The Relationship between medical residents' attitudes toward psychosocial issues in patient care and psychosocial interv...

MISSING IMAGE

Material Information

Title:
The Relationship between medical residents' attitudes toward psychosocial issues in patient care and psychosocial interview skill level following communication training programs
Physical Description:
viii, 82 leaves : ; 29 cm.
Language:
English
Creator:
Behen, Joseph Matthew, 1967-
Publication Date:

Subjects

Subjects / Keywords:
Internship and Residency   ( mesh )
Physician-Patient Relations   ( mesh )
Communication   ( mesh )
Attitude of Health Personnel   ( mesh )
Education, Continuing   ( mesh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 72-81).
Statement of Responsibility:
by Joseph Matthew Behen.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
oclc - 50514854
ocm50514854
System ID:
AA00011215:00001

Full Text











THE RELATIONSHIP BETWEEN MEDICAL RESIDENTS' ATTITUDES
TOWARD PSYCHOSOCIAL ISSUES IN PATIENT CARE
AND PSYCHOSOCIAL INTERVIEW SKILL LEVEL
FOLLOWING COMMUNICATION TRAINING PROGRAMS

















By

JOSEPH M. BEEN
















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

1994














To the Behen family














ACKNOWLEDGEMENTS


I would like to acknowledge Nathan Perry and Stephen

Boggs for serving as the cochairmen of this project, and the

three other committee members, James Rodrigue, Cynthia

Belar, and Lynn Romrell, for their assistance during the

course of this project. I also wish to thank Michael

Geisser for his help, although he was not able to stay until

the end. Particular gratitude goes out to Steve for the

well-balanced and important direction and support he offered

as my acting chairman. Additionally, I would like to

acknowledge my deep appreciation to Jim for the consistent

and valuable guidance which he provided throughout my

graduate experience. Finally, I wish to express my

appreciation to the Arthur Vining Davis Foundation for

providing the financial support that allowed this study to

take place.















TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS ....................................iii

ABSTRACT .............................................vi

CHAPTERS

1 REVIEW OF LITERATURE ............................1

Introduction ....................................1
Overview of Physician-Patient Relations.........2
Physician-Patient Relations:
Empirical Findings ...........................5
Communication Training Programs ................10
Attitude Theory................................12
Attitude Definition......................12
Attitude Formation .......................15
Attitudes and Behavior...................16
Attitude Assessment......................17
Attitudes Toward Psychosocial Issues
in Patient Care: Empirical Studies.........18
Methodological Considerations...................24
Proposed Model.................................25
Specific Aims and Hypotheses...................27

2 METHOD .........................................31

Subjects..................................... ... 31
Materials..................... ....................32
Training Program...............................36
Design and Procedure...........................40

3 RESULTS .........................................42

Participants versus Nonparticipants............43
ANOVAS and ANCOVAS .............................44
Correlation and Regression Analyses............48

4 DISCUSSION .....................................51
















APPENDICES ................................... .......70

A CORRELATION MATRIX BY RATER .................... 70

B BELIEF SCORES BY SPECIALTY..................... 71

REFERENCES ...........................................72

BIOGRAPHICAL SKETCH .................................. 82














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

THE RELATIONSHIP BETWEEN MEDICAL RESIDENTS' ATTITUDES
TOWARD PSYCHOSOCIAL ISSUES IN PATIENT CARE
AND PSYCHOSOCIAL INTERVIEW SKILL LEVEL
FOLLOWING COMMUNICATION TRAINING PROGRAMS

By

JOSEPH M. BEEN

December, 1994

Chairman: Nathan Perry
Cochairman: Stephen Boggs
Major Department: Clinical & Health Psychology

Central to the task of successful patient medical care

is the quality of the relationship between physician and

patient. A growing body of research has demonstrated that

effective physician-patient communication is a goal that can

be achieved through the development of teachable skills.

The present study sought to examine the role of a variable

thought to be important in the learning of clinical

communication skills, specifically, physician attitudes

toward psychosocial issues in patient care.

Based on recommendations from social psychological

research on the attitude construct, the proposed study

examined two components of physicians' psychosocial

attitudes, beliefs toward psychosocial issues in patient

care and behavioral intentions to carry out the behaviors

vi








taught in communication training programs. The specific

aims of the study were to assess the extent to which medical

residents' psychosocial attitude components may change as a

result of communication training and the extent to which any

changes are maintained over time, and to clarify the

relationship between residents' psychosocial beliefs and

their behavioral intentions and interview skill level

following training in communication skills.

Sixty-seven medical residents were randomly assigned to

one of two communication training groups or a waitlist

control group. Beliefs and behavioral intentions were

assessed pre- and post-training and at follow-up. Interview

skill was assessed following communication training. The

results indicated that the components of residents'

attitudes (beliefs and behavioral intentions) toward

psychosocial issues in patient care did change as the result

of brief psychosocially-oriented training interventions

compared to a control group, and that the initial changes in

residents' attitude components were maintained at follow-up

assessment. The results also revealed that the attitude

components did not relate to ability to demonstrate

interview skill during a simulated patient interview.

The findings have implications for the training of

medical residents in interviewing skills. Specifically,

residency educators are likely to find that communication

training programs with the goal of improving residents'

interviewing skills may be most effective when the emphasis

vii








is concurrently on both teaching residents specific skills

involved in acquiring psychosocial information efficiently

and accurately and emphasizing to residents the value of

psychosocial information in patient care.


viii















CHAPTER 1
REVIEW OF LITERATURE

Introduction

Central to the task of successful patient medical care

is the quality of the relationship between physician and

patient. A growing body of research has demonstrated that

effective physician-patient communication and improved

physician-patient relations are both goals that can be

achieved through the development of teachable skills.

However, despite the widespread acceptance that

communication problems are common in medical practice and

that the quality of clinical communication significantly

impacts health outcomes, traditional medical education has

proven generally ineffective in teaching clinical

communication. This state of affairs calls for research to

clarify the important variables involved in effective

teaching of physician-patient communication skills.

The following review addresses the role of one such

variable, physician attitudes toward psychosocial issues in

patient care, as it may impact skill acquisition during

communication training. Clarification of the role of this

variable in physician training has implications both for the

development of efficacious training methods and the

selection of individuals for training opportunities. What








follows is, first, a literature review of physician-patient

relations, communication training programs, relevant

attitude theory, and empirical research which has

investigated physician attitudes toward psychosocial issues

in patient care. Second, a study is proposed to investigate

the role of such attitudes in skill acquisition during

communication training.

Overview of Physician-Patient Relations

The nature of the relationship between physicians and

their patients has undergone extensive change in the last

century. Currently, this relationship continues to evolve

as the communicative processes between physicians and

patients increasingly become a topic of interest and

investigation.

Stoeckle and Billings (1987) have reviewed the process

by which the physician-patient relationship has been

gradually transformed by reforms within and outside the

medical profession during this century. According to their

review, obtaining the medical history has progressed from a

discourse based primarily on interrogation about the

patient's bodily feelings and medical facts as exclusive

data needed in the search for disease. Reforms from

psychiatry during the 1940s moved the relationship from

simply fact-finding and history-taking to what could now be

appropriately referred to as the medical interview. This

change involved obtaining information that could be used in

psychological treatment by the use of nondirective








questioning and explicitly including assessment of the

patient's personality, emotional reactions, and inter-

personal conflicts as part of the interview. During the

1970s behavioral and communication scientists, including

clinical and social psychologists, helped transform

physician-patient communication into an interaction that was

even more explicitly psychotherapeutic, educational, and

participatory. This was accomplished by teaching commun-

ication skills such as empathic responding to providers,

attending to the factors involved in the communication of

information between physician and patient, and highlighting

the reciprocal nature of the communication process.

