthy older individuals (Adelman et al.

1991). Ageism among providers also results in attributing disease processes to normal

aging, leading to inappropriate medical treatment (Chiang 1998; Davidhizar and Giger-

Newman 1996; Ryan and Butler 1996).

The problems of ageism are compounded by age-related changes in communication

skills, such as hearing and vision impairments, changes in the ability to process various

types of information, and memory impairments (Ryan and Butler, 1996). Some hearing

loss is common for 50% of adults over age 65, and some degree of visual impairment is

also common (Adelman et al. 1991). The actual incidence of dementia increases with age

from about 2.5% for people in their sixties, to about 22% for those in their eighties

(Adelman et al. 1991). These sensory and cognitive difficulties have important

implications for communication and can impose serious limitations on health related

communication for otherwise healthy adults. Because these changes are socially

stigmatized and seen as symbols of a generalized decline, they are also sometimes

inappropriately assumed to be indicator of the patient's capacities in other unrelated areas









(Adelman et al. 1991). Adding to the complexity of ageist perceptions and physiologic

changes are gender stereotypes, and racial and ethnic stereotypes.

Gender Issues in Physician-Older Patient Communication

Gender issues in physician-patient interactions are of tremendous importance for

the aging patient due to the age differentials in longevity. Consequently the problems of

aging are largely the problems of women (Allman et al. 1999). Women of all ages

generally visit the doctor's office more often than men for multiple reasons relating to

health and social roles (Jenks 1993). A CDC report on women's use of ambulatory

medical care in 1997 and 1998 found that even after excluding pregnancy-related visits,

regardless of race or ethnicity "women were 33 percent more likely than men to visit a

doctor, although this difference decreased with age. The rate of doctor visits for such

reasons as annual examinations and preventive services was 100 percent higher for

women than for men and medication patterns differed significantly" (Brett and Burt

2001: 12). Women often see doctors in the process of caring for children, husbands, and

parents as non-patient clients because of their social role of family caretaker (Jenks

1993).

Several studies and reviews suggest that both physicians and patients are influenced

by gender stereotypes in the clinical encounter (Cline and McKenzie 1998; Hall and

Roter 2002; Haug 1996; Jenks 1993; Street 2002). The gender of the physician relates to

how they interact with patients and to how patients perceive them, and the gender of the

patient impacts how physicians relate to them as well (Hall and Roter 2002; Street 2002).

West (1984) found that doctors consistently underestimate the medical knowledge of

their patients, and tend to talk "down" to their patients. They frequently underestimate

the abilities of patients whose ethnicity, gender, or socioeconomic status differs









significantly from that of most physicians. Many physicians expect women to be difficult

and demanding, and to have more psychosomatic illness episodes than men (Cline and

McKenzie 1998; Haug 1996; Jenks 1993). Women frequently receive less thorough

diagnostic work-ups and less aggressive treatment than men, and women typically also

receive less technical/medical information (Adler 1998; Cline and McKenzie 1998;

Lorber 1997). Female patients often must struggle to be considered competent

individuals when interacting with male physicians, since their concerns are often taken

less seriously than those of male patients (Cline and McKenzie 1998). Women spend

more time asking questions of their doctors (Jenks 1993), but the amount of time women

spend asking questions is "disproportionate to the information yielded" (Cline and

McKenzie 1998: 64).

Women Physicians/Men Physicians

The vast majority of studies on the influence of the physician's gender on

interaction patterns in the medical interview indicate that the gender of physicians and

patients make a difference in medical practice (Adelman et al. 1991; Cline and McKenzie

1998; Hall and Roter 2002; Haug 1996; Lorber 1997; Roter and Hall 1998; Street 2002;

West 1984). In most studies that consider the influence of the sex of the physician,

female physicians appear more apt to provide support and information to their patients,

are more egalitarian, and give more time to consultation than their male counterparts

(Adelman et al. 1991; Haug 1996; Lorber 1997; Roter and Hall 1998). They also tend to

interrupt their patients less than male physicians, and show more interest in what their

patients are trying to communicate (Cooper-Patrick, Gallo, Gonzales, Ve, Powe, Nelson,

and Ford 1999; Haug 1996; Lorber 1997; West 1984). West (1984) found distinct

differences in interruption frequency between interactions with female patients and male









patients, as well as differences between male physicians and female physicians. Male

physicians initiated 67% of interruptions, where as patients initiated 33%. When the

gender and race of the patient was thrown into the equation the asymmetries were

magnified.

Women physicians spend more time with patients and engage in more positive

communication practices, such as agreement, approval and laughter, and are more likely

to build partnerships with their patients by asking for their opinion, and expressing

understanding (Lorber 1997). This has led to a preference for women physicians in

primary care, as well as the traditional primary care specialties of obstetrics and

gynecology, and pediatrics, where direct, hands-on, first-line care is needed.

Although the research indicates that communicative differences do exist between

female and male physicians, the differences are typically not momentous, and on the

whole women and men physicians are more alike than they are different (Roter and Hall

1998; Street 2002). These gender differences in communication styles are most likely an

effect of life-long socialization and interaction patterns (Lorber 1997). Physician and

patient gender stereotypes and attitudes likely generate a priori assumptions about their

capacities and desires for both gender concordance and discordance in clinical

communication (Street 2002). "For example, people generally believe that female

physicians are less likely use aggressive communication strategies (commands,

directives, negative opinions, controlling behaviors) than are male physicians" (Street

2002: 204). While women physicians also do not tend to interrupt their patients as often

as their male counterparts, women physicians are interrupted more often than men

physicians by their patients (Hall and Roter 2002; West 1984). Stereotypes and beliefs









about women and men physicians, combined with the assumption that women are more

interested in psychosocial aspects of health (Roter and Hall 1998), could explain why

patients tend to talk more, divulge more emotional and relational information, ask more

questions, and are more involved in health care decisions when interacting with women

physicians (Street 2002).

Some of the problems associated with frequent interruptions by physicians in the

medical interview are related to contemporary time constraints imposed by HMOs and

hospital administrators (Lazarus 1988; Paget 1993), but also because of the standard way

that physicians shape the meaning of illness for their patients through their use of

question and answer practices (Clark 1996; Good and Good 2000; Paget 1993).

"Importantly, in the traditional clinical encounter the representation of the reality of the

patient's condition is (re)defined in the cognitive structures of the discipline of

medicine-a reality that may be very different from that understood by the patient"

(Clark 1996: 751). This hierarchical exchange is in part developed in medical schools

where students learn to reconstruct the clinical narrative in the context of diagnostic and

therapeutic procedures (Good and Good 2000). Clark (1996) suggests that as the patient

tells his or her "story" to the physician, the physician begins to manipulate the

information conveyed by interrupting to pose strategic questions that shape the

information to conform to the format of the medical report. These structural and

institutional impediments to positive and empathetic health care communication are often

magnified by racial/ethnic and socioeconomic differences of the participants.

Gender, Ethnicity and Social Class in Medical Encounters

The issue of social class intersects with age, ethnicity and gender, because

minority/ethnic groups are more likely than the white majority to be in poverty, women









are more likely to be poor than men, and older ethnic/minority women, are the most

impoverished of all (U.S. Bureau of the Census 1995). More older women than men are

Medicare and Medicaid beneficiaries, and since Medicare pays for only 80 percent of

physicians' services, physicians do not have a "financial incentive for spending optimal

time and delivering optimal care-employing optimal communication" with this class of

patients (Allman et al. 1999: 334); and not all physicians accept Medicare payments

effectively trapping some women in an unproductive physician-patient relationship.

A study of 20 female patients age 65 and older found that women who had only

Medicare coverage had the perception that their physicians did not really care about them

as individuals (Allman et al. 1999). Adler's study (1998) of an ethnically and

economically diverse group of older women with breast cancer, found that some women

tended to behave more passively with their physicians because they lacked other health

care options due financial constraints and poorer health leading to a sense of "forced

reliance" on their physician. Some women may not attempt to assert themselves in the

medical encounter and may feel trapped due to various social, situational and economic

reasons. These issues of interpersonal dynamics and economics are perhaps most

troublesome for ethnic/minority groups.

Minority Elders and Health Communication:

Not only is the elderly population growing, the older ethnic/minority population is

growing faster than the majority elderly population (Damron-Rodriguez 1998; Jackson

and George 1998; Tripp-Reimer 1999) and will likely reach 25% of the citizens aged 65

and older by the year 2030 (Davidhizar and Giger-Newman 1996). Recent studies show

that for patients with similar clinical findings there are differences by ethnicity in

treatment, amount of information provided, and empathy displayed by physicians (Adler









1998; Cline and McKenzie 1998; Cooper-Patrick et al. 1999; Schneider, Zaslavsky and

Epstein 2002; Woo et al. 2004). Jackson and George (1998) argue that Blacks, women

and members of lower social classes are often assigned lower social worth by health care

professionals, and receive the worst care in hospitals. As a result, more impoverished

groups, including the elderly, often delay seeking treatment to avoid negative appraisals

by health care providers (Jackson and George 1998, Lazarus 1988).

Cooper-Patrick et al. (1999), in their consideration of race and gender concordance

and discordance and its impact on patients' perceptions of physicians' participatory

decision-making (PDM) styles, found that patients aged 40-65 years rated their visits as

more participatory than patients younger than 30 years. However, African Americans

and other minorities rated their physicians as having lower PDM scores than did white

patients. Although the differences were lessened when there was race and gender

concordance between physician and patient, African Americans still rated their visits as

less participatory than whites (Cooper-Patrick et al. 1999). Even after controlling for

SES African American and other minority patients had less participatory visits with white

physicians. Another more recent study confirms the above findings. In a cohort study

using pre-visit and post-visit surveys and audiotape analysis, race-concordant visits tend

to be longer and were rated as having more positive affect (Cooper, Roter, Johnson, Ford,

Steinwachs and Powe 2003). However, race/ethnicity is not the only factor affecting the

quality of physician-patient interactions. One nationally representative survey that over-

sampled for Black, Hispanic and Asian households found that the quality of interactions

was lower for non-White patients, particularly Hispanics and Asians, even after

controlling for differences between respondents with and without health insurance (Saha,









Arbelaez and Cooper 2003). Yet race/ethnicity discordance contributed minimally to the

quality of these interactions, and was better explained by physicians' behaviors such as

treating patients with greater respect, spending more time with patients, listening and

involving patients in the decision-making process.

Although communication may be enhanced when the physician and patient belong

to the same ethnic group (or for that matter the same gender, or age group, or social

class), this does not insure that both will have the same explanatory models of illness or

the same expectations for the medical visit (Cooper-Patrick et al. 1999; Ikels 1998). A

study by Green et al. (1994) of older patient satisfaction shows the interrelatedness of

SES, racism and ageism in health encounters.

In this sample of mostly poor, non-white women patients.. .there may be little
expectation that the young white physician working in the highly respected tertiary
care medical center study site would share in decision-making. These data should
not be construed to mean that older patients do not desire to participate in decision-
making. Rather, these data may reflect the past experiences of these older patients
with other physicians and their perceptions of what they can expect from
physicians in this setting. (Greene et al. 1994a: 1284)

Therefore, differential interaction patterns between physicians and their patients are not

only shaped by cultural, generational and gender differences; they are also constructed

out of the experience and history of social inequality.

Although different genders, age groups and ethnic groups vary in their expectations

regarding medical encounters, researchers should be cautious in making cultural value

generalizations that do not consider how the experience of social inequality shaped those

values. The health care system is itself culturally constructed and functions within the

parameters of the larger society. Placing the health care transaction into its larger cultural

context clarifies the social processes that shape the meaning and outcome of the

encounter.









Improving Health Communication

Effective (and affective) communication is key to all human relationships and yet it

is frequently not achieved in health care relationships. Despite good intentions, doctors

often do not listen to or understand their patients' concerns, and therefore, do not

incorporate those concerns into the healing process. Regarding the physician-patient

transaction as hierarchical and intercultural highlights the social and structural

distinctions between individuals and groups within the larger society. Problems in the

doctor-patient relationship have to be understood as a reflection of the power relations of

the larger society (Rhodes 1996). Regardless of ethical pronouncements by the American

Medical Association to treat "all of humanity without distinction by race, gender, age,

religion, socioeconomic status, or sickness condition," there is evidence of substantial

institutionalized inequity (Hahn 1995: 143).

Health communication scholars, linguists and medical sociologists and

anthropologists have suggested a multitude of ideas and techniques to improve health

care communication and ultimately health care outcomes. The approach that appears to

incorporate most of these suggestions is the shift from the traditional biomedical

"problem-based" approach to a "goal-oriented" holistic approach (Clark 1996: 762). In

the goal-based model, health is defined not as an outcome, but as a process, incorporating

"the values of the individual-as embodied in his or her personal and life goals which

can best be determined through a dialogue" between patient and providers (Clark 1996:

763). This processual approach to health and health care empowers patients to assume

greater responsibility for their health, and ideally allows both provider and patient to

transcend some of the barriers of difference that separate them (Clark 1996). The key in

this approach is listening. Medical professionals are often not taught how to listen and









believe that it amounts to something a kin to good manners or note taking (Elgin 1999).