The evolution in physician-patient communication within

the medical interview has resulted in common agreement about

what the modern physician-patient relationship should

entail. Educational texts, professional behaviors, and

social attitudes each reflect how the current ideology of

the relationship may be characterized (Stoeckle & Billings,

1987). The elements of the modern relationship include

allowing the relationship to be more democratic by giving

patients choices in decisions about the scope of diagnosis

and the alternatives of treatment, developing patient

participation in the relationship by transmitting

appropriate information about illness and treatment which

enables patients to make choices, and negotiating with

patients about their requests and choices (Lazare,

Eisenthal, Frank, & Stoeckle, 1976). It also involves








attending to patients' feelings about illnesses and

treatment, and responding to their feelings with positive

regard, genuine concern, and empathy; providing helping

actions by eliciting, acknowledging, and responding to

patients' own perspectives of their illness and care;

conveying respect to patients without regard to their class,

gender, race, ethnicity, or age; promoting health education,

self-help, and preventive behaviors by communicating

information about diagnosis, treatment, health maintenance,

and prevention; and being self-reflective in the acts of

questioning, listening, and talking in order to modulate the

effect of the doctor's own feelings and prejudices and to

reframe inquiries and responses to patients (Schon, 1983).

The current ideology of the physician-patient

relationship has impacted medical interviewing approaches.

Bird and Cohen-Cole (1990), Lazare (1989), and Lazare et al.

(1990) have described a three-function approach to medical

interviewing. These interview functions represent an

operationalized application of a biopsychosocial approach

designed to facilitate physician-patient communication

processes. They suggest that the skilled physician must be

able to accomplish several different objectives. The first

function is to gather data to understand patients' problems.

This involves collecting accurate information in an

efficient manner. The second function is to develop rapport

and respond to the patient's emotions. Demonstration of

sufficient emotional support serves to relieve acute








distress and facilitate the development of trust. The third

function is the education of patients about their illness

and the development of patient motivation to adhere to

treatment plans.

Physician-Patient Relations: Emoirical Findings

Findings from reports investigating physician-patient

relations suggest that the quality of current physician-

patient interactions is often poor. Increasing public

dissatisfaction with the medical profession is significantly

related to deficiencies in the clinical communication

(Simpson et al., 1991). It has been found that most

complaints by patients deal with communication problems

(Richards, 1990).

Several studies have shown that patients' visits are

usually physician-centered rather than patient-centered.

Platt (1979) referred to the "high control style" in which

the physician asks many questions and maintains control over

the interaction by not letting the patient speak at any

length. Waitzkin (1984) noted that physicians often

maintain a style of high control which involves many

physician-initiated questions, interruptions, and neglect of

patients' life circumstances. Research has suggested that

language used by physicians is unclear to patients, both in

terms of the use of jargon and in relation to a lack of the

expected shared meanings of relatively common terms

(Simpson, 1980).








Research has demonstrated that patients' concerns are

frequently not addressed. Beckman and Frankel (1984) noted

that physicians play an active role in regulating the

quantity of information elicited at the beginning of the

clinical encounter, and use closed-ended questioning to

control the discourse. They reported that of 74 office

visits studied, in only 23% was the patient provided the

opportunity to complete his or her opening statement of

concerns. In 69% of the visits the physician interrupted

the patient's statement and directed questions toward a

specific concern. They suggest the consequence of this

controlling style is the premature interruption of patients,

resulting in the loss of relevant information. Frankel

(1984) found that 94% of physician interruptions resulted in

the physician's taking charge of the conversation. Burack

and Carpenter (1983) found that such an approach prevented

physicians from learning all but 6% of primary problems that

were ultimately determined to be psychosocial.

The educational needs of patients have been found to be

unmet during physician-patient interactions. Fletcher

(1980) reported that up to 60% of patients have been found

to be dissatisfied with information provided by physicians.

Waitzkin (1984) suggested that physicians tend to

underestimate patients' desire for information and

misperceive the process of information-giving, resulting in

a low proportion of visits with physicians that includes

patient education.








During recent decades, contributions from the social,

behavioral, and communication sciences have improved the

diagnostic, educational, and therapeutic potential of the

medical interview (Stoeckle & Billings, 1987). Overall,

studies have found the quality of physician-patient

interactions to be related to the quality of patient care

and positive health outcomes. Specifically, effective

interviewing skills have been associated with a variety of

positive effects.

Improved physician-patient relations has been

demonstrated to increase patient satisfaction. In a meta-

analysis of correlates of provider behavior in medical

encounters, Hall et al. (1988) found that satisfaction had

the most consistent relation to physician behavior.

Satisfaction was found to be related to information giving,

technical and interpersonal competence, partnership

building, immediate and positive nonverbal behavior, social

conversation, positive talk, and more communication overall.

Enhanced physician-patient communication has been found

to result in improved accuracy of diagnosis. A series of

studies by Cox and his colleagues (1981) suggested that

physician use of open-ended questions and allowing patients

to express their concerns resulted in better quality factual

information on which to base diagnostic decisions. Goldberg

(1980, 1982) found that primary care physicians trained in

interview skills using videotape feedback were better able

to make accurate diagnoses. Roter and Hall (1987) found








greater use of open-ended questions to be related to

increased patient disclosure of medical information.

Kaplan, Lipkin, and Gordon (1988) argued that establishing a

therapeutic relationship with patients is critical to both

diagnosis and treatment.

Patient compliance has been shown to increase as a

result of improved physician-patient communication. Hall's

(1988) meta-analysis demonstrated that compliance was

associated with information provision by physician, fewer

physician questions overall (but more questions about

compliance in particular), more positive talk, and less

negative talk. Additional evidence suggested compliance was

increased when providers took a more dominant role.

Effective physician interviewing skills have been

demonstrated to have beneficial effects on patient

education. Hall's (1988) meta-analysis suggested that

patient recall and understanding were best predicted by

information giving, less question asking, more partnership

building, and more positive talk. Other researchers report

that more recall and understanding are also related to

technical competence (Roter, Hall, & Katz, 1987),

interpersonal competence (Bartlett, Grayson, & Barker,

(1984), fewer direct orders (Heszen-Klemens & Lapinska,

1984), more immediate nonverbal behaviors (Larsen & Smith,

1981), and less negative talk (Carter, Inui, & Kukall,

1982).








Enhanced physician-patient communication has been shown

to influence favorable health outcomes. Greenfield, Kaplan,

and Ware (1985) demonstrated that an intervention which

increased patient involvement in the interaction with the

physician resulted in fewer limitations imposed by disease

on patients' functional ability, and also increased

preference for active involvement in medical decision

making. Kaplan (1989) found that "better health," measured

physiologically, behaviorally, and subjectively, was

consistently related to specific aspects of physician-

patient communication, including more patient control,

expression of affect, and information provision. Orth,

Stiles, Scherwitz, and Hennrikus (1987) demonstrated that

reduction in blood pressure was significantly greater in

patients who, during visits to the doctor, had been allowed

to express their health concerns without interruptions.