Elgin, a linguist, argues that the crisis in health care is a crisis of language, because of

misunderstandings, a failure to listen and believe the stories of patients, and not

communicating effectively and empathetically. Kleinman suggest, "If we think of the

clinical encounter as a transaction between patient and practitioner conceptualizations of

sickness and 'clinical reality,' then a core clinical task becomes making these explanatory

models explicit and thereafter translating and negotiating between them" (1979: 61).

However, Kleinman cautions that this approach to health care may require practitioners

to recognize that biomedical conceptualizations may be inappropriate in some

circumstances or require adjustments to fit with the patient's experience. Thus culturally

competent care is not about memorizing the attributes of different ethnic and cultural

groups, but about learning to use a "patient-centered approach that incorporates

fundamental skills and attitudes that may be applicable across ethnic boundaries" (Saha et

al. 2003:1718). These skills are crucial to improving health care communication about

sexual health issues, which the following section notes is poorly attended to for older

persons.

Part Two: Issues in Physician-Patient Communication About Sexual Health

An interest in exploring the dynamics of provider-patient communication about

sexuality issues has taken on new significance in the wake of the AIDS pandemic. This

interest is manifest in numerous studies with women and men of all ages, and with their

health care providers, to determine how and if they address topics that are sometimes

difficult to articulate, embarrassing, and even stigmatizing (Epstein et al. 1998; Gerbert,

Maguire, and Coates 1990; Haas and Coe 1997; Makadon and Silin 1995; Metz and

Seifert 1990; Kerr et al., 1996; Nusbaum, Gamble, Skinner, and Heiman 2000; Nusbaum,









Gamble and Pathman 2002; Skiest and Keiser, 1997). These studies have explored what

physicians do in their practices with regard to HIV prevention as well as more general

discussions of sexuality, sexual function and sexual health (Epstein, et al. 1998; Haas and

Coe 1997; Frank and Harvey 1996; Kerr, et al. 1996; Makadon and Silin 1995; Penson, et

al. 2000; Ross, et al. 2000).

Health care practitioners have the potential to be an excellent source of HIV

prevention information for a broad spectrum of persons since adults in the US average

almost three visits a year to a physician (Makadon and Silin 1995). However, this is

seldom the case. "In spite of the recognize need for provider involvement in prevention

education, many obstacles exist, including clinicians' narrow conceptions of medical care,

discomfort with discussing sexuality and drug use, attitudes about HIV/AIDS, and time

and resource constraints" (Strombeck and Levy 1998: 918). Many physicians simply

lack good interviewing skills, especially when dealing with the older patients (Putnam

1996). Furthermore, many providers practicing today were trained before the HIV/AIDS

epidemic began, and thus lack sufficient knowledge and skills to address HIV risks.

Additionally, as is true of the general population, some physicians have ageist attitudes

and assume that those older patients are not at risk for HIV infection (Strombeck and

Levy 1998: 918). However, the continuing rise of cases among those typically not

considered "high risk", indicates that many primary care physicians are, whether they are

aware of it or not, interacting "with HIV-positive patients and other patients who are at

risk for infection that may not fall into the traditional' high-risk categories" (Haas and

Coe 1997: 403).









Although many health care practitioners have begun to incorporate important

disease prevention issues into their practices, sex and drug-related discussions are still

relatively low.

In a recent study, the CDC reported that the overwhelming majority of primary-
care physicians 'usually' or 'always' asked their adult patients about cigarette
smoking (94%) and alcohol use (84%). In areas most directly related to HIV risks,
physicians were more reticent, however. Many fewer inquired about illicit drug use
(49%), condom use (31%), number of sexual partners (22%), or sexual orientation
(27%). (Strombeck and Levy 1998: 918-19)

Despite some improvement in preventive medicine, physician training has

traditionally given higher priority to immediate acute needs rather than to long-term

behavioral interventions. Narrow conceptualizations of the physician's role tend to focus

on patients' complaints rather than offering health promotion services (Makadon and Silin

1995).

Patient Perspectives on Physician Communication About Sex and HIV Prevention

The vast majority of patients surveyed and interviewed over the past decade have

indicated that they would like to talk with physicians about sexual health concerns, but

most have never done so. These studies show that most adult women and men want to

discuss sexual health issues, including HIV, with their health care providers, but would

prefer that their doctor raise the issue (Haas and Coe 1997; Loehr Verma and Seguin

1997; Makadon and Silin 1995; Marwick 1999; Metz and Seifert 1990; Nusbaum et al.

2002; Waterhouse 1993; Zweifler Padila and Schafer 1998). One nationwide survey of

patients revealed that only 15% reported that they had discussed HIV risks with their

doctors in the previous five years, even though 94% had visited a physician during that

time. Nearly three-quarters (72%) of the discussions about HIV that did take place were

at the patients' urging (Gerbert, Maguire, and Coates 1990). Other smaller studies









focusing on reports from both physicians and patients reveal similar finding with regard

to clinical communication about sexual health issues.

In a nursing study, Waterhouse (1993) surveyed a convenience sample of 53 men

and 20 women (age 29-58 years, mean 42 years) who were either current or retired

university employees. This study examined the attitudes of healthy individuals toward

physicians and nurses discussing sexual concerns with clients. The majority of this

mostly white (96%), married (77%), and well-educated (68% with bachelor's degree or

higher) sample thought discussing sexual concerns with nurses was appropriate, and all

subjects thought that physicians should discuss sexual concerns with patients (p < .0001).

Twenty-six percent felt nurses should "always" or "almost always" discuss sexual

concerns with their patients, and 66% responded "always" or "almost always" that

physicians should discuss sexual concerns with their patients. Women showed more

positive attitudes towards discussion of sexual concerns by physicians and nurses than

males, but this difference was not significant (data not shown). Age, marital status, and

educational level had no significant influence on attitudes.

Nusbaum et al. (2002) assessed women's interest and experience with discussing

sexual concerns with their doctors, and identified what they perceived to help or hinder

this discussion. The 964 women who completed the questionnaires ranged in age from

18 to 87 years old and were seeking routine gynecological care. Most participants (78%)

wanted to discuss their sexual concerns with their physicians, and 97.2% felt that their

physicians should be knowledgeable about sexuality. More than half (57%) had

discussed some of their sexual concerns with a physician, and 89.3% of these reported the

discussion was helpful. Another 24% had been interested in discussing concerns but had









not done so, and 35% reported missed opportunities to discuss concerns when they

attempted to raise the subject by physicians did not seem interested or to understand.

Although most respondents (69.7%) reported feeling too embarrassed to bring up the

topic; these women were over 10 times more likely to respond that they would have

discussed their concerns if the physician had asked compared to those not reporting

embarrassment. Patients reported feeling comfortable talking about sexual concerns with

both female and male physicians, but physician gender was reported as effecting their

willingness and ease in discussing these concerns.

In a much smaller, qualitative study, Loehr et al. (1997) investigated the attitudes

of women over 60 toward sexuality and intimacy, including how the medical profession

addresses sexuality and intimacy issues. Fourteen English-speaking women age 60 to 85

in Kingston, Ontario, Canada, from senior apartment complexes or nursing homes, senior

community organizations and volunteer groups participated in focus groups to discuss

sexuality issues. On the whole, the participants felt that sexuality is, and will continue to

be, an important part of their lives. However, like the women in the Nusbaum et al. study

there was uncertainty and hesitation about the prospect of volunteering sexual

information in clinical context. The women in this study suggested that they would feel

more apt to talk about sexuality with their physicians if they were not being rushed; if

they felt their complaint was important to the doctor; if doctors used layman's terms to

explain things, and; if they were not addressed by terms of endearment or age-related

terms like "Granny".

Men's expectations about physician interactions mirrored those of women,

particularly in the desire for physician-initiated communication about sexuality. Metz









and Seifert (1990) surveyed by mail a small sample of primarily white, middle-class,

well-educated men (N = 62) ranging in age from 18 to 73 years. While 97% of the men

reported prior sexual concerns, only 19% had discussed these issues with their doctor.

However, 35% had consulted one or more professionals for assistance with a sexual

concern (psychologists/ psychiatrists 26%, primary care physicians 19%, sex therapist

12%, marital and family therapist 8%, and the clergy 3%). Many were hesitant about

talking to their doctors, but wanted to deal with sexual problems. Only 31% of the man

reported that they were likely to initiate a conversation about sexual concerns with a

physician. They overwhelmingly preferred that the physician initiate the discussion of

sexuality issues (85%). Only 18% indicated they did not appreciate their physician

inquiring about sexual matters. Fewer than a quarter of the men reported that their

physicians had inquired about sexual matters, and only a quarter of the patients have

raised sexual questions themselves.

Zweifler et al. (1998) also focused exclusively on men, exploring the reasons that

diabetic Latino men do not discuss their sexual difficulties with their doctors. Zweifler

and colleagues conducted interviews with 82 Latino (mostly Mexican-American) male

patients who were diagnosed with diabetes mellitus. The men ranged in age from 19 to

86 years, with a median age of 56 years.

Most of the men reported erection difficulties, 59% always had erection difficulties

and 82% indicated that they frequently (more than one-half the time) had erection

difficulties. Only 47% of patients had discussed these issues with their physicians.

Significantly, among the patients with frequent erectile dysfunction, 90% were interested

in receiving information about impotence and 85% were interested in trying a new









method to improve erections. However, less than half of had discussed it with their

physician. Men did not initiate discussion in most cases because they were either

embarrassed or because they expected their physician to initiate the dialogue.

These studies indicate that in most cases half or fewer than half of patients had

actually discussed any of their sexual concerns with their doctor. Repeatedly, one study

after another indicates that most patients would like for their physician to provide a more

comfortable environment for discussing sexuality issues by gently raising the topic.

These studies also reveal that many women and men have sexual issues that they would

like to discuss with a physician (Loehr et al. 1997; Nusbaum et al. 2002; Metz and Seifert

1990; Zweifler et al. 1998); in some cases a particular type of doctor, therapist, or nurse

is felt to be more appropriate than others (Metz and Seifert 1990; Waterhouse 1993) and

gender concordance between patient and physician was preferred more often than not

(Nusbaum et al. 2002; Metz and Seifert 1990; Zweifler et al. (1998). Indeed, physicians

are routinely credited as the most trusted source of healthcare information by the general

public (Haas and Coe 1997; Makadon and Silin 1995). Although physicians can be good

educators for their patients, most have little or no training in how to design, developed,

and present information effectively to others (Haas and Coe 1997).

What Physicians Do: Discussion of sexuality and Other HIV Exposure Possibilities
in Clinical Contexts

Several qualitative and quantitative studies, and review articles, indicate that

physicians and other health care providers are uncomfortable talking about sexuality

issues, including HIV, with their patients (Epstein et al. 1998; Kerr et al. 1996; Makadon

and Silin 1995; Penson et al. 2000; Ross et al. 2000). In particular, primary care general

practitioners, who increasingly have to address HIV prevention and sexual dysfunction









issues in their practices, report difficulty and discomfort in addressing their patients'

sexual concerns (Epstein et al. 1998; Sadovsky 2000). This is particularly evident with

older patients. In 1996 a survey of primary care physicians (n = 124; 72 internal

medicine and 52 in family practice) in Dallas County, Texas, reported that most

physicians rarely or never discussed HIV/AIDS with patients older than 50 years of age,

and were less likely to discuss symptoms of HIV infection or to council older patients for

HIV testing than their younger patients (Skiest and Keiser 1997).

Certainly the physician's specialty is predictive of their likelihood of engaging their

patients about sexuality, but the sex and ethnicity of the physician and patient, as well as

the recentness of medical training also appear to be important factors. A national

random-sample survey of 2,545 primary care physicians to determine the extent to which

they provide HIV prevention and clinical services found that male physicians and the

physicians' perception that patients would be offended by sexual history questions were

predictive of not asking new patients about their sexual behaviors and drug use (Kerr et

al. 1996). OB/GYNs were the most likely to ask these questions, whereas GPs were the

least likely to ask these questions. OB/GYNs were more than six times and those in

Pediatric/Adolescent medicine were four times more likely to ask about condom use than

all other specialties. Additionally, physicians who treated mostly White patients were

less likely to report asking their patients about their sex and drug behaviors. Black and

White physicians were less likely than Hispanic, Asian and other physicians to ask adults

about sex partners. In general, White physicians, male physicians, physicians whose

patients are mostly age 65 and older, physicians who treat mostly White patients, and

physicians who believe that their patients would be offended if they asked questions









about sexual behaviors and /or drug use were less likely to engage their patients on

HIV/AIDS related topics.