Agreement between physician and patient in identifying the

nature and seriousness of the clinical problem has been

related to improving or resolving the problem (Bass,

Garland, & Otto, 1986).

Improved physician-patient interaction has also been

found to have positive benefits for physicians. Researchers

have reported that improving physician-patient relations

results in making medical practice more professionally

rewarding (Almy, 1989; Gerber, 1987) and in a decreased

number of malpractice suits (Altcheck, 1988).








Professional medical organizations have affirmed that

the teaching of interviewing skills and the psychosocial

content of patient care are essential aspects of physician

training. The American Board of Internal Medicine (1983)

issued a statement describing and outlining the humanistic

qualities required of the internist and guidelines for their

evaluation in practice. The Society of General Internal

Medicine, through its Task Force on the Medical Interview

(Lipkin, Quill, & Napodano, 1984), published a model

curriculum for teaching interviewing skills to internal

medicine residents. Other medical organizations calling for

such training include the American Board of Family Practice,

the Association of American Medical Colleges, and the

American Board of Pediatrics (1987).

Communication Training Proarams

In the past fifteen years there has been a

proliferation of efforts to teach interviewing skills and

the psychosocial dimension of medical practice to medical

students and residents. Among programs that have been

developed, there exists a considerable amount of variation

in content, quantity, teaching methods, organization,

faculty responsible for training, financial support,

training setting, and evaluation. Overall, the methods that

have been used have been shown to improve trainee

performance (Kern et al., 1989).

The quality of evaluations of training programs within

the literature is highly variable. Problems with








evaluations include the failure to use objective measures of

training effectiveness, not keeping trainees blind to

evaluations, failure to use a control group, insufficient

reliability and validity of trainee performance assessment

measures, and lack of generalizability to other settings.

Kern et al. (1989) reviewed training programs which

were randomized controlled evaluations in order to identify

which educational methods demonstrated effectiveness. He

suggested that among those investigations yielding positive

results, the most commonly used educational method was

feedback based on a trainee's performance. Feedback was

typically used in these programs in combination with

videotape reviews, audiotape reviews, live observations of

role playing, or trainee interactions with real or simulated

patients. Supplemental training methods often included

discussion, didactic sessions, provision of readings or

syllabus, role modeling, or videotaped examples.

Research has evaluated the impact of training on

patient care outcomes. Four randomized controlled studies

have demonstrated effectiveness of training programs with

respect to patient variables. Goldberg et al. (1980) found

that training improved the ability of physicians to assess

accurately their patients' psychologic distress. Evans and

colleagues (1987) reported that satisfaction was greater and

anxiety less among patients of trained physicians. Cope and

coworkers (1986) found patient satisfaction to have improved

with physician training. Roter et al. (1990) found trained








residents asked more open-ended questions and fewer leading

questions, summarized main points more frequently, did more

psychosocial counseling, and were rated as having better

communication skills than untrained residents. The use of

more focused and psychosocially directed questions and fewer

leading questions was associated with more accurate

diagnosis and management recorded in the medical chart.

Attitude Theory
The attitude construct has a long and influential

history in the field of psychology, particularly social

psychology. Unfortunately, the study of attitudes in the

health professions has utilized this research only to a

limited extent. In order to avoid following this trend,

relevant issues and background concerning the attitude

construct is provided.

Attitude Definition

The study of attitudes and attitudinal processes has

long preoccupied social scientists. Despite the concept's

influential history, consensus on precisely what an attitude

is and how it can be identified has proven to be somewhat

elusive (McGuire, 1985). However, among researchers on

human attitudes, there is agreement generally that an

attitude or combination of attitudes involves a set or state

of readiness or willingness that predisposes an individual

to respond with certain verbal and/or other behaviors under

certain conditions; and that attitudes are learned, and

relatively enduring (Tinker, 1991). The most prominent








feature of an attitude is its evaluative character, the

disposition to respond toward an object in a positive or

negative manner. Attitudes can thus range from very

favorable to very unfavorable on an evaluative continuum

(Ostrum, 1984).

Various attempts have been made to conceptualize the

components of human attitudes, and to develop theory and

research about these factors. Rosenberg (1965) popularized

the view that attitudes are composed of three classes of

response (affective, cognitive, and behavioral) to a

stimulus, or attitude, object. They suggested that an

attitude consists of how we feel, what we think, and what we

are inclined to do about an attitude object. Azjen (1984)

noted that attitude is a hypothetical, unobservable

construct which must be inferred from measurable responses

that reflect evaluations of an attitude object. He

distinguished between three categories of such responses,

including cognitive, affective, and conative responses.

Cognitive responses are beliefs that reflect the individual's

perception of, and information about, the attitude object.

Affective responses are evaluations of, and feelings toward,

the object. Conative responses are behavioral intentions,

tendencies, and actions with respect to the object.

Zimbardo and Leippe (1991) emphasized that the inter-

connectedness of attitudes, which are typically the most

important component of attitude systems, cognitions, affect,

behavioral intentions, and behaviors into organized systems








allows for changes in one facet of the system to often cause

changes in other facets.

A competing approach to attitude definition focuses on

conceptualizing attitudes as consisting of a single

component. Researchers have approached this issue from

divergent perspectives, each of which emphasizes either

affect, cognition, or behavior. An early definition by

Thurstone (Thurstone & Chave, 1929) suggested that attitudes

consist of evaluative or affective responses to attitude

objects. Fishbeck and Azjen (1975) have popularized this

one-component view by proposing that affective responses are

based upon cognition, specifically beliefs. Zajonc (1980)

proposed that affect can be the basis for one's attitudes

and preferences. Bem (1972) has emphasized that attitudes

are often inferred from past behaviors, taking into account

the conditions under which the behavior was performed.

Thus, the literature on attitudes is marked by two

conflicting definitions. Zanna and Rempel (1988) have

attempted to reconceptualize the attitude construct to

address these definitional issues. They view their

reconceptualization as an attempt to take existing notions

in the literature and combine them in a new and

heuristically useful way. They define attitude as the

categorization of a stimulus object along an evaluative

dimension. They propose that such evaluation can be based

upon three classes of information: 1) cognitive

information, 2) affective information, 3) information








concerning past behaviors. They suggest that these classes

of information can determine evaluations separately or in

combination. An advantage of this model is that it

specifies conditions for when a particular class of

information may be a more relevant focus of study.

Specifically, it suggests that when evaluations are based on

beliefs about the attitude object, the model should be

reduced to a one-component formulation such as that proposed

by Fishbein and Ajzen (1975). When evaluations are based

primarily on affects associated with the attitude object,

the model should resemble the one-component formulation of

Zajonc (1980). Finally, when evaluations are based on

inferences from past behavior, the model should be similar

to that proposed by Bem (1972).

Attitude Formation

Azjen (1984) has identified three separate approaches

social scientists have taken to studying attitude formation.