Although, physicians appear more likely to discuss sexual issues, including HIV

prevention, with their younger patients than with their older patients, Epstein et al. (1998)

found that physicians' discomfort with discussing these topics results in ineffective

evaluation of the risk status of even their younger patients. In a convenience sample of

17 Family Physicians and Internists and 26 patients age 18-45 who indicated a concern

about HIV risk on a pre-visit questionnaire, videotaped interaction revealed that

physicians did not elicit enough information to characterize a patient's HIV risk status in

73% of the encounters. This appears to be directly related to physician discomfort with

discussing sex related issues. "Nonverbal barriers to effective communication were

frequent. For example, physicians were observed to shake their heads 'no,' avoid eye

contact, or turn their back to the patient while asking about sexual behavior and to initiate

discussions about HIV during a genital examination" (Epstein et al. 1998: 438).

Typically, physicians avoided HIV discussions by changing the subject or simply

ignoring patients' stated concerns and cues that they were interested in discussing HIV

risks. "Patients in our study wanted their physicians to introduce the topic of HIV, but

few physicians had a comfortable standard question or statement with which to introduce

the topic" (Epstein et al. 1998: 440).

One 1990s study, however, indicates that more recently trained physicians may be

more comfortable discussing sex and drug-related issues with their patients. The mail-in

survey indicated that most (81%) recent medical school graduates indicate that they are

moderately or very comfortable discussing HIV/AIDS topics, including sexual behavior









issues, with their patients, but assessed only 0-25% of their patients to be "at-risk"(Haas

and Coe 1997). This survey included both recent medical school graduates and

preceptors from all areas of medicine. The 103 physicians who graduated between 1950

and 1984 were less likely to discuss HIV-related topics than those who graduated after

1985. Both recent graduates and well-established physicians discuss HIV/AIDS with the

patients they perceive as at risk, however recent medical school graduates consider more

of their patients at risk, and felt more comfortable with HIV/AIDS topics, and discussed

these topics with a higher percentage of their patients. For the patient discussion

questions, most selected the 0-25 percentage of their patients category with which they

discussed the following: general HIV/AIDS; HIV/AIDS risk factors; modes of

transmission; "safe sex"; and, abstinence.

Only 9.8% answered 'None' for percentage of patients with whom they discussed
general HIV/AIDS. However, this number increased to 10% did not discuss risk
factors, 12.1% mode of transmission, 17.3% did not discuss 'safe sex' with patients,
and 25% reported not discussing abstinence with patients. For all of these
questions the [St. Louis University School of Medicine] alumni report discussing
these HIV/AIDS-related topics with a higher percentage of their patients than the
preceptors group. This only reached statistical significance in the risk factors and
modes of transmission questions, however. (Haas and Coe 1997: 405)

It is encouraging to learn that more recent medical school graduates are more

comfortable with discussing HIV/AIDS, though their misperceptions about who is at-risk

is not so encouraging. The reluctance of health care providers to discuss HIV/AIDS with

their patients, especially their older patients is a significant disadvantage to prevention

efforts with midlife and older persons.

Indeed, larger quantitative studies indicate that Family Practice physicians (FPs) do

not routinely recommend HIV testing for the majority of their sexually active patients. A

survey of FPs in Rhode Island and Mississippi showed that although 93% tested their









"high risk" patients for HIV, only 7% of their sexually active patients aged 18-50 had

been tested in the previous year (Simmons, Ma, Stumpff and Flanigan "Routine HIV

Testing in the Primary Care Outpatient Setting," poster presented at the October 2004,

North American Primary Care Research Group (NAPCRG) annual meeting, Orlando,

FL.). The difficulty appears to be narrow conceptualizations about which patients that

are at-risk for HIV. These authors conclude, "In order to not stigmatize any specific risk

groups, nor to miss any patients who are unable to be identified as being 'at high risk',

routine testing in the primary care setting should be encouraged."

These studies indicate that most primary care physicians who treat patients aged

45 and older do not discuss HIV prevention topics or sexuality topics with their older

patients (Frank and Harvey 1996; Kerr et al. 1996; Skiest and Keiser 1997; Wilson and

Kaplan 2000). However, those that consider the physician's gender (Frank and Harvey

1996; Kerr et al. 1996; Wilson and Kaplan 2000) noted that women physicians discuss

sexuality issues more than men physicians regardless of physician specialty. These

studies point to the fact that many physicians do not perceive of their older patients as

sexual beings who may have sexual concerns or questions, or may be at risk for STDs.

Indeed what often comes to mind when the dating scene for single women and men is

mentioned are folks in their 20s and 30s. However, this perception is contrary to the

reality as noted in a very recent AARP survey of 3,501 single women and men between

ages 40-69. The survey focused on their lifestyles, dating and romance patterns,

including risky behaviors. The study notes that over a third of all persons in the US ages

40-69 are single, and divorcees make-up a plurality of this group (AARP 2003). About a

third of single women and men in the AARP study were in exclusive dating relationships,









and almost the same percentage (32%) were dating non-exclusively. Many others

indicated an interest in finding a new relationship. This report shows how out-dated

physician and even societal views are regarding sexuality and aging and HIV/STD risks

for older individuals.

"In spite of the recognized need for provider involvement in prevention education,

many obstacles exist, including clinicians' narrow conceptions of medical care,

discomfort with discussing sexuality and drug use, attitudes about HIV/AIDS, and time in

resource constraints" (Strombeck and Levy 1998: 918). Although there are excellent

articles (Dustow 1995; Wilson and Castillo 2004) and books (Miller, Mayer, and

Makadon 1999; Ross et al. 2000) that offer practical suggestions for discussing the

subjects of sex, sexuality, safer sex and HIV with a diverse population of patients, few

providers appears to use them. Below is a review of the epidemiology of HIV/AIDS

among midlife and older individuals and the trends in HIV/AIDS infection and HIV-

related knowledge, indicating the salience and timeliness of incorporating these

communication tools into clinical practice.

Part Three: Epidemiology of HIV/AIDS Among Persons Aged 50+ and Age, Gender
and Ethnicity in HIV/AIDS-Related Knowledge and Risk

In 1989, Catania et al. argued persuasively that the prevention of HIV transmission

"is the single most important AIDS-related public health goal," and that greater attention

should be paid to preventing transmission of HIV among late-middle-age and elderly

individuals (1989: 77). Since that time the need to target prevention efforts at mature

persons, has only increased, with adults age 50 and older consistently representing

approximately 10% of the cumulative AIDS caseload (Strombeck and Levy 1998). In

1998, Ory et al. continued this theme by pointing out that persons age 50 and older are









still largely ignored in both HIV/AIDS scientific research and intervention efforts, despite

being among one of the fastest growing segments of the AIDS caseload.

In the mid-90s, when the number of adults diagnosed with AIDS showed their first-

ever decline, new incident cases among the midlife and older population (50 years of age

and older) rose twice as fast as those 49 years of age and younger (22% and 9%

respectively) (Ory et al. 1998). Nevertheless, calls for prevention efforts, particularly

among health care providers, go largely unheeded. Although sexuality issues are

increasingly visible in the media, news and entertainment, as well as taught in schools of

medicine, nursing, social work, public health and psychology, the skills for talking about

sex and diagnosing sexual problems or counseling patients with sexual concerns remain

under emphasized for many medical practitioners (Ross et al. 2000). Due to the lack of

intervention efforts that have been directed toward midlife and older populations (see

Strombeck and Levy 1998), both within and outside the clinic, it is not surprising that

these populations tend to have less information and more misinformation about

HIV/AIDS risks, and that physicians often underdiagnose and misdiagnose their older

patients because they do not perceive them to be "at-risk" for HIV (Ory et al. 1998). As

among individuals in other populations, gender issues, sexuality issues, structural,

cultural, sociodemographic and economic issues impact how mature individuals are

affected by the HIV/AIDS epidemic. Recognition of the diversity of the older adult

population needs to replace ageist assumptions that misrepresent and homogenize this

group for more effective intervention and prevention strategies.

Surveillance Reports on HIV/AIDS Among Persons Age 50 and Over

Surveillance data from the Centers for Disease Control and Prevention (CDC)

estimates the number of persons living with AIDS increased 7.9% from 1999 to 2000









(CDC 2001b: 3). By the end of 2002, "an estimated 384,906 persons in the United States

were known to be living with AIDS" (CDC 2002: 5). Death rates for persons with AIDS

have continued to decline, mostly due to the use of highly active antiretroviral therapy

(HAART). "During 1998-2002, the estimated number of deaths among persons with

AIDS declined 14%" (CDC 2002:5)

Although the estimated number of AIDS cases diagnosed annually declined

substantially from 1996 through 1999, the rate of decline observed between 1999 and

2000 slowed relative to previous years (CDC 2001b), by 2002, AIDS diagnoses increased

two percent (CDC 2002). Additionally, these declines are not significant among all age,

sex, ethnic and regional categories.

The leveling in overall AIDS incidence is occurring as the composition of the
epidemic is changing. AIDS incidence declined in most populations but increases
were observed in some groups, notably women and persons infected through
heterosexual contact... .From 1999 to 2000, the number of estimated deaths among
persons with AIDS declined 11%; this decline, however, was less than that
observed from 1996 to 1997 (42%)...when the widespread introduction of highly
active antiretroviral therapy (HAART) had its greatest impact... Declines in deaths,
which have been attributed in part to the impact of HAART, also reflect declines in
AIDS incidence which in turn reflect both underlying trends in new HIV infections
as well as delays in disease progression among HIV infected persons who are
receiving treatment. Other factors associated with the slowing of declines in deaths
from 1999 to 2000 compared with those previously observed may include delayed
test seeking among certain populations, limited access to or use of health care
services, and limitations of current therapies among persons in care. (CDC 2001b:
3)

Among the issues mentioned above, delayed test seeking and access to health

services are particularly salient for minorities, older age groups, and women. Even with

delays in testing and reporting, the percentages of minorities, minority seniors and

women infected with HIV, and diagnosed with AIDS continue to increase (CDC 2001a,

2001b; Florida Department of Health 2003a, 2003b, 2003c). As of 1996, the estimated

adult/adolescent AIDS incidence, by race/ethnicity was 46.1% for the White, non-









Hispanic population, Black, non-Hispanics accounted for 34.8%, and Hispanics

accounted for 17.8%. However, by the year 2000, the incidence of AIDS cases had

undergone a significant shift in ethnic make-up. The percentages for White, non-

Hispanics dropped to 37.8% of all estimated cases, Black, non-Hispanics rose to 41.2%,

and Hispanics rose to 19.6% (CDC 2001b: 30). Newer surveillance data from 30 areas in

the U.S. and territories that have instituted confidential-name-based HIV infection

reporting since 1998 through 2002, shows that the estimated number of diagnoses of

HIV/AIDS by race is now 54% Black, 32% White, 13% Hispanic, and 1% other (CDC

2003).

From 1998 through 2002, "AIDS incidence decrease 61% among children and 24%

in the age group 25-34 years, and remained stable in the age group 35-44 years. AIDS

incidence increased in the following age groups: 13-14, 15-24, 45-54, 55-64, and 65

years and older" (CDC 2002: 6). In 2001, 10.9 % of all cumulative AIDS cases (based

on age at diagnosis) were of men and women age 50 and older (20.7% of all persons age

45 and older). Women age 50 and over accounted for 9.5% of all reported AIDS cases in

2001; African American women alone accounted for 50% of this total. Men age 50 and

over accounted for 11.2% of all reported cases (CDC 2001b). If we drop the cut off age

down to 45, then women accounted for 17.3% of all reported AIDS cases in 2001, and

men age 45 and over accounted for 21.4% of all reported cases (CDC 2001b). The

numbers of midlife and older women and men infected, particularly minority women, is

more significant in the southern United States. Throughout the epidemic Florida

continues to be one of a handful of states with the highest HIV/AIDS rates.









Epidemiology in Florida

The South has remained the region with the highest estimated number of persons

living with AIDS, accounting for 38.2% in 2000, though the Northeast follows with a

close behind at 29.8% (CDC 2001b: 38). Florida ranks third behind New York and

California in total adult/adolescent cases (Florida Department of Health 2003b). Florida

accounts for about 11% of the estimated number of persons living with HIV/AIDS in the

US (Florida Department of Health 2003b). However the gender and ethnic trends in

HIV/AID incidence are notable. African Americans account for 55% of cumulative

reported HIV cases in Florida compared to 40% nationally, and women account for 36%

of cumulative reported HIV cases in Florida compared to 29% nationally (Florida

Department of Health 2003b). Of women in Florida diagnosed with HIV and with AIDS

through 2003, 72% were Black, 16% and 17% (respectively) were White, and 11% and

10% (respectively) were Hispanic (Florida Department of Health 2003c).

In Florida, through 2003, 14% of AIDS cases and 12% of HIV cases were reported

in persons age 50 and over. The racial trend noted for all HIV/AIDS cases in Florida are

mirrored in its midlife and older population. Significantly, Black, non-Hispanics account

for almost half (51%) of all seniors (age 50 and over) that have been diagnosed with

AIDS in Florida (Florida Department of Health 2003a). White, non-Hispanics account for

33% and Hispanics represent 16% of senior AIDS cases. Unfortunately, most persons age

50 and over have not been tested for HIV. This is in part related to age and ethnic trends

in HIV/AIDS knowledge and risk perceptions.









Intersection of Age, Gender And Ethnicity in HIV/AIDS Risk & HIV/AIDS-Related
Knowledge

In the early years of the HIV/AIDS epidemic, the source of infection for many

older adults was often through the receipt of contaminated blood or blood products. This

source of infection distinguished them from other age groups who were not as likely to be

blood product recipients (Strombeck and Levy 1998). Unfortunately, when these older

adults became infected through blood transfusions, many health care providers and policy

makers falsely assumed that the infection would not spread based on the stereotype of

older adults as disengaged from sexual activities (Laskin Siegal and Burke 1997).

Laskin Siegal and Burke argue that thishs myth that older adults infected through

transfusion would not go on to infect others sexually has been one basis for the lack of

education of the older population regarding the risks from this phase of the epidemic."

(1997:156).

The good news is, due to improved blood screening methods in place by 1985, the

AIDS caseload of Americans diagnosed at age 50 and older who had become infected

through blood transfusions, decreased from 17% in 1988 to approximately 6% by the end

of 1996 (CDC 1988, 1996, as cited in Strombeck and Levy 1998). The bad news is, HIV

transmission for newly identified AIDS cases in adults over 50 years of age is primarily

through high-risk sexual behavior (Strombeck and Levy 1998), and despite this trend

most midlife and older adults are largely still not targeted by clinical and public health

prevention efforts (Ory et al. 1998, Linsk, Fowler and Klein 2003)

As for the general population, the trend towards more heterosexual transmission of

the virus that causes AIDS appears to be increasing among both sexes and all ethnic

groups. By comparing the most recent trend in AIDS incidence by exposure category









and race/ethnicity diagnosed in between 1998 and 2002, each ethnic group showed slight

increases in the heterosexual mode of transmission and slight declines in intravenous

drug use (IDU) transmission, with the exceptions of Asian/Pacific Islanders and female

and American Indian/Alaska Native whose IDU mode of transmission increased (CDC

2003).

HIV/AIDS in older persons

Mack and Bland (1999) present four interrelated issues that tend to distinguish

older persons from younger persons with regard to HIV/AIDS: 1) the progression from

HIV infection to AIDS occurs faster among older persons, and the progression of AIDS

opportunistic illnesses often progress at a faster rate as well; 2) AIDS, sometimes called

the "new great imitator", causes many of the same symptoms typical of the chronic

diseases that are common among older persons, making correct diagnose challenging; 3)

health care workers do not expect their older patients to be "at-risk" for HIV/AIDS, and

therefore miss many prevention and diagnosis opportunities, because of ageist

assumptions about sexual and/or drug use behaviors among older persons; and 4) more

older people die from AIDS related illnesses in the same month that they are diagnosed

than younger people because of the three issues listed above.

The problem of late diagnosis and misdiagnosis is illustrated in some retrospective

analyses of hospital records for older patients who died without HIV/AIDS diagnosis. Of

170 patients (ages 60 to 79) in Harlem Hospital Center in New York City, blood samples

showed that 6% of the men and 9% of the women were HIV positive. "Nearly all the

cases were misdiagnosed, and few had known risk factors for HIV" (AIDS Alert

1995:79). Another retrospective study at Grady Hospital in Atlanta found that 32 HIV-

positive elderly patients (age 60 or older, mostly Black and male) that presented with









signs or symptoms of AIDS were not tested for HIV until months later (AIDS Alert

1995). "Most of the patients acquired HIV through sexual intercourse (38%), primarily

through homosexual contact, and injection drug use (16%). Blood transfusions accounted

for only 9 % of the infections" (AIDS Alert 1995: 79).

Mack and Bland (1999) point out that age-related issues in HIV infection and

diagnosis are perhaps more problematic for older women because of social and

physiological factors that place them at greater risk of infection, and earlier death from

infection. Physiologic changes in vaginal tissues make tearing during intercourse more

likely, and this increases the risk of HIV infection, particularly when barrier methods are

not used. Also, life expectancy for older women with AIDS is less than it is for older

men, "in part, because older men tend to make more antibody that older women" (Mack

and Bland 1999: 688).

Compounding these physiologic factors is the fact that older women are less likely

than younger women to use condoms since they typically associate them with pregnancy

prevention and/or they do not perceive themselves to be at risk for HIV and other STDs

that barrier methods help to prevent (Mack and Bland 1999, Yates et al. 1999, Zaboltsky

1998). Additionally, women (of all ages) are sometimes at a disadvantage in protecting

themselves from HIV because of their structural position in heterosexual relationships

and cultural values defining their sexual interactions with men (Lear 1997). Financial

dependency on a male partner may put some women at a disadvantage in negotiating

condom use (Florida Department of Health 2003c), and the mere suggestion that condom

protection is needed puts relationship trust into jeopardy.









Midlife and Older Women and HIV/AIDS

The trends in the AIDS incidence rate for midlife and older persons are especially

troublesome for women. Strombeck and Levy (1998) report that from 1991 to 1996 the

number of AIDS cases attributed to heterosexual contact increased 106% and IDU

increased 75% for women age 50 and older, whereas the number of reported AIDS cases

in older men attributed to heterosexual contact increased 94% and IDU increased 53%.

For mature women, particularly ethnic minority women, the trends of HIV/AIDS

infection among those aged 50 and older are particularly devastating. Zaboltsky (1998)

reports that in 1989, more than three-fourths (77%) of women age 65 and older with

AIDS were White (non-Hispanic). This had declined to slightly less than half (48%) as

of 1997. At the same time, the proportion of women in the same age group who were

Black increased from 16% in 1989 to 38%, and Hispanic women increased from 7% to

12% over the same time period. Significantly, younger age groups in the 50+ category

show similar patterns; from 1989 to 1997, non-Hispanic white women age 50-54 declined

from 37% to 23%, whereas Black women increased from 47% to 54% and Hispanic

women increased from 14% to 21%. This pattern is continued in each 50+ five-year age

cohort in the number and percentage distribution of AIDS cases.

The reason for these patterns are unclear; they may be due to women of color aging

with their infection and/or being diagnosed at older ages, or from more of these women

contracting HIV later in life from an HIV-positive man (Zaboltsky 1998). Results from

one national study of women age 50 and older, with heterosexually acquired AIDS,

compared to those age 49 and younger (most of the women from both age groups were

Black and with incomes below $10,000), found that the older women were more likely to

be widowed, separated, or divorced; to live alone; to have less than 12 years of schooling;









and to report having been exposed to HIV through sex with a man whose risk of exposure

was unknown (Schable, Chu, and Diaz 1996). Compared to younger women in the

study, women 50 years of age and older were less likely to have had a male sex partner in

the past year, less likely to have a history of other STDs, more likely to have never used

condoms, and more likely to been tested for HIV while hospitalized. The age-related

difference in condom use is consistent with other survey data indicating very low

prevalence of condom use among midlife and older women and relates to age-related

difference in perceived risk for HIV infection, and the use of barrier methods for disease

prevention rather than pregnancy prevention (Binson, Pollack and Catania 1997; Theall,

Elifson, Sterk and Klein 2003).

Among a subsample of women age 40-75 who participated in the 1990 (wave 1)

and the follow-up 1992 (wave 2) National AIDS Behavioral Surveys nabsS), only 8%

of women had any risk factors for HIV infection (Binson, Pollack and Catania 1997).

While this means that 92% had no risk factors, the remaining 8% represents 4.5 million

midlife and older women in the U.S. The findings were less optimistic for minority

women, and African American women, in particular, had the highest prevalence of at

least one risk factor (12.4%) if they lived in a "high-risk city" (high-risk city (HRC)

refers to 23 metropolitan areas that have large numbers of AIDS cases and large Hispanic

and African American populations, which were used to over-sample minorities). The

length of sexual relationship had little relevance for risk behavior, because women who

had been partnered for 40 years or more were equally as likely to report a risk behavior as

women who had been partnered for less than 5 years. When this finding is combined

with the fact that 13% in the national sample and 22% in the HRC sample did not know









the risk status of their primary sex partner the potential for exposure goes up. The authors

note that when women who did not know the risk status of their primary partner are

included in the definition of risk, "the prevalence of risk increased to 19.6% in the

national sample and 22.7% in the HRC sample" (Binson, Pollack and Catania 1997: 350).

This study also points to a low perception of risk among these women; 90%

reported never having been tested for HIV, and 90% reported never having used condoms

(75% of those who lived in "high-risk cities"). "Moreover, about 90% of the women who

did not know the risk status of their primary partner indicated that they were not at-risk

for HIV" (Binson, Pollack and Catania 1997: 349)

In a study comparing two age cohorts (ages 18-29, and ages 40-71) of mostly

unmarried African American women (85%) on their perception of HIV risk, found that

most of the older women (55%) reported no chance of infection compared to 29% of the

younger women (Theall et al. 2003). Several issues increased the perception of risk for

both groups women including: those who had been cheated on by their most recent

partner, or they had cheated on their partner; know someone with HIV/AIDS or someone

who has died of AIDS; sexual preference for males only; having been tested for HIV;

were treated for an STD in the past year; and perception of low efficacy of condoms or

disliked male condoms. However, most of the women in the study regardless of age

perceived themselves to have no risk or to be at low risk for contracting HIV, even those

reporting at least one risk factor.

Although midlife and older women have been at relatively low risk of HIV

infection throughout the epidemic, there overall rate of exposure has been climbing over

the past decade and this is especially true for ethnic minority women. Age, gender,









ethnicity, and SES all play a role in determining risk behavior, but these demographic

characteristics should not be the foundation for categorizing one group "at-risk" and

another "not-at-risk", particularly for prevention education strategies. Instead, the

prevalence of behaviors, along with the underlying beliefs, customs and behavioral

history, should be understood so that intervention and prevention efforts are relevant.

Midlife and Older African-Americans and HIV/AIDS

Rodgers-Farmer (1999) examines the problem of categorizing an ethnic group as

"at-risk" without contextualizing that risk in her analysis of AIDS knowledge, attitudes

and behavior among African Americans aged 55 and older. Rodgers-Farmer points out,

African-Americans are not disproportionately affected by AIDS because they are
African-Americans, but because prevention programs aimed at this population have
not been culturally specific.. .Nor have these programs adequately addressed
pervasive beliefs in the African-American community that may prevent African-
Americans from participating in AIDS prevention programs ... (1999: 2)

African-Americans may be reluctant to participate in government-sponsored programs on

HIV/AIDS because of common knowledge about past medical abuses, such as the

Tuskegee experiment, and pervasive beliefs that AIDS is part of the white conspiracy to

commit genocide against African-Americans (Rodgers-Farmer 1999). Additionally,

misconceptions about who is "at-risk" for HIV also impede AIDS prevention efforts in

this community (Rodgers-Farmer 1999).

Based on analysis of an African-American subsample (n= 745) from the National

Health Interview Survey (NHIS-AIDS), which explored AIDS knowledge and attitudes

from half of the families who participated in the NHIS, Rodgers-Farmer examined AIDS

knowledge, prevalence of HIV antibody testing, and prevalence of HIV-related risk

among those age 55 and older.









Ninety-eight percent of these respondents perceived their chances of getting

infected with HIV as "low/no chance", and the same percentage perceived their chances

of currently being infected with HIV as "low/no chance" (Rodgers-Farmer 1999). This

may help explain why African-Americans at risk for HIV infection are underutilizing

HIV antibody testing, with only 7.6% of those age 50 years and older with a risk factor

for HIV having been tested compared to 28.9% of those 18-29 years and 26.4% of those

age 30-49 (Grinstead et al. 1997). The Grinstead et al. study also used the NHIS-AIDS,

but made comparisons using the whole sample of 2,717 heterosexual African-Americans

between age 18 and 75. Only 14% of the NHIS-AIDS respondents age 55-75 reported

that they had been tested for antibodies to HIV (Rodgers-Farmer 1999).

Overall, the results suggest that those persons with less than 12 years of education,
those persons whose family income is less than $20,000.00 per year, and those
persons who perceive themselves as not being at risk for HIV infection have low
levels of AIDS knowledge... A substantial percentage of respondents failed to
recognize that persons could still look healthy and have the AIDS virus. This
particular finding is of interest because it has implications for engaging in
unprotected sexual intercourse and the initiating of HIV antibody testing prior to
engaging in sexual intercourse. (Rodgers-Farmer 1999: 11-12)

These African Americans tended to be very knowledgeable about the major modes

of transmission (sexual intercourse, perinatal contact, and sharing needles for drug use),

with 80% choosing the correct answers for questions about modes of HIV transmission.

However, they still had many misconceptions about transmission through casual contact.