The earliest of these is a functional approach which

attempted to identify the origins of attitudes based on the

needs or functions that attitudes serve. Functions which

attitudes were assumed to provide include instrumental

(allowing individuals to attain rewards and avoid

punishments), knowledge (serving to organize and simplify

one's experience), expressive (enabling emotional release),

and ego-defensive (protecting and enhancing the self). The

major limitation of this approach is argued to be its

circularity of reasoning in that attitude related needs are








inferred from the attitudes people are known to hold, and

then these inferred needs are used to explain the observed

attitudes.

The second approach to understanding attitude formation

is followed by behaviorally oriented scientists. This

approach bases attitude development on principles of

classical conditioning. Specifically, repeated and

systematic association between the attitude, or conditioned

stimulus, and a positively or negatively valued event, the

unconditioned stimulus, is assumed to produce a favorable or

unfavorable affective reaction, or attitude, to the object.

The third approach toward attitude formation, the

cognitive approach, has followed the general trend toward

cognitive, information-processing explanations of social

behavior. This approach stresses the role of information as

a basis of attitude formation. According to this view,

beliefs, which represent one's subjective knowledge of

oneself and his world, are the primary determinants of

attitudes. Generally, the ratio of the number of beliefs

which associate the object with positive attributes to those

which associate the object to negative attributes determines

the resultant attitude.

Attitudes and Behavior

The attitude construct has typically been used to

explain social behavior. Given that attitudes are

considered behavioral dispositions, it follows that they are

assumed to direct, and to some extent, determine action.








However, it was not until the 1960's that the assumption

that attitudes can be used to predict and explain behavior

was empirically challenged. In his review of the research

exploring this issue, McGuire (1985) asserted that only

within limited circumstances do attitudes account for more

than 10% percent of behavioral variance.

Azjen (1984) has reached a more optimistic conclusion

regarding the relationship between attitudes and behavior.

He noted that as a result of the negative evidence about the

relationship, researchers were forced to reexamine the

nature of attitude and its relation to behavior. He

suggested that renewed recognition that attitude is an

unobservable, hypothetical construct which must be inferred

from measurable responses to the attitude object resulted in

increased understanding. He suggested that as attitudes and

actions are both expressions of an underlying disposition, a

strong empirical relation can only be expected if

measurements of each assess exactly this common disposition.

Therefore, he concluded that general expressions of

attitudes are found to be strongly related to aggregate

measures of behavior when such is the case, and attitudes

are usually found to be quite accurate predictors of

subsequent actions.

Attitude Assessment

Davis and Ostrom (1984) have presented theoretical

issues to consider when assessing attitudes. They suggest

that in carrying out an attitude assessment the researcher








must specify the following: First, the attitude object;

second, the conceptual attributes of the attitude construct

relevant to the aims of the research; and third, the

response domain. The attitude object refers to any of a

variety of behaviors, ideas, concepts or entities which may

be specific and concrete, refer to a social category, or be

broad and abstract. The attitude construct refers to the

operational definition of attitude on which the study

relies. The response domain refers to the three categories

of response identified by attitude researchers: affective

responses which refer to feelings and physiological

reactions one has to the attitude object, cognitive

responses including information, beliefs, and inferences

made about the object, and conative responses which refer to

the behaviors one initiates or intends to initiate in regard

to the attitude object.


Attitudes Toward Psychosocial Issues in Patient Care:
Empirical Studies

During the past forty years attitudes of physicians in

training and the extent to which they change over the course

of the educational experience has been investigated. A

common purpose of these studies has been to understand those

characteristics which predispose to effective medical

practice.

Early psychological investigations focused on whether

medical students' psychosocial attitudes changed during the

course of medical school. These studies described the








presence or absence of attitude change generally through the

use of standardized multidimensional personality inventories

such as the Jackson Personality Inventory (Rosenberg, 1965)

and the Minnesota Multiphasic Personality Inventory (Glaser,

1951). Student attitudes on authoritarianism, cynicism,

dogmatism, compassion, and humanitarianism were of

particular interest in these studies. Results typically

indicated that medical school had an adverse effect on

student attitudes. Studies suggested that students tended

to increase in cynicism and decrease in humanitarianism

during their medical school experience (Eron, 1958; Christie

& Merton, 1958; Gordon & Mensh, 1962). Other research

indicated that although such change was notable, the

students' attitudes were situational in nature and developed

as a reaction to the medical school environment (Becker &

Geer 1958; Fox, 1957; Reinhardt & Gray, 1972). Attempts

were also made at this time to influence students' attitudes

in more positive directions through more comprehensive

training programs which exposed students to psychosocial

issues (Merton, Reader, & Kendall, 1957; Reader & Gross,

1957; Hammond & Kern, 1959). Findings of these programs

suggested that attitude changes which resulted were short-

lived. Overall, early research on student attitudes

suggested that medical school contributed to the development

of negative attitudes toward psychosocial issues in students

and that attempts to reverse this trend were typically

unsuccessful (Rezler, 1974).








More recent investigations have continued to examine

the attitudes of medical students. These studies differ

from their predecessors in that they tend to use attitude

scales rather than personality tests. This approach is

based on the idea that valid measurement requires test items

which are specific to the relevant environment (Parlow &

Rothman, 1974). Additionally, this research has addressed

issues including whether attitudes change with certain

interventions, how attitudes differ between various medical

and health professionals, and what correlates exist for

different types of attitudes.

Research investigating whether interventions during

training can change medical students' attitudes has provided

equivocal results. Blizek and Finkler (1977) reported no

changes in medical student attitudes towards human values

after a course on moral problems of medicine. Markham

(1979) failed to find changes in medical students attitudes

toward patients and physician-patient relations after a

behavioral science course as initially negative view were

maintained. However, Maisiak et al. (1980) found that

first-year medical students attitudes toward the social and

behavioral determinants of patient health, although

initially positive, did improve after a behavioral science

course. The results of the study suggest that such a course

affects both students' attitudes and knowledge toward

psychosocial issues. Dornbush (1984) reported that initial

attitudes of medical students were positive, and that these








remained positive after an in-depth clinically and

humanistically oriented course in the behavioral sciences.

In a follow-up study Dornbush (1985) reported that initially

positive attitudes of medical students toward psychosocial

issues had been maintained during the course of medical

school. Zeldow (1987a) reported evidence that students who

enter medical school with adequate levels of compassion and

nurturance leave with these qualities intact, while those

who enter deficient in these qualities complete their

medical school training no more patient-oriented than when

they matriculated. Dwyer, Detweiler, & Kosch (1988) found

psychosocial attitude scores to improve pre- to post-

intervention with third-year medical students following a

four week family practice rotation. De Monchy (1988) found

medical students in their final year of study to be more

patient-centered than students at earlier stages of

training.

Fewer studies of psychosocial attitudes and attitude

change have taken place at the residency level of training.

Shapiro (1991) studied changes in self-assessed resident

psychosocial attitudes and behavior after participation in a

month-long behavioral science rotation. The findings

suggested that certain basic psychosocial assessment skills,

especially those that residents initially evaluated

themselves as lacking, could be successfully taught through

a structured format. However, lack of clinically

significant improvement in terms of attitude change








suggested the limitation of using this format for such a

purpose. Yet, it was evident that residents assessed

themselves as having an initially positive orientation

toward psychosocial issues in medicine. Smith et al. (1991)

reported that an intensive, comprehensive month-long

training program which focused on communication and

relationship-building skills with first-year residents

resulted in improved attitudes toward psychosocial issues

toward patient care compared to a control group. Rich

(1987) reported that residents had initially positive

attitudes toward physician-patient relations at the onset of

residency, and that this remained over the course of the

entire residency.