For example, over 90% believed that there was a possibility of contracting AIDS
from being cared for by someone who had AIDS; from being bitten by mosquitoes
or other insects; and from being coughed or sneezed on by someone who has AIDS.
Moreover, 80% of the respondents believe that there was a possibility of
contracting AIDS from eating and restaurant with a cook who has AIDS; from
using public toilets; and from sharing eating utensils with someone who has AIDS.
Finally, over 70% believed that there was a possibility of contracting AIDS from
working near someone who has AIDS, and from attending school with a child who
has AIDS. (Rodgers-Farmer 1999: 8)









Many respondents (45%) also believed that there was vaccine to prevent AIDS. Yet, few

respondents had any knowledge about the available drugs to treat HIV, or knowledge

about the benefits of testing and early treatment (Rodgers-Farmer 1999).

As for the perceived effectiveness of condoms, 40% indicated that they did not
know if condoms were 'effective or not'; 35% believed that condoms were '
somewhat' effective; 21% believed that condoms were 'very effective'; and 4%
believed that condoms were' not at all' effective... Only 15 percent of the
respondents knew the oil-based lubricants can cause latex condoms to
break... (Rodgers-Farmer 1999: 8).

The implications of misinformation in HIV/AIDS knowledge are supported in the

recent data on the incidence and prevalence of HIV/AIDS among African Americans.

Although lack of income and lower education status appear to be correlated with lack of

accurate knowledge about certain HIV/AIDS issues, an ethnic comparison of people with

AIDS with different income levels and education suggest that misinformation is only one

of many factors that affect HIV knowledge and transmission rates.

As noted in the previous section on epidemiology of HIV/AIDS, African

Americans suffer disproportionately for this epidemic, particularly in Florida. A recent

study that looked at basic demographic and social information about people with

HIV/AIDS (PWA) over the age of 40 years in one Florida community, found some

notable differences between middle-age and older African-Americans and Whites (Speer

et al. 1999). Not surprisingly, the two groups differed significantly on several

sociodemographic variables. Sixteen percent of the participants from both groups were

60 years of age or older, and 28% of the African Americans and 16% of the Whites were

female. African-Americans were more likely to have high school education or less,

whereas Whites were more likely to have college education. The average annual income

of African-Americans is half that of Whites in the sample at $7,015, and 81% of African-









Americans fell below the poverty level versus 37% of Whites. As for sexual orientation,

82% of African-Americans described themselves as heterosexual, whereas only 28% of

Whites described themselves this way, and 59% of Whites described themselves as

homosexual or bisexual as opposed to 11% of African-Americans (Speer et al. 1999).

More African-Americans thought they had been infected by an opposite-sex partner

(20%), an infected sex worker (14%), a stranger (16%) or by sharing needles (29%),

while most Whites thought they had been infected through sexual relations with the

same-sex partner (44%), and a stranger (23%). Although the groups did not differ

significantly on pre-infection HIV prevention education, 43% of African-Americans and

32% of Whites reported not receiving any prevention education (Speer et al. 1999).

The dramatic differences between these two groups on both socioeconomic factors

and risk behavior factors, suggest that prevention education efforts need to be sensitive to

social and ethnic differences in order to be effective. In particular, African Americans in

this sample were concerned about being assumed to be gay because of their HIV status.

The social stigma associated with HIV/AIDS may be stronger in certain communities and

is often related to misconceptions about the virus and who is "at-risk".

The most recent data from the Behavioral Risk Factors Surveillance System

(BRFSS) for the state of Florida suggest that in terms of HIV testing, African Americans

are ahead of the other two major ethnoracial groups. The BRFSS shows that women in

Florida are slightly more likely to have been tested for HIV than men (49.2%, 47%

respectively) and 59.7% of Black, non-Hispanics reported have been tested, compared to

44.2% of White, non Hispanics, and 52.7% of Hispanics (Florida Department of Health

2003d). These data do not include persons over age 64, and age is clearly an important









factor in the likelihood of being tested in the percentages above with 57.3% of those ages

18-44 being tested, compared to 33.2% of those ages 45-64. Lower income and higher

level of education were also correlated with a greater likelihood of being tested. Notable

for this research project are the data on health professionals discussing STD prevention

through condom use. In the past 12 months, only 14.9% of men and 17.5% of women

reported that a health professional had talked to them about preventing STDs through

condom use. Blacks and Hispanics reported more of this communication (29.9% and

29% respectively) than did Whites (10%), and Black women and Hispanic men reported

the most (both 34%). Again, age is a significant factor with 21.3% of those ages 18 to 44

years reporting this communication compared to 8.8% of those 45 to 64 years of age.

Income is also influencing the communication patterns of health care professional with

31.6 % of those with incomes below $25,000 compared to 7% of those with $50,000 or

more in annual household income reporting that their providers talked to them about STD

prevention.

Clearly, health care professional are targeting specific age and ethnic groups, and

poorer populations with this message. While it may be appropriate to target these groups,

it should not be at the exclusion of other groups. This sort of focus tends to maintain the

stereotypes about those "at-risk" and "not-at-risk" so that older, more affluent

populations may be missing this important message.

Perception of Risk: The perils of the "at-risk" label

Misperceptions about HIV risks and transmission can be dangerous. Although

national level data suggest that most older people in the general population do not engage

in HIV-related risk behaviors, the approximately 10% who do are less likely to report

using condoms during sex, to have ever been tested for HIV, or to of change their









behavior in the past 12 months than younger persons who are "at-risk" (Ory and Mack

1998). This behavior is due in part to older persons not being assigned to the "at-risk"

category by public health campaigns to combat the spread of HIV.

At the closing of the second decade of the HIV/AIDS pandemic AIDS-related

knowledge has increased for all ages, however older people are still less knowledgeable

and younger people. For example, persons age 50 and older are more likely to think that

HIV/AIDS can be contracted through casual contact; they are more likely to believe that

they can be infected by being coughed on, sharing food utensils, or using public toilets

than younger people (Ory and Mack 1998).

The studies detailed in the previous pages note that risks behaviors and perceptions

are often incongruous for many midlife and older persons. Many women and minorities

that are in the older age groups studied about do not think that they are at risk even when

they are engaging in high-risk behaviors.

A study by Dancy (1996) on African American women and their AIDS-related

knowledge, showed that despite adequate knowledge about HIV risks, these women often

did not apply that knowledge to their own behaviors because they decided that it did not

apply to them. Theall and colleagues, suggest that this is the case in their study of risk

perceptions between two age cohorts of women. "Engagement in risky behavior despite

HIV/AIDS knowledge and with no perceived risk of infection is likely a result of denial

and rationalization of behavior" (Theall et al. 2003: 424).

These findings about the perceptions of HIV risks among midlife an older women

are cautionary. As women in the baby boomer generation turn 50, it is unlikely that their

sexual behaviors and attitudes will change radically (Yates et al. 1999). They will









probably continue to engage in more risky sexual and drug-related behaviors than those

in older cohorts. However, "the effectiveness of prevention efforts targeted at younger

adults may carryover into middle age" (Ory and Mack 1998: 661).

The question of how to change perceptions, encourage early testing and safer

behaviors with a diverse population of midlife and older persons remains. Below some

of the more common strategies are considered and how life-course stage may affect

prevention strategies.

Testing the Behavioral Change Theory for Older Adults

There are several theoretical models that have strongly influenced the CDC's AIDS

prevention research: the health belief model, social cognitive theory, the theory of

reasoned action, and the transtheoretical model of behavioral change (Fishbein and

Guinan 1996). In general these models support the examination of beliefs underlying

attitudes and behaviors, as well as the contextual factors that encourage individuals to

behave in a certain way. "According to these models, individuals who are at greater risk,

perceive themselves at greater risk, have more resources, and perceive or actually have

health problems are more likely to be tested for HIV than their counterparts" (Mack and

Bland 1999: 689). All of these models are focused on the individual rather than social

groups, and therefore in many ways neglect the cultural and structural factors that affect

beliefs and behaviors.

Mack and Bland (1999) tested the assumptions of the various models with two

different age groups to assess whether theories of health behavior should take life-course

stage into account. Their models test whether perceived risk, resources, and health

problems affect the likelihood of being tested for HIV among persons age 50-64. Their

analysis of data from the 1996 Behavioral Risk Factors Surveillance System (BRFSS)









compares adults aged 18-49 (77.6% of sample) with adults age 50-64 (22.4% of sample)

on their self-reported chance of getting infected with HIV, change in sexual behavior the

last 12 months, perspectives of condom efficacy in preventing HIV, and marital status.

The percentage of older adults who had ever been tested for HIV was 26.6% and
for the younger adults was 46.8%. The percentage of adults aged 50-64 who are
voluntarily tested was less than half that of adults aged 18-49. Much of this
difference is due to the fact that pregnancy was not being given as a reason for the
last test by older women... The percentage of respondents who report that their
chance of getting HIV/AIDS is high or medium declines with age. The percentage
of adults aged 50-64 who changed their sexual behavior the last 12 months because
of what they know about HIV/AIDS is one-third of the percentage for adults aged
18-49 (5% versus 16.6%). (Mack and Bland 1999: 690-691)

Most older men and women did not rate their chance of getting AIDS as being high

or medium (4.1% and 3.8% respectively). Consistent with other research findings,

Blacks were more likely than Whites to have ever been tested for HIV (38.4% versus

24.4%). Older Blacks rated their chance of getting AIDS as high or medium only slightly

more so than older Whites. "Perhaps the most notable difference by race for adults aged

50-64 was that 16.3% of Black respondents reported that they had changed their sexual

behavior in the last 12 months on the basis of what they knew about HIV compared with

only 3.4% of White respondents." (Mack and Bland 1999: 692). Although the model

confirmed the research results for the younger portion of the sample, (males, Blacks,

those who perceived themselves as having high or medium risk, those who have more

resources, and who have health problems were all more likely to have been voluntarily

tested for HIV other than respondents) the predictive model of health behavior was not

very robust for older adults based on this analysis, as measures of resources and health

status had to be dropped from the model because they made no significant contribution.

Only perception of greater risk was predictive of being tested for HIV (Mack and Bland

1999). This suggest that life-course stage may be important when designing intervention









strategies, and that current programs directed at the general population are not reaching

older persons.

Theall and her colleagues note that their study of risk perceptions among two age

cohorts of women had implications for different types of intervention and prevention

messages/efforts aimed at younger and older women. Among younger women, "HIV risk

behaviors were associated with younger women's perception of risk, suggesting that

knowledge is high but not related consistently to practice... in contrast to the preceding,

our data revealed that HIV risk behavior was not as salient an issue for the older women's

perception of HIV risk" they suggests that programs working with older women should

focus on risky drug use behaviors and sexual orientation issues, since older women

associated drug-related risk with their level of perceived HIV risk, suggesting that these

women were aware of the risks associated with drug using behaviors. They also suggest

that educational efforts for younger women "may include issues such as pregnancy,

dating, and establishing new romantic/sexual relationships, whereas for older cohorts,

topics such as partner communications/negotiation, dating/sex after divorces, and risks

associated with HIV infection for older individuals may be highlighted" (Theall et al.

2003: 426).

Targeting Older Persons with HIV/AIDS Prevention Messages

Several experts in the field of prevention have noted that HIV/AIDS programs are

seldom prepared to consider aging issues, and that most programs serving older adults are

not prepared to address HIV infection as an issue (Linsk et al. 2003; Ory et al. 1998; Stall

and Catania 1994; Strombeck and Levy 1998). Few local or national programs include

screening for HIV-related risk behaviors, such as alcohol and substance abuse or STD

history. Nor do they address the concerns of many older persons about life transitions,









such as loss of a partner through divorce or death, which can lead to changes in social and

behavioral patterns. Although senior focused programs may serve many newly widowed

or divorced individuals, there is little information available addressing the issues of

socialization and intimacy.

Negotiating friendships with possible intimate partners is a new behavior for many,
and educational and support programs may help older adults avoid detachment and
withdrawal. A need exists for information for older adults about how to reduce
HIV-related risk, maintain personal safety, and make decisions about sexual
relationships. Given that older adults are vulnerable to STDs, a basic need exist for
HIV/STDs prevention education integrated with information about meeting one's
social needs safely and avoiding risks HIV/AIDS service organizations,
however, may have information about risk and even about making social contacts,
but these materials may not be age sensitive or even may be overtly ageist by
excluding the possibility of HIV infection/AIDS and older adults. (Linsk et al.
2003: S245)

One area where rapid change needs to occur is in the medical professions. Since

older persons, particularly women, see health care providers more frequently than

younger populations, this is an ideal place to target older persons with HIV prevention

and testing messages. Clearly from the studies reviewed in section two of this chapter,

physicians rarely assess their midlife and older patients as "at-risk" for HIV and other

STDs. Many also express discomfort in discussing sensitive issues, including sexuality

issues, with this class of patients. Linsk and colleagues urge the better education of

primary care physicians and other providers, which are the front line of care and services

for many older persons in the community, to understand and address the fact that HIV

and other STDs can and do infect older people. "Good patient care for sexually active

older persons includes discussions about STDs (e.g., chlamydial infections, gonorrhea,

and syphilis) as well as HIV and how to prevent infection. Health practitioners should do

routine risk assessments and offer HIV/STDs testing to older as well as younger patients

who may be at risk and also should support patient requests for such testing." However,









this practice is rarely the reality. Providers and patients have described how older

persons are routinely discouraged or even ridiculed when trying to initiate such services.

One well-publicized story of a 54-year-old heterosexual woman, who asker her longtime

family practitioner if she should be tested for HIV because she had engaged in a few

intimate relationships after her divorce, only to be told "not you, you don't need that test."

Two years later, when she applied for new medical insurance "but was rejected because a

routine blood tests... revealed a 'significant blood abnormality.' It was then left to the

family practitioner to inform her patient that [it] was, indeed, HIV infection" (Linsk et al.

2003: S246).

In general, based on the information presented in the preceding pages, the rate of

HIV infection and AIDS diagnosis in the U.S. is growing among those who engage

primarily in heterosexual activity; this is most significant for women and minorities

particularly midlife and older women minorities. This is compounded by higher rates of

poverty, age-related social and biological factors, gender-power dynamics, ethnic and

cultural perceptions of sexuality, HIV/AIDS risks and public health initiatives. The lack

of accurate knowledge about HIV transmission and prevention practices among midlife

and older persons points to the need for better education efforts directed at a diverse

aging population. While there is an obvious need for broad public health efforts to target

a general population of seniors through retirement organizations and more specific

regional populations through community outreach; there is also a need to bring health

care providers in to the front line of prevention education in the course of their work.

Although HIV/AIDS and aging often are not perceived as being interrelated, the
increase in the number of cases of HIV infection/AIDS in older adults indicates that
elders are often overlooked in terms of HIV prevention efforts.... As individuals
live longer with HIV infection/AIDS, this new group of long-term survivors will






63


join those affected by HIV infection/AIDS in their later years. Service models for
older persons and HIV/AIDS service models can benefit from targeted education
and better integration of prevention, care, and supportive services. (Linsk et al.
2003: S249).














CHAPTER 3
DATA COLLECTION: THEORIES AND METHODS

Theoretical Considerations

Introduction

Many cultural anthropologists, linguistically trained or not, focus their attention on

what people say and the words they use to express themselves, in their efforts to explain

culture and cultural patterns. James Spradley noted in 1979, in his treatise on

ethnographic interviewing, that in most ethnographic fieldwork, researchers make

cultural inferences from what people say; the way people act; and the artifacts people use.

Since both tacit and explicit culture are revealed in language, ethnographers routinely

focus much of their efforts on what people say, in both casual conversation and lengthy

interviews, in their work to describe culture (Spradley 1979). Linguistic anthropologist,

Deborah Cameron, likewise notes that many anthropologists regard language as a key to

understanding certain aspects of culture, with vocabulary sometimes serving as a

summary of a people's concerns and beliefs about the world around them (1992).

This project relies on what people say to reveal their cultural perceptions. I

combine qualitative and quantitative ethnographic methods to describe cultural models of

mature female sexuality. Data collection includes a short demographic survey; in-depth

interviews following the presentation of health care encounter vignettes; and cultural

consensus modeling techniques including free-list, pile-sort, and rating procedures.

In the first sections of this chapter I explain the rationale for the methods used and

discuss some examples of these methods that relate to the study of health perceptions.









Secondly, I describe the methodology, the sampling and setting and why it was an

appropriate choice. Finally, I describe the entree to the multiple research settings and

provide some basic information about the participants.

Ethnography

Ethnographic interviews begin with the ethnographer talking to one person and

then going to the next, usually finding something different, as well as many things in

common (Handwerker 2001).

We keep track of similarities, note variability, and keep data until we decide that
we've exhaustively identified significant cultural variation... Then, we construct a
story from the inferential generalizations we've arrived at about the people we
worked with, their lives, and the circumstances in which they have lived; about
what those people now think, feel, and do; about who agreed with whom about
what to and to what extent; and, so, about who is similar to whom and to what
extent and how they differ from others and to what extent. (2001: 11-12)

To accomplish the activities described in this passage this project combine structured and

unstructured interview techniques. Bernard (1995) describes unstructured interviews as

the best for getting people to open up and express themselves on their own terms.

However, he suggests that in situations where you might only get one chance to interview

someone, a semistructured interview is best. "It has much of the freewheeling quality of

unstructured interviewing, and requires all the same skills, but semistructured

interviewing is based on the use of an interview guide. This is a written list of questions

and topics that need to be covered in a particular order" (Bernard 1995: 210). This

approach proved to be the best fit for the project I undertook, since in most cases I would

only meet with each informant once. I will provide more detail about the interviewing

techniques in second half of this chapter. As a cognitive anthropology project I also

included the cognitive methods of cultural consensus modeling, which rely on structured

interview techniques to reveal cultural domains.









Cognitive Theory and Methods

This project employs the cognitive method and theory of Cultural Consensus

Modeling (CCM). CCM is based on using mathematical models that simultaneously

record an estimate of a respondent's cultural knowledge about a particular domain and an

estimate of how much they agree with other respondents on what is culturally known

about the domain (Bernard 1995; Weller and Romney 1988). The central theme of

consensus theory is that the correspondence between any two informants is a function of

the extent to which each has knowledge of the culturally appropriate responses for the

interview questions (Weller and Romney 1988).

Owing to its cognitive science foundations the concept of culture in cultural

consensus theory is one that emphasizes what people know and how it affects what they

do, with less emphasis on behavior itself. Handwerker (2002) draws attention to what he

describes as the "largely overlooked and undervalued implications" of Edward Tylor's

1871 definition of culture: "(1) the culture that specific people use to live their lives

constitutes an evolving configuration of cognition, emotions, and behavior unique to

themselves; and (2) cultures consist of evolving configurations of cognition, emotion, and

behavior at the intersection of individually unique cultural sets" (2002:106). He goes on

to explain that while culture is learned and shared by groups of people, "only individuals

learn, and individuals embody and constitute the only source of cultural data. To make

the world we live in sensible, all of us assemble out of our individual sensory experiences

ways of thinking about what we have experienced" (109). In other words, while culture

is shared by groups of people, the experience of knowing one's culture is unique for each

individual.









CCM builds on the concept of culture as shared "cultural model" (D'Andrade 1995;

Dessler and Bindon 2000). Dressler and Bindon (2000) describe a cultural model as a

"highly schematized, skeletal representation of some cultural domain" in that it defines

the elements that make up a cultural domain. A semantic or cultural domain is basically

any subject matter that can be coherently defined (Weller and Romney 1988). According

to Weller and Romney a domain is "an organized set of words, concepts, or sentences, all

on the same level of contrast, that jointly refer to a single conceptual sphere" (1988:9).

Cultural domains can be organized around almost anything; they can be kinship terms,

animal categories, beliefs about symptoms of disease and so on. Among other things,

understanding a domain helps us understand how the world around us works and how to

function in it. Dessler and Bindon describe it this way,

In deciding what to do in any given situation, or in deciding what the behaviors of
others mean in any given situation, we draw on our understanding of the world as
that understanding is structured by our cultural model. So, if we are heading off to
the bait shop, or a baseball game, or a graduation ceremony, or a committee
meeting, we have a pretty good idea what, in broad outlines, to anticipate in others'
behaviors, and what repertoire of behaviors we will draw on. (2000: 247)

Individuals are part of multiple cultural groups, in the sense that they may at any given

time occupy several social roles that require them to have specific cultural knowledge

about several domains. Each individual draws on a large set of cultural models, models

they have learned and shared with others-however incompletely-to get through each

day (Dressler and Bindon 2000). From one person to another, there is variation in the

knowledge of the elements of any given in cultural model, because models, although

shared, their components are also shaped and understood through individual experience

(Handwerker 2001, 2002; Dressler and Bindon 2000).

There are three sources of variability in cultural models within a society. First, as
noted, we incompletely share cultural models; therefore, some people will know









more (or less) about some domain than others. Second, there may be systematic
variability in cultural models across social contexts ... Third, individuals may be
limited for a variety of reasons in their abilities to act on cultural models; that is,
they may know the model, but they may be unable to act in accordance with it (by
economic constraints, for example). (Dressler and Bindon 2000: 248).

The techniques of CCM provide a systematic way to highlight the shape and scope of

cultural domains and the degree of agreement and variability within them.

Since 1986, when Romney, Weller and Batchelder introduced the consensus model

of culture, several other researchers have used and added to the theory and techniques

(Chavez Hubbell, McMullin, Martinez, and Mishra 1995; Chavez, McMullin, Martinez,

Mishra, and Hubbell 2001; Dressler 1996; Garro 1986, 2000, Handwerker 1997, 2002;

Weller and Baer 2001; Weller and Romney 1988). The consensus model of culture has

contributed to anthropological theory by highlighting the shared nature of culture as a

defining feature of cultural knowledge (Dressler and Bindon 2000).

For this research project I use cultural consensus modeling (CCM) to explore if

there is a shared cultural domain of sexuality for midlife and older women, and if primary

care physicians also share that domain, or to what degree they share it. The cultural

domain of sexuality, or sex-life as some preferred to describe it, for midlife and older

women is a bit of a departure from traditional uses of CCM. However, several other

researchers have used consensus theory and methods on abstract concepts and have used

the concept of schemas to shed light on the reasons for cultural variation highlighted

through CCM.

Consensus Theory and Cultural Schemas

As stated above, CCM focuses on shared knowledge and helps to show the

variability within what is shared but not the reasons for the variability. To explore the

factors associated with intra-cultural variation it is helpful draw on traditional qualitative









methods to illuminate the reasons for variation. Ethnographic interviews provide a useful

complement to CCM data collection and analysis. Below is a brief discussion of how

several researchers have successfully combined the quantitative and qualitative features

of CCM with more traditional ethnographic methods.

Linda Garro (2000) used consensus theory to look for patterns and variation when

she studied a group of Ojibway's beliefs about hypertension, but unlike her earlier work

using CCM to examine the shared knowledge of curers and non-curers in Mexico (1986),

this time she used combined methods to get at the nuances of intra-cultural variation.

Using interviews along with CCM, and drawing on Kleinman's work on explanatory

models, she mapped the consensus model for the causes of hypertension among the

Ojibway. However in doing so she also revealed the patterns of variation in the model,

notably demonstrating how individual experience cause some to reject specific elements

of the model or to reframe it based on differing criteria. Garro's work with the Ojibway

led her to develop an analytical approach to some of the shortcomings of consensus

theory. With its focus on what is shared, CCM has limited utility for describing the

elements of difference within a domain; differences do not just indicate variation in

cultural knowledge about a domain, but perhaps a different way of using the domain or

putting the domain together. Garro describes this in terms of schemas or cultural models

theory where variation between individuals reflects different schemas or different aspects

of the same schemas.

CCM and schemas are not in conflict; rather they are complementary tools for

understanding cultural similarity and variation. In a study of gender schemas and

prostate cancer among veterans Stansbury et al. (2003) used CCM to examine cognitive









schemas about gender. With a reliance on primarily emic categories of masculinity the

researchers mapped the masculine ideology of the veterans, showing both the shared

ideology between prostate cancer patients and non-patients, and the variation in that

ideology brought about by the experience of disease. This research did not have the rich

ethnographic data to explore the nuances of the variation that Garro was able to

investigate. Nevertheless the data are supported by schema theory in cognitive

anthropology (D'Andrade 1995; Garro 2000), by showing how the findings support

earlier more qualitative research on the affect of illness on gender constructs. The

research lends support to contention that schemas are "flexible interpretive states that

reflect the mixture of past experience and present circumstances" (D'Andrade 1995:

142).

Several other researchers have used CCM to consider health and health care beliefs,

and some have ventured to determine if beliefs are related to health care behavior.

Hurwicz (1995) compared physicians with elderly patients to see if the two shared similar

ideas about what illness symptoms necessitated office visits. She also explored how the

patients' models of when to visit the doctor's office related to their behavior. The two

groups did share a similar domain about doctor visits, however the patient behavior did

not conform to the health-care seeking norms described in their model. Her analysis of

the health seeking behavior showed that "treatment choices.. .were more a result of the

characteristics of the [illness] episodes than of the different criteria being applied to the

decision" (1995: 232). Hurwicz's analysis of this difference between the model and the

behavior points lends further support to the theory that models of cultural domains are









shaped by a life-time of experience and the factors of current circumstance, influencing

how individuals interpret cultural models and react to them.

Chavez et al. (2001) also address the topic of cultural beliefs and behavior in their

analysis of beliefs about cervical cancer risk factors and the use of Pap exams. For this,

they drew on their 1995 research that looked at intra-cultural variation in beliefs about the

risk factors for breast and cervical cancer among three sub-groups of Latinas, Anglo

women and physicians. They found Latinas, particularly immigrants as opposed to

Chicanas born in the US, held beliefs about cervical cancer that diverged from those of

physicians and those held by Anglo women. In the 2001 report, they found that the

closer the Latinas' beliefs were to Anglo women the more likely they were to have had a

Pap test in the last two years, yet the closer their beliefs mirrored the physicians' the less

likely they were to have had and exam. The distinction here appears to be related to the

emphasis that physicians put on sex-related behavior as an important risk factor vs. Anglo

women who down-play sex-related factors in favor of other factors, such as heredity.