The attitudes of physicians-in-training and practicing

physicians toward psychosocial aspects of patient care have

been compared to other health. professionals. Medical

students appear to be more similar to dentistry and pharmacy

students and less similar to nursing and social work

students with respect to attitudes toward doctor-patient

relations (Parlow & Rothman, 1974). Medical students have

been found to demonstrate equal or greater concern on

attitudes to issues of prevention and doctor-patient

relationships compared to other student groups (Ewan, 1987).

Different medical specialties have been found to vary in

their nurturant-empathic attitudes toward patients with

psychiatric residents having the most positive attitudes,

followed by pediatric residents, and then surgical residents








(Roskin & Marell, 1988). Medical specialties have also been

found to differ with regard to attitudes toward the

importance of psychosocial aspects of patient care.

Levinson, Kaufman, and Dunn (1990) found academic general

internists and psychiatrists to be significantly more

psychosocially oriented than practicing internists and

surgeons. Ashworth, Williamson, and Montano (1984) found

that out of four specialties assessed, psychiatrists were

the most psychosocially-oriented, followed by family

physicians, pediatricians, and internists.

Recent research has studied attitudinal correlates of

medical students. Streit (1980) found that in medical

students positive attitudes toward psychosocial issues in

medicine were negatively correlated with a relatively closed

value system and positively correlated with empathy,

sensitivity, and expedience. Zeldow (1987b) found that

medical students who scored high on positive attitudes

toward doctor-patient relations were liked and respected by

their classmates, knowledgeable about the behavioral

sciences, and described themselves as warm, caring, and

tolerant of ambiguity, while placing a premium on rapport

with patients. Low scorers were more concerned about

appearing calm and in control of their feelings and the

situation and emphasized emotional and interpersonal

detachment in their relations with patients. Dwyer et al.

(1988) found that female gender, planning to enter a primary








care specialty, and having a doctor-parent were correlated

with higher psychosocial attitude scores.

In summary, the research on physicians-in-training

attitudes toward psychosocial issues suggests that this

population's attitudes are generally positive, vary between

specialties and other professions, and have particular

correlates. Less clear is the extent to which attitudes may

change with psychosocially-oriented interventions, and

whether such changes are maintained over time. Also

unaddressed by this body of research is the relationship

between attitude and skill acquisition during

psychosocially-oriented training programs. However, while

these findings do indicate that physician attitudes are an

important topic upon which to focus, conclusions from this

body of research must be tempered as a result of

methodological limitations.

Methodological Considerations

While the aforementioned findings involving physicians-

in-training's psychosocial attitudes appear promising,

conclusive statements regarding the role of such attitudes

are not possible as a result of the various methodological

problems which characterize this area of study.

Specifically, research findings are limited by the following

problems: First, few studies have included the necessary

controls by using an experimental design. Research in this

area has been either quasi-experimental or correlational in

design. As a result, studies have been subject to the








threats to internal and external validity which accompany

those designs. Second, studies which directly assess

attitudes toward psychosocial issues using measures with

acceptable reliability and validity have been the exception

rather than the norm. A significant percentage of the

studies in this area have used the approach of developing

their own attitude measures. Recently, a measure, the

Physicians' Belief Scale, which objectively describes where

a physician falls on a dimension of acceptance versus

rejection of psychosocial tenets in medical care and which

has acceptable reliability and validity has been developed.

Finally, researchers have not been guided by the findings

from social psychological research regarding the attitude

construct which has accumulated over the past several

decades. A reconceptualization of the attitude construct by

Zanna and Rempel (1988) recently has emerged which

incorporates the main ideas of past conceptualizations in a

way that capitalizes on the strengths of several of the most

prominent, current models and provides a framework for

future research.

Proposed Model
Effective communication with patients requires the

physician to be able to accomplish several objectives during

an interview. The skilled physician must be able to collect

accurate information efficiently, demonstrate emotional

support, educate the patient, and encourage adherence.

Thus, the behavior of the physician during an interview has








various instrumental purposes. Research findings suggest

that when behavior was undertaken for instrumental purposes,

evaluations about attitude objects were best predicted by

beliefs about the attitude objects (Miller & Tessar, 1985).

As noted earlier in this review, Zanna and Rempel (1988)

recommend that when evaluations about an attitude object are

based on beliefs, the model should be reduced to a one-

component formulation such as that proposed by Fishbein and

Ajzen (1975).

The Fishbein-Ajzen model (1975) emerged in the mid-

1970's and has continued to guide attitude research since

that time. The foundation of their conceptual framework is

provided by their distinction between beliefs, attitudes,

intentions, and behaviors. The major concern of the

conceptual framework is the relationships between these

variables.

Beliefs serve as the fundamental construct in the

conceptual structure. Specifically, belief is a probability

judgement that links an object or concept to some attribute.

The sum of an individual's beliefs serves as the

informational base which ultimately determines the person's

attitudes, intentions, and behaviors.

In this model an attitude is a bipolar evaluative

judgement of the object. It is essentially a subjective

judgement of general evaluation or feeling of favorableness

or unfavorableness toward an object. Specifically, one's

attitude toward an object is based on his/her salient








beliefs about that object. An individual's attitude is

determined by his/her beliefs that the object has certain

attributes and by his/her evaluations of those attributes.

This framework suggests that a person's attitude is related

to the set of beliefs about the object but not necessarily

to any specific belief.

An intention is a probability judgement that links the

individual to some action. Intentions to perform a

particular behavior result from the individual's attitude

toward the behavior and the person's subjective norm

concerning the behavior. As with a belief, the strength of

an intention is indicated by the person's subjective

probability that he will perform the behavior in question.

The term behavior refers to observable acts of the subject

that are studied in their own right.

The current study sought to use two of the variables

suggested by the Fishbein-Ajzen model, beliefs and

behavioral intentions, and assess their relationship to

resident behavior following communication training programs.

Specifically, the beliefs were those involving psychosocial

issues in patient care. The behavioral intentions referred

to intentions to carry out the behaviors taught by the

programs.

Specific Aims and Hypotheses

This study sought to build upon the body of research

that exists regarding physicians' attitudes as well as to

address questions previously unaddressed empirically.








Despite the attention that has been paid to physicians'

attitudes toward psychosocial issues during their training,

the relationship between attitudes and skill development has

remained relatively unexamined. It seems likely that such

physician attitudes may be related to the development of

effective communication skills. Residents who believe that

psychosocial issues are important in patient care may be

more motivated to participate in training programs that

teach methods for enhancing competence in communication.

Residents may also more readily acquire the skills presented

in such programs. If residents believe that skill in

handling the psychological care of patients is a natural,

untrained ability possessed by individuals, they are

probably less likely to become engaged in learning

activities designed to bolster skills in this area

(Levinson, Kaufman, & Dunn, 1990). The specific aims of

this study were to assess the extent to which residents'

psychosocial attitudes changed as a result of communication

training and to clarify the relationship between resident

physician attitudes toward psychosocial issues in patient

care and interview skill level following training in

communication skills. Additional aims of the study were to

evaluate whether residents who volunteered to participate in

the communication training programs differed from those who

did not and to assess the maintenance of attitudes at

follow-up.








The specific hypotheses for the study were the

following:

1) Residents who participated in the communication skills

training programs would differ in their beliefs toward

psychosocial issues in patient care from residents who did

not volunteer to participate. It was predicted that the

residents who participated in the study would score in a

more psychosocially oriented direction on the beliefs toward

psychosocial issues measure than residents who did not

volunteer to participate.

2) It was predicted that the three experimental groups

would differ in belief change from before the training

interventions to after the interventions. Specifically, it

was anticipated that the affect + content group would show

the greatest changes, followed by the content group, and

then the control group.

3) It was predicted that the differences hypothesized in

number two above would be maintained at three month follow-

up.

4) It was predicted that there would be positive

relationships between beliefs and skill level following

training within each of the training groups and between

behavioral intentions and skill level following training

within the affect + content group.

5) It was predicted that there would be differences in the

relationship between beliefs and skill level following

training between each of the three groups. The affect +






30


content group would show the strongest relationship,

followed by the content group, and then the control group.
















CHAPTER 2
METHOD

Subjects
The subjects were 67 medical residents recruited from

various divisions within the residency training programs at

Shands Teaching Hospital. The mean age of participants was

30.5 (Sg=4.1, range 24 to 45), and the mean year of

residency was 3.0 (SE=1.6). The sample was predominantly

male (50 or 75%) and Caucasian (51 or 76%; African-American

6 or 9%, others 10 or 15%). Slightly over half of the

residents were single (35 or 52%). Residents from 13

internal medicine specialties comprised the largest group of

residents (32 or 48%), followed by pediatrics (19 or 28%),

psychiatry (14 or 21%), and surgery (2 or 3%). Residents

were solicited through three articles in the Shands

Housestaff Newsletter, three mailings of project

descriptions to all housestaff, letters to all Residency

Training Directors and all Chief Residents, posters

displayed in all clinics in which medical and surgical

residents provided clinical services, and personal

interactions by research personnel at rounds and on hospital

units. For their participation in the study, the residents

each received $100. Table 1 provides demographic inform-

ation on the residents for each experimental group. An









additional 28 residents, matched on type of training

program, level of training, and gender, were recruited to

complete pre-training and demographic questionnaires in

order to assess the possibility of sample bias.

Table 1
Demoaraphics of Residents by Group



Male:Female Mean Mean Specialty
Ratio Age Years in Medicine:
(SD) Residency Pediatrics:
Psychiatry:
Surgical (n)

Group 1 14:6 30.3 3.05 10:4:5:1
(Control) (3.6)

Group 2 18:6 31.0 3.29 11:8:5:0
(Content) (5.7)

Group 3 18:5 30.3 3.00 11:7:4:1
(Content (3.1)
+ Affect)




Materials

The materials used in this study included two

physician self-report measures and a behavioral coding

system. These included the following:

The Physician Belief Scale (Ashworth, Williamson, & Montano,

1984)

The PBS is a 32-item, rationally developed, self-report

scale designed to measure beliefs about psychosocial aspects

of patient care held by physicians. The instrument

objectively describes where a physician falls on a dimension

of acceptance versus rejection of psychosocial tenets in








medical care. The development of the scale was based upon a

theoretical framework concerning the physician's role, what

the patient wants, and physician's reactions to their

patients as people. A pool of 79 statements was developed

using these categories as starting points and presented to

180 physicians. Statistical characteristics of the initial

data were used to reduce the item pool to only those items

that would best differentiate among respondents. Twenty-two

items were deleted for reasons of abbreviated range of

response or low variability. Following a factor analysis,

twenty-five items were deleted due to negative average

intercorrelations or having loaded on unique factors. The

final set was 32 items scored individually on a five point

Likert scale of "disagree" to "agree." Scale scores could

range from 32 (maximum degree of psychosocial orientation)

to 160 (minimum psychosocial orientation). The total score

on this measure for each respondent was used for this study.

Ashworth et al. (1984) reported an internal consistency

analysis of their sample which resulted in a reliability

coefficient of i = .88, suggesting a highly internally

consistent scale with individual items tapping the same

dimension. The average scale score was 74.3 with scores

following an approximately normal distribution. The average

item variance was .86 and average inter-item correlation was

= .19.

Construct validation of the PBS was carried out by

comparing scores across internal medicine specialties








(Ashworth et al., 1984). Physicians from four disciplines

obtained scale scores congruent with expectations regarding

the psychosocial orientations of those disciplines. The

results indicated that the measure has the ability to

differentiate among respondents. Concurrent validity has

been demonstrated by comparing the Physician Questionnaire

(Levinson, Dunn, Parker, & Kaufman, 1988) with the PBS. The

correlation of the two measures was found to be r = .69.

Behavioral Intention Questionnaire

The BIS is a 24 item questionnaire rated on a 5 point

Likert scale from "never" to "always" which asks physicians

to rate to what extent they would take time in the interview

to carry out behaviors in a medical interview that are the

focus of the training programs. The individual items are

based on specific skills that are directly taught during the

content and affect modules of training. This measure takes

into consideration the need for physicians to determine

priorities during an interview due to time limitations. This

measure was developed for use in this study. It has a range

of scores from 24 to 120. The total score was used for the

purposes of this study. Only the content + affect group

received this measure. This decision was made to ensure

that the other two experimental groups were not exposed

through this questionnaire to skills that were taught in

modules in which they did not participate.








Physician Rating Scale

This measure is a behavioral coding system which was

developed for this project. It was used to assess each

physicians' skill level during a videotaped interaction with

a simulated patient following training. The effect of the

training program on physician skill was assessed by

videotaping an interview between the physician and a

simulated patient. The simulated patient was a professional

actress who was carefully coached by a physician and a

psychologist to present physical symptoms of diabetes onset

and to exhibit the initial emotional sequelae of receiving a

diagnosis of chronic illness. The interviews were videotaped

and then scored by two graduate research assistants

intensively trained to use the direct observation system

developed for the project by the research team. This direct

observation system allowed scoring of the use of specific

skills that loaded into a Total score and a Summary score

for each of nine subdomains of skill attainment. The items

being rated are behaviors that fall into the following

subdomains of skills: introductory skills, eliciting

information, active listening, nonverbal behaviors,

demonstration of respect for the patient, genuineness of

interaction, communication of empathy to the patient,

information giving, and closing skills. The Total score for

each physician was used for this study. Interobserver

agreement statistics were used to estimate numerically the

extent to which the two graduate research assistant raters








agreed on the domains being rated. As the ratings were

based on session totals, the Pearson product-moment

correlation was used to assess interobserver agreement. The

ratings of the two raters correlated at r=.79 for the total

score. It was determined a priori to randomly select one of

the raters for use as the criterion. Rater 1 was thus

selected. Appendix A provides a correlation matrix of

results comparing Rater 1, Rater 2, and the averages of both

raters. This table indicates that the findings were the

same for both raters and the averages of the two raters.

Training Proaram

The present investigation was part of a larger study

the purpose of which was to develop, implement, and evaluate

communication skills training programs designed for

physicians at the residency level of training. The project

was funded by a grant provided by the Arthur Vining Davis

Foundation and carried out by a research team comprised of

personnel from the Departments of Clinical & Health

Psychology and Medicine at the University of Florida.