The researchers' earlier ethnographic data show that the immigrant Latinas hold more

conservative moral values about sexual behavior than Chicana and Anglo women, and

this influences their health-seeking beliefs and behaviors. However the researchers are

quick to point out the strong influence of structural factors on health-seeking behaviors

such as lack of health insurance, educational level, and acculturation.

These findings point to the importance of the complementarity between structural

and cultural explanations in understanding health beliefs and behavior. Chavez et al. also

credit their early ethnographic interviews with a small sample of Latinas for the richness

of the data and its usefulness in clarifying the reasons for the intra-cultural variation in









the cultural model, and the generalizability of ethnographic findings on health-beliefs to

much larger populations based on smaller ethnographic studies.

The measurement of shared cultural knowledge has been operationalized in the

cultural consensus model and verified for small sample sizes by several anthropological

researchers (Chavez et al. 1995, 2001; Garro 1986; Handwerker 2002; Handwerker and

Wozniak 1997; Romney, Weller and Batchelder 1986; Weller and Romney 1988). This

project uses a relatively small sample as well to explore the domain of sexuality for

midlife and older women. To address the nuances of intra-cultural variation within the

model the CCM data is combined with an interpretive analysis of the vignette-based

interview data to contextualize the domain. What follows is a description of the methods

used in this project.

Methods and Sampling

Introduction

I created vignette-based interviews to address women's and physicians' ideas about

the clinical encounter, as well as general perspectives on aging and sexuality. These

semistructured interviews relied on open-ended questions and probes to elicit a broad

understanding of similarity and variation in aging and sexuality concepts among the

"recently single" women in a displaced homemaker program (DHP) and among primary

care physicians (referred to as physicians) in the same city. The vignette-based

interviews with both samples explored variation in ideas about provider-patient

communication, beliefs about aging and sexuality, and individual and social factors that

may enhance or inhibit communication about sexuality issues, including HIV/STD issues

in clinical context.









I also administered a survey (Appendix A) with the women that gathered

demographic data on age, ethnicity, marital status, number of children, number in current

household, level of education, years lived in the area, health status, recentness of doctor

visits, and type of doctor visited. Surveys gathered from physicians (Appendix B)

provided information on the physician's age, recentness of medical training, type of

practice, demographic profile of their patient population, perceptions of time constraints

on patient-physician interaction, perceptions of HIV risk for their patients, sources of

information on sexuality and aging, and ideas about patient and physician responsibility

regarding discussing sexual health issues.

Vignette-based interviews

I read each DHP woman two vignettes involving a women and her clinical

encounter, 1) a story about a women who undergoes a hysterectomy and suffers painful

intercourse after the surgery, and 2) a story about a women not being tested for HIV by

her primary care physician and only being tested after she is hospitalized. After each

story, I asked the respondents to tell me their opinion of the case, other more detail

questions, and then general questions about health care communication and sexuality and

aging (Appendix C).

Physicians participated by either completing the survey and one of the CCM

procedures, or were interviewed using the same vignettes discussed above (Appendix D).

However, in order to save time I divided the interview schedule into two parts: the first

group of physicians who completed the full interview (n=9) was read the HIV story, and

the second group was read the hysterectomy story (n=8). Some of the interview

questions were similar to those asked of the DHP women sample, but several medical









practice questions were also asked concerning their current views and practices regarding

communication about sexual health issues with midlife and older women.

Analysis of the interview and survey data

All of the survey data and key responses from the interview data were coded and

input in to the Statistical Package for the Social Sciences (SPSS, v.12.0, 2003). The

interview data were coded for text units in N6 (QSR N6, v 6.0, 2002) so that the

appropriate context of the aggregate responses could be explored. In several instances I

sorted the interview data in N6 so that it could be added to SPSS. For example, the

question about whether or not the DHP thought that other women their age were aware of

the risk for HIV/STDs was an interview question and generated long responses, which

were analyzed and grouped under "yes", "yes, most", "no", "no, not most", and "don't

know." This interpretive analysis was applied to all interview responses that could be

ordered in such a way based on the aggregate of replies.

The SPSS data are analyzed for frequencies of the survey and interview questions

and presented graphically in chapters five and six. The two samples are analyzed

separately using frequencies and Chi-square tests to measure internal agreement, and the

samples are combined on variables that are common to both groups to measure

agreement between groups. For the DHP women, marital status, age groups, ethnic

groups, and health status are used to determine if these variables influence their responses

to the vignette interview questions. The physician sample is analyzed in much the same

way as the DHP women, but on fewer variables (residency status, age groups, and

gender). The coded N6 text data are used to highlight the nuances of the statistical data.

The text data also provides crucial insights to understanding the consensus data,









described below, ultimately linking the various data sets to reveal the cultural schemas of

mature sexuality.

Cultural consensus modeling techniques

The interview format for the women and the physicians also included the cultural

consensus modeling (CCM) procedures. Approximately half of the physician sample

participated in the survey/interviews/CCM procedure and the other half were only asked

to participate in the CCM procedures and complete the survey.

The respondents were asked to participate in one of three systematic data collection

procedures for CCM. The first procedure, free listing (qualitative) of terms and phrases

associated with the domain of mature sexuality, occurred with the first group of DHP

women interviewed (n=18), and physicians (n=14). These free-lists were then be

combined and distilled into a shorter version of the lists by collapsing repeated terms into

a single term and/or removing terms and phrases mentioned by only one respondent (see

Borgatti 1999), to be used in the second step, pile sorting. This second procedure the

next group of DHP women (n=15) and physicians (n=7) to sort the terms and phrases into

categories or "piles." The list is then used in the final step, the rating procedure. The

remaining participants (n=14 DHP women and n=10 physicians) were asked to rate the

relevance of the terms to describe sexuality for women age 45 and older. However, due

to the small sample size for this study, I asked three of the DHP women who completed

the free list procedure to assist me by also completing the rating procedure. These

procedures and their results are described in chapter four.

Analysis of cultural consensus data

The results of the CCM data are analyzed using a variety of methods, all of which

are described in greater detail along side the results in chapter four. All of the CCM data









are imported and analyzed using ANTHROPAC (Borgatti 1996a). The initial free listing

techniques elicits each individual's notions about the particular domain in question

without imposing the researcher's own concepts. The results of the combined lists from

all of the respondents that completed the free list task are then analyzed for frequency and

salience of terms to find the core of the domain.

The pile sort and ratings data are used to highlight the structure of the domain and

help to determine if there is one cohesive domain and/or if there are subdomains and

multiple domains. The single pile-sort task (Borgatti 1999; Weller and Romney 1988))

elicits the way that the individual respondents organize the 48 most frequently mentioned

terms from the earlier freelist data. The results provide the coordinates for a two-

dimensional representation, using multidimensional scaling (MDS); showing how the

respondents believe the domain ought to be structured. These results are also analyzed

with cluster analysis, which is another way to show how items are grouped together.

The ratings data are used to show the degree of agreement among the respondents,

in this case, their perception of the relevance of the terms for describing women's midlife

sexuality. The results of this task generate a profile matrix for cultural consensus

analysis (Romney, Weller and Batchelder 1986). This provides a way to systematically

measure the degree of shared perceptions among respondents and ascertain the reliability

of the data from the average intercorrelation among respondents (Weller and Romney

1988). Additionally, this analysis can be used to determine degree of agreement within

and between groups; in this case between the DHP women and the physicians.

Finally, the ratings and pile-sort data can be used together to test hypotheses about

the characteristics that shape respondents' notions about the similarities among the items









tested using property-fitting (PROFIT) (Borgatti 1996b). PROFIT is a multiple regression

technique that uses the coordinates from the MDS pile-sort data as independent variables

and the attribute used in the ratings test as the dependent variable. This analysis assesses

whether or not the hypothesized attribute was influencing the suspected similarity among

the items. These methods and their analyses will be described in greater detail in the next

chapter.

Suitability of Methodology

Due to the sensitive nature of this research topic, the combined methods of cultural

domain analysis and in-depth interviews with the vignettes were appropriate for this

project and this population because they permit respondents to depersonalize their

responses if they so choose, on this highly personal topic. This is because the vignettes

describing a hypothetical encounter were followed with general questions regarding

respondent opinions on the pros and cons of the encounter. Respondents were also told

that their responses would be kept confidential, that there were no right or wrong

answers, and that their responses did not need to be based on their personal experience,

but instead may just reflect their general knowledge and opinions about doctor-patient

communication. In this way the respondents were able to choose for themselves how

much or how little personal information they wished to provide. The cultural consensus

procedures, which revolve around questions about salient elements of sexuality and aging

generally, not individually, were also designed to limit personal information.

Additionally, although it may seem counterintuitive to suggest that face-to-face

interviews are an ideal data collection strategy for talking with mature women about

sexuality, because they are presumed to be more uncomfortable discussing sexual issues,

several researchers report the contrary (Gibble, Rogers, Miller and Turner 1998; Herzog









and Rogers 1988; Levy and Albrecht 1989). Preliminary findings suggest that the utility

of self-administered surveys for younger respondents is not duplicated with older

respondents (Gibble, et al. 1998; Herzog and Rogers 1988; Levy and Albrecht 1989).

Gibble et al. (1998) analyzed data from a large survey experiment investigating the

impact of the privacy of self-administered vs. interviewer administered survey modes.

The experimental results hint that it may be unwise to generalize findings on survey

mode effects on younger populations to older populations. In-person interviews have the

advantage of allowing the researcher to build an empathetic relationship with a

respondent in the course of gathering information. The use of open-ended questions in

unstructured or semi-structured format also allows the researcher to probe for information

that may otherwise be neglected, and to clarify the meaning of responses (Bernard 1995;

Levy and Albrecht 1989; Spradley 1979).

Sampling Frame

As an ethnographic research endeavor that includes cultural consensus modeling

this project is an exploration about cultural patterns that are expected to have high

agreement. It therefore does not need a random sample to serve as a valid measure of

cultural pattern. Although each individual assembles out of their own experience ways of

thinking about those experiences and construct an understanding of the world around

them, their various ideas, emotions, and behaviors intersect in multiple ways with the

culture of other individuals that have shared similar experiences (Handwerker 2001).

This is the point of a sampling frame that encompasses certain life experiences-in this

case midlife and older women becoming single again after divorce or widowhood-"that

may influence the patterns of social interaction through which people construct culture

that include explicitly measured internal validity confounds (Handwerker 2001: 99). The









number needed to ascertain the cultural domains depends on the degree of cultural

consensus in the population considered. Borgatti notes that "if every informant gives the

exact same answers, you only need one," but since this is unlikely he suggest computing

the frequency counts of free-list data after obtaining about 20 or so lists until the

frequencies of the top items remain the same as informants are added (1999:122-123).

The age range, ethnic makeup, marital status and gender specific focus of this

project is deemed appropriate based on epidemiological data showing a growing number

of women, particularly ethnic minority women, in this age range as HIV+ through

heterosexual transmission, and due to the lack of prevention programs targeted at people

age 50 and older (Rich 2001; Levy 1998; Ory and Mack 1998; Zaboltsky 1998). The age

range is also deemed appropriate for cohort comparisons because as women baby

boomers enter middle-age, it is predicted that they will be more sexually active and

engage in more high risk behaviors than what is occurring in the current older

populations (AARP 2003; Levy 1998; Yates et al. 1999).

Five factors combine to make women age 45 and older who join the Displaced

Homemakers Program in North Florida an ideal sample and location for investigating

patient and physician beliefs about aging and sexuality, and HIV issues:

* Florida as a whole is the Bellwether state for aging issues due to the fact that it is
the oldest state in the union in terms of demographics, which means that it is an
ideal place to study concepts about aging and sexuality among physicians and
patients;
* The spread of HIV/AIDS among midlife and older persons, particularly among
women and minorities has been increasing over the last decade, and this is
particularly true in Florida;
* The African American population in this area accounts for almost 30 percent of the
total population, far exceeding the national average of 12 percent, and the Florida
average of 14.6 percent (U.S. Census Bureau 2001), thus providing a larger sample
of minority participants;









* Many recently divorced, separated and widowed women are reentering the arena of
sexual activity in an age very different from the years before the discovery of HIV
and its routes of transmission (Rich 2001), which suggest that their beliefs about
aging and sexuality are not only of clinical importance, but that their views may be
quite different from that of physicians; and
* The percentage of persons divorced and separated in this area slightly exceeds that
of the state as a whole (United States Census Bureau 2001), although the widowed
population is smaller (Table 3-1). The number of women in each of these
categories exceeds men. All of which suggests that the issues of this project may
be salient for this population.