The training modules that were designed integrated

existing research on effective communication strategies and

incorporated methods of teaching that have been demonstrated

to maximize learning potential. Two training modules were

developed, one designed to train content-oriented

communication skills and one designed to train additional

affect-oriented communication skills. Each of the two

training modules had three major components: didactic








presentation of specific skills, videotaped vignettes of

physician-patient interactions, and behavioral rehearsal and

feedback through supervised trainee roleplays.

A lecture format with slide presentation was chosen for

the formal didactic component of the modules. The skills

covered in each training module were included after a

careful review of the extant literature on communication

interactions in health care settings. The lectures were

written to delineate specific skills and to provide clear

explanations of the functions of each skill presented. At

the end of each didactic presentation, an extended

physician-patient interview was provided on videotape and

participants rated the physician's use of the skills

discussed using a checklist given to them by the instructor.

This checklist detailed the structural and stylistic

components of good interviewing discussed in that lecture,

providing a final review of the skills covered in the

didactic presentation.

Interspersed throughout the didactic presentation were

videotaped vignettes of physician-patient interactions that

demonstrated the skills discussed in the lecture

presentation and that also presented examples of "failures"

to communicate if the skills were not used. Actors for the

films were chosen through formal auditions after advertising

in local and campus newspapers for experienced actors. The

realism of the vignettes was assured by recruiting an

experienced R.N. to produce written scripts that represented








likely practitioner-patient interactions. These scripts

were then reviewed by a physician on the research team to

affirm medical integrity of the content. Professional

editing and production of the vignettes was provided by the

University of Florida's Learning Resource Center.

The third component of each module was behavioral

rehearsal involving roleplays in which the resident received

direct feedback and coaching from both the instructor and

other residents in the training group. Roleplays were

conducted in triads, with each small group member

sequentially playing the roles of physician, patient, and

observer. The observer for each roleplay completed a

checklist that incorporated all skills that should be

demonstrated during the interaction. After the interview

was complete, all members of the triad critiqued the

interaction using the observer's checklist and notes as a

guide.

The format of the program was as follows. The 67

residents were assigned to one of the three experimental

groups. Random block assignment was used to ensure that

specialty (medical, surgical, psychiatric, and pediatric),

level of resident training (as indicated by years in

residency), and gender were evenly distributed across the

three groups.

All groups, with the exception of the control group,

participated in three ninety minute training sessions. These

sessions were led by faculty of the University of Florida








(four psychologists and one physician) supported by four

graduate and three undergraduate research assistants. Those

assigned to the content + affect group attended a ninety

minute session which focused on content skills. The

components of this session included a didactic lecture on

content-related skills, videotape models of appropriate and

inappropriate physician communication skills, and group

discussion. This group then attended a session which

focused on the development of affect-focused physician

communication skills, which used a similar outline. The

third part of this program was a ninety minute

roleplaying/behavioral rehearsal session during which

residents broke into groups of three and practiced the

skills taught in the previous two sessions.

Those assigned to the content group attended a session

which focused on content skills identical to that provided

to the content + affect group, a ninety minute control

session during which residents viewed three videotaped

vignettes used in the affect module and discussed how

content-related skills were used by the physician (designed

to equate the time spent "in training" between the two

groups), and a ninety minute roleplaying/behavioral

rehearsal session during which they practiced the skills

taught in the first session.

In order to accommodate the work schedules of the

residents, six content-focused modules, six affect-focused

modules, six review sessions, and seven








roleplaying/behavioral rehearsal sessions were conducted

over the span of eight months.

Those assigned to the control group did not receive any

training, however, they did receive the entire assessment

battery.

Design and Procedure

Data collection for the study occurred at three separate

times: prior to training, following training subsequent to

evaluation of skill acquisition, and beginning at three

month follow-up. The study was carried out in the following

phases:

1) Residents completed self-report questionnaires at the

time of recruitment.

2) The residents were randomly assigned to one of three

groups: a content-only training program, a content +

affect-training program, and a waitlist control group.

Residents assigned to the training groups were then

trained.

3) Following training each resident was videotaped

conducting an interview with a simulated patient. The

videotape was coded to assess the skill level following

training of each resident. At this time residents once

again completed the self-report questionnaires to assess any

changes in attitude (beliefs and intentions measures).

4) Three months following the conclusion of the training

programs residents once again completed the attitude

questionnaires. Questionnaires were sent to the residents






41

via interdepartmental mail and returned in the same manner.

Residents who did not return the questionnaires at this

point were contacted through subsequent mailings and again

asked to complete and return the questionnaires.

5) Following the final training session a matched group of

residents who did not participate in the training was

recruited to complete the Physician Belief Scale. The

residents were matched on specialty, gender, and level of

training. Questionnaires were sent to the residents via

interdepartmental mail and returned in the same manner. A

total of 130 residents were sent questionnaires. 28

returned completed questionnaires for a response rate of 22

percent.















CHAPTER 3
RESULTS

Statistical analyses served several primary purposes.

First, the sample of residents who agreed to participate in

the training program was compared to a matched group of

residents who did not participate to assess the potential

for sample bias. Second, analysis of variance tests, with

covariates as appropriate, were conducted to examine

differences between training groups on belief and intention

change from pre- to post-training and at follow-up. Third,

correlation and multiple regression analyses were carried

out to determine the relationships of beliefs and intentions

and skill level following training within groups and the

differences of such relationships between groups.

The Kolmogorov-Smirnov Goodness of Fit test was used to

test the assumption that the scores for the Physician Belief

Scale, the Behavior Intention Scale, and the Physician

Rating Scale were normally distributed in the population.

The hypothesis that the scaled scores approximated normal

distributions was not rejected for any of these measures

(E>,.05, n=67, (n=23 for BIS)). Table 2 presents the means

and standard deviations for all measures by group and for

the total sample.








Table 2
Means and Standard Deviat s


for 5tud Measures u


and for Total Sample



PBS-pre PBS-post PBS-diff PRS
(SD) (SD) (SD) (SB)

Group 1 61.45 68.35 6.90 25.05
(Control) (12.62) (15.40) (8.98) (6.1)
n=20

Group 2 70.33 66.29 -4.04 34.04
(Content) (12.41) (13.70) (9.16) (5.68)
n=24

Group 3 75.26 70.26 -5.00 34.48
(Affect) (15.69) (14.95) (8.95) (6.90)
n=23

Total 69.37 68.27 -1.10 31.51
n=67 (14.59) (14.53) (10.34) (7.47)



BIQ-pre BIQ-post BIQ-diff

Group 3 92.57 102.30 9.74
(Affect) (10.63) (9.32) (6.73)
n=23


Note.
PBS-pre = Physician Belief Scale pre-training
PBS-post = Physician Belief Scale post-training
PBS-diff = Physician Belief Scale post pre training
PRS = Physician Rating Scale
BIQ-pre = Behavior Intention Scale pre-training
BIQ-post = Behavior Intention Scale post-training
BIQ-diff = Behavior Intention Scale post pre training



Participants versus Nonparticinants

The first hypothesis of the study suggested that

residents who participated in the communication skills

training programs would differ in their beliefs about








psychosocial issues in patient care from residents who did

not volunteer to participate. It was anticipated that

residents who participated in the study would score in a

more psychosocially oriented direction on the beliefs

measure. A paired sample t-test was conducted to test for

differences in beliefs (as measured by the PBS) between

residents who participated in the training program and a

matched group of residents (matched on gender, specialty,

and level of training) who chose not to participate. The

results indicate that the two groups (training group

1=68.77, SE=16.04, matched group M=67.12, SD=16.67) did not

differ in beliefs about psychosocial issues in patient care

_(df=27)=-.41, R=.68.