Table 3-1. Disrupted Marital Status of State and Research Community
Florida Metropolitan North Florida
Divorced % 11.6 13.4
Separated % 2.4 2.9
Widowed % 7.9 6.4

Therefore the ideas about aging and sexuality are relevant and timely for this community,

and the findings will serve the clinical communication and HIV prevention needs this

underserved population.

Additionally, primary care physicians working in general practice or family

medicine are generally the most frequently visited care providers. Drawing physician

participants from a hospital-based clinic that serve an older, poor, largely minority and

mostly female population, was appropriate for addressing the issues in this project.

The in-depth approach of this qualitative project made the use of a very large,

randomly selected sample prohibitive. Although cultural consensus modeling is an

astonishingly powerful method with a small number of respondents, the results cannot be

generalized to a national cohort of "recently single" women between the ages of 45-70,

nor to a general population of primary care physicians. Additionally, relationships

among some demographic variables are expected to be anecdotal because of sample size

and the lack of applicability of some standard statistical analysis.









Sampling and Recruitment

The DHP women participants for this project were selected through a purposive

sampling method aimed at women age 45 to 70 who participated in the state sponsored

Displaced Homemakers Program in a large, north Florida city between October 2002 to

June 2004.

Women were recruited in-person at group meetings at the community college

where the Displaced Homemakers Program meeting takes place. I introduced myself and

explained my student status and what my research was about. I passed around a paper for

the women to write their names and contact information if they were interested in

participating. I explained that the interviews could occur in their homes, at the college or

an alternative location of their choosing. Participants were compensated for their time

with a $20 grocery gift certificate as an incentive to participate.

The research plan initially proposed to select 60 or more DHP women over the

duration of the project (20 African-American, 20 Latino, and 20 Euro-American women),

however the number of Latinas attending the program during the year and a half that I

recruited subjects was extremely low, and of those in attendance few met the age (45 or

older) and marital status criteria (divorce, separated, widowed or the equivalent of these

statuses for a long-term consensual union). Therefore the final sample of DHP women

respondents is 20 self-identified Black/African American women and 23 self identified

White/Euro-American women, and one "other". The ethnic group percentages of this

sample generally reflect the attendees of this program, except that the Latinas tended to

be younger during the research time period. The primary demographic difference is that

the program caters to women age 35 and older whereas this research project included

only those age 45 and older. The average age of my final sample was 52, whereas the









average for the program tends to be closer to 45 (personal communication with program

director). (Tables 3-2 and 3-3 show characteristics of the DHP participants).

The research plan initially proposed to select a physician sample of 40 in order to

have enough participants to complete the CCM procedures. The final sample ended up at

31 due to the difficulty of recruiting physician participants. I ultimately accepted lower

than ideal numbers of participants to run the CCM analyses.

Thirty-one primary care medical residents and physicians were recruited to assess

current physician views and practices regarding aging and sexuality. Almost two-thirds

of the physicians participated by completing a short survey, in-person interview and one

of the CCM procedures, and the rest were asked to participate by completing the survey

and a CCM procedures only. For their time they received either a $10 or $20 restaurant

gift certificate, depending on whether they chose the long or short interview format.

Research Setting

The setting of this research project is not isolated to one small community as is

often typical of anthropological research. Rather the setting of this research occurs in a

multitude of locations. The primary entree into the "field" begins at one of the buildings

of the campus of the community college for the DHP women, or at the family practice

clinic for the physicians. The community college is where the displaced homemakers

program meets for three-week sessions every other month.


Table 3-2. Mean Age of DHP Participants, Years in Union and Household Composition.
Age at interview Mean (SD) Range
Total 52.23 (5.31) 45-68
White/Euro-American 52.48 (5.59) 45-68
Black/African-American 51.95 (5.24) 47-65
Other 52 (NA) NA

Years in marital union 18.05 (9.58) 1-34
Household composition 1.97 (0.99) 1-4












Table 3-3. Frequencies and Percentages of DHP Women Sample (n=44)
Self-Defined Ethnicity Frequency Percentage
White/Euro-American 23 52.3
Black/African-American 20 45.5
Other 1 (Middle East Nationality) 2.3
Latino Identity
White-Puerto Rican 1 2.3
Non-Latino 43 97.7
Age at interview
45-49 17 38.6
50-55 17 38.6
56-60 7 15.9
61-70 3 6.8
Marital Status Total White Black Total White Black
Divorced 22 (1"other") 15 6 50 65.2 30
Widowed 11 4 7 25 17.4 35
Separated 9 4 5 20.5 17.4 25
Ended Consensual Union 2 0 2 4.5 0 10
Educational level Total White Black Total White Black
Some High school 5 1 4 11.4 4.3 20
High school/GED 8 7 1 18.1 30.4 5
Vocational ED/college 23 (l"other") 10 12 52.3 43.5 60
College grad/grad school 8 5 3 18.2 21.7 15
Household composition Total White Black Total White Black
Live alone 17 (1"other") 10 6 38.6 43.5 30
2 in household 16 10 6 36.4 43.5 30
3 in household 6 1 5 13.6 4.2 25
4 in household 5 2 3 11.4 8.7 15
Self-rated health status Total White Black Total White Black
Excellent 9 (1"other") 6 2 20.5 26.1 10
Good 19 9 10 43.2 39.1 50
Average 8 5 3 18.2 21.7 15
Fair 5 2 3 11.4 8.7 15
Poor 3 1 2 6.8 4.3 10
Employment status
Full-time 13 29.5
Part-time 6 13.6
Retired 1 2.3
Unemployed 18 40.9
SSI 5 11.4
Other 1 2.3
Time lived in N.E. FL
Entire life 13 29.5
20 yrs. or more 13 29.5
5-15 years 8 18.2
Less than 5 years 10 22.7



Most of the physicians were recruited through a community health and family

practice clinic of the university-based hospital. The behavioral science coordinator on









staff there facilitated this recruitment, by introducing me to attending physicians and

giving me access to the noon meetings for the residents. I was unable to duplicate this

useful relationship in any of the other residency programs in the area. A few other

physicians were recruited through an email announcement that went to primary care

physicians in multiple specialties at the hospital, but this sort of recruitment strategy was

of very limited utility and cancellations were much more frequent with those that

expressed interest. Interviews with physicians occurred in their office or clinic.

The physician sample consisted of 13 attending physicians, and 18 resident

physicians. Naturally, attending physicians were, on average, older than residents, but

they were also more predominately males. Women have begun to predominate in the

enrollment in the residency program in recent years, and my sample reflects this trend

somewhat with 10 women and 8 men in the residency portion of the sample. There has

been a trend of increasing numbers of women medical doctors since the 1970s and 1980s,

but the earlier male dominance in the profession is seen in the attending physician portion

of the sample with only four women attending physicians. Additionally the female

participation in my study was low compared to the female-male ratio in the overall

residency program, which was just slightly less than 2:1 for 2003 and 2004. (Tables 3-4

and 3-5 show characteristics of the physician participants).

Most of the DHP meetings took place in a room located at the end of the hall on the

second or third floor, where slight ramp leads up to the door of the end room. A small

vertical rectangular window in the door reveals a large U-shaped table arrangement, and

the diversity of women seated at the bottom of the U, who turn to see that me peering

through the window in the door. The room is bright and large with windows flanking the










South and East sides of the room. Invited in by the program coordinator, I would sit

quietly off to the side of the table waiting for a break in the program.

Table 3-4. Frequencies and Percentages of Physician Sample (n=31)
Full interview or Survey Frequency Percentage
interview/survey 18 58.1
Partial interview/survey 1 3.2
Survey only 12 38.7
Attending/Resident
Attending Physician 13 41.9
Resident Physician 18 58.1
Resident Physician Detail
1st year resident 8 25.8
2nd year resident 7 22.6
3rd year resident 3 9.7
Sex/Gender
Female 14 45.2
Male 17 54.8
Self-reported Ethnicity
White/Euro-American 17 54.8
Black/African-American 2 6.5
Asian/Pacific Islander 7 22.6
Other 5 16.1
Latino Ethnicity
Cuban 1 3.2
South American 2 6.4
Non-Latino 28 90.3
Other Ethnicity
Guyanese-West Indian 1 3.2
Hispanic (South American) 1 3.2
Venezuelan 1 3.2
Iranian 1 3.2
Pakistani 1 3.2

Table 3-5. Mean Age of Physicians and Year of Medical School Completion
Age at interview Mean (SD) Range
Total (N=31) 39.42 (8.71) 27-57
Attending (N=13) 47.62 (6.37) 39-57
Resident (N=18) 33.50 (4.09) 27-42

Year completed med school 1991.87 (9.64) 1972-2002

Once the program director or coordinator of the class found a place to take a break,

she introduced me by name and my UF affiliation. I passed a one-page description of the

project with my contact information around to all of the participants and then proceeded

to summarize the information on the handout by telling them the title of the project, what









they would need to do to participate and how long it will take, the age and marital status

criteria and that they would be compensated with a $20 grocery store gift-certificate.

Then I passed around a sheet of paper and asked that anyone that was interested in

participating give me their name and phone number. I always brought a copy of the

informed consent with me for anyone who wanted to read it over. I also let them know

that if they were uncertain at the moment that my contact info was on the handout and

they could contact me in the future if they wanted to. This only happened once; usually I

had to call the participants to find a time to meet.

Most of the women attending the displaced homemaker program were suffering

either emotionally, financially, and/or spiritually, but they came to this free program to

get their lives back on track, or to learn new skills, or to simply improve their self-

esteem. I attended several of the class sessions, but each group differs from the one

before it, just like any teacher knows that every class has its own collective personality.

All of the women I spoke with said that they benefited from the program in some way.

Once in a while, when I visited the program to recruit new participants, I would get

a very inquisitive group that asked lots of questions, like why I am doing the research,

why the specific ages, what will I do with my degree once I have it and so on. Other

times, there would be no questions and only one person signed-up. I found it useful when

the timing worked to my advantage to ask if anyone would like to schedule a time to

meet right then, though this was not always appropriate and I tried very hard not to

interfere with the program.

The field setting changed every time I scheduled an interview. Scheduling

interviews often proved difficult since many were not available when I called and several









never returned my phone calls. This was true for both DHP women and physicians.

Sometimes I visited the homes for the participants, other times we met at the college or

some other location. The physicians were all interviewed in their offices.

The homes I visited in this research encompassed a range of economic levels and

the locations stretched to the outer edges of the service area of the displaced homemaker

program. Most of the participants live in urban or urban-suburban areas whereas only a

handful lived in more rural areas of the region. In some cases I visited homes in up-scale

gated communities, others lived in very modest homes and apartments. Sometimes the

homes were spotless, and it was obvious that the occupants had taken great care in

decorating their home. Other times the homes were in total disarray (perhaps a testament

to the difficult transition some of the women were undergoing) and I was instructed to

climb over boxes to get to a seating area. However, not all of the interviews took place at

participants homes; I met one woman at her place of work on her lunch break, and about

a third of the DHP participants chose to meet me in the building where the DHP classes

took place. This was convenient for both of us since we knew the place and the kind

people at the community college helped us find a quiet place to talk.

Unlike the interviews with the DHP women, the physician interview location was

fairly static. I interviewed all of the physicians at their place of work, whether this was

their clinic office, hospital office, or in the case of the residents, in a quiet meeting room

at the clinic or the office of the behavioral science coordinator that was assisting me. As

for the physicians that only completed the survey and CCM procedure, I met with them

when they had a few minutes between meetings or patients at the clinic. I also was given

access to 20 minutes of one of the noon meetings and convinced six residents to complete









the survey and the rating procedure. It was not uncommon for the physician to delay or

cancel our scheduled meetings. I made several trips to the clinic and hospital that yielded

little more than a chance to reschedule.

Most of the physicians in my study were residents or attending physicians in a

family practice clinic that is in a largely African-American neighborhood. The building

consists of two wings, one side is the clinic itself and the other is where the residency

program resides. The clinic itself opens with the reception staff to the right and with a

large waiting area to the right. The waiting area is stocked with countless brochures on

every conceivable health issue. There are two TVs and a play area filled with toys for

children. On my first visit there, the coordinator who was assisting me gave me a grand

tour and introduced me to several physicians. Most of the physician offices are located

on the residency side, and this was my primary point of entree.

The interview process typically followed the same pattern for both DHP women

and physicians. For the DHP women, I gave them the informed consent to read and sign,

explaining any elements that they needed clarified. I always offered to read it to them if

that was their preference, to save them any embarrassment if they had difficulty reading.

Once the informed consent was taken care of, I proceeded to the survey, which usually

only took a couple of minutes, then I began the vignette-based interview. Before reading

the first story, I made clear that the stories were fictional, that there were no right or

wrong answers to the questions that followed the stories, and that they could make their

responses as personal or impersonal as they wished. All of the interviews were audio

taped with the participants' permission, and initialed on the informed consent. The

interview section of the encounter usually took from 30 minutes to an hour. One of the