ANOVAS and ANCOVAS

The second hypothesis of the study stated that the

three experimental groups would differ in belief change from

before the training interventions to after the

interventions. It was hypothesized that the affect +

content group would demonstrate the greatest changes,

followed by the content group, and then the control group.

An ANOVA performed on the pre-training PBS indicated

significant differences between the three experimental

groups E(2,64)=5.54, 2=.006. Therefore, an ANCOVA was

conducted to test for differences between the three groups

on belief change pre- to post-training as assessed by the

Physician Belief Scale (PBS). The ANCOVA method was chosen

to control for the significant differences that emerged








between the three groups on the pre-training PBS. The post-

training PBS score was used as the dependent variable,

group membership and pre-training PBS score were used as

independent variables, with pre-training PBS score used as a

covariate. The results of the ANCOVA suggested that

significant differences existed between groups in belief

change from before the training programs to after the

programs E(2, 63)=7.41, R<.001. Both the content group

(_(df=42)=3.9, p<.001) and the content + affect group

(.(df=41)=4.34, g<.001) significantly differed from the

control group in belief change from pre- to post-training.

There was no significant difference between the content and

the content + affect groups (t(df=45)=.36, p=.72) in belief

change. Paired sample t-tests within each group on pre-

training beliefs to post-training beliefs indicated that the

content (t(df=23)=2.16, E<.05) and content + affect

(f(df=22)=2.68, g<.01) groups became significantly more

psychosocially oriented, while the control group became less

psychosocially oriented (t(df=19)=-3.4, E<.01) from before

to after training.

The second part of this hypothesis, that the content +

affect group would demonstrate a significant increase in

behavioral intentions from pre- to post-training, was tested

with a paired sample t-test. This group did demonstrate a

significant increase in behavioral intentions

(t(df=22)=4.49, R<.001).








The third hypothesis of the study stated that the

differences in belief change would be maintained at three-

month follow-up. Table 3 presents the means and standard

deviations for the attitude measures (PBS and BIQ) by group

for the subset of residents who completed follow-up

questionnaires.

Table 3
M. and SD for Study Measures by Group for Residents Who
Completed Follow-up Ouestionnaires



PBS-pre PBS-post PBS-fu PBS-diff2



Group 1 63.18 70.82 71.46 8.27
(Control) (11.02) (15.53) (14.41) (7.49)
n=ll

Group 2 69.19 64.81 66.18 -3.00
(Content) (12.75) (13.91) (14.65) (7.49)
n=16

Group 3 74.09 71.00 68.82 -5.27
(Affect) (18.41) (19.69) (16.26) (13.49)
n=11


BIQ-pre BIQ-post BIQ-fu BIQ-dif2


Group 3 89.64 103.09 99.00 9.36
(Affect) (12.78) (11.05) (10.92) (7.28)
n=11

Note.
PBS-pre = Physician Belief Scale pre-training
PBS-post = Physician Belief Scale post-training
PBS-fu = Physician Belief Scale follow-up
PBS-diff2 = Physician Belief Scale follow-up pre-training
BIQ-pre = Behavior Intention Scale pre-training
BIQ-post = Behavior Intention Scale post-training
BIQ-fu = Behavior Intention Scale follow-up
BIQ-dif2 = Behavior Intention Scale follow-up pre-training








For the residents who completed the follow-up

questionnaires, an ANCOVA was conducted to test for the

maintenance of belief change differences between groups at

follow-up. As with the previous hypothesis, the ANCOVA

method was chosen to control for the significant differences

that emerged between the three groups on the pre-training

PBS. The follow-up PBS score was used as the dependent

variable, group membership and pre-training PBS score were

used as independent variables, and pre-training PBS score

was used as a covariate. The results of the ANCOVA suggest

that significant differences existed between groups in

belief change from before the training programs to follow-up

several months after the programs E(2,34)=4.87, D<.01. Both

the content group (t(df=25)=3.8, E<.001) and the content +

affect group (t(df=20)=2.9, E<.01) significantly differed

from the control group in belief change from pre-training to

follow-up. There was no significant difference between the

content and the content + affect groups (t(df=25)=.56,

R=.58). Paired sample t-tests within each group on pre-

training beliefs to follow-up beliefs indicated that the

content (f(df=15)=1.6, E=.13) and content + affect groups

did not become significantly more psychosocially oriented,

although the control group became significantly less

psychosocially oriented (t(df=11)=-3.65, p<.01) from before

training to follow-up. Additionally, the content + affect

group demonstrated a significant increase in behavioral

intentions from pre-training to follow-up as suggested by









the results of a paired sample t-test (t(df=10)=-3.42,

R<.01).

Correlation and Regression Analyses

The fourth hypothesis of the study stated that there

would be a positive relationship between beliefs and skill

level following training within each of the experimental

groups and between behavioral intentions and skill level

following training within the affect + content group.

Pearson correlations were obtained to assess the

relationship between beliefs and intentions and skill level

following training. Table 4 presents the correlation matrix

for these variables.

Table 4
Correlation Matrix for Beliefs. Intentions, and Skill Level
Following Training



Skill Level Following
Training (PRS Total Score)


Beliefs

Group 1
(Control)
PBS-pre -.09
PBS-post -.13
PBS-difference -.09

Group 2
(Content)
PBS-pre .20
PBS-post .12
PBS-difference -.10

Group 3
(Affect)
PBS-pre -.12
PBS-post -.19
PBS-difference -.10








Table 4. continued

Total
PBS-pre .19
PBS-post -.06
PBS-difference -.36*

Intentions

Group 3
(Affect)
BIQ-pre -.35
BIQ-post -.14
BIQ-difference .23



*<.*01



The lone signicant result involved the entire sample of

residents. An increase in psychosocial orientation from

pre- to post-study was related to higher skill level

following training (r=-.36, p=.003). However, this finding

was not significant with the Bonferroni adjustment to

control for Type 1 errors in multiple comparisons (p=.18).

The physicians' beliefs and intentions, measured prior to

and subsequent to training, were not significantly related

to skill level following training for any of the groups or

the total sample.

The fifth hypothesis of the study purported that there

would be differences in the relationship between beliefs and

skill level following training between each of the three

groups. It was suggested that the affect + content group

would show the strongest relationship, followed by the

content group, and then the control group. A multiple

regression analysis with a dummy variable to code for group






50

membership was used to test for differences in the

relationship between beliefs and skill level following

training. A cross-product interaction term (PBS-pre x

group) was included in the regression equation to test for

differences between groups. Physician Rating Scale Total

score was the dependent variable, with PBS-pre and group

membership as independent variables. The results of this

analysis revealed a nonsignificant regression equation.

Evidence of interaction between the independent variables

did not exist, Y(2,61)=.65, R=.52, thus indicating no

significant differences between groups in the relationship

between beliefs and skill level following training.















CHAPTER 4
DISCUSSION

This was an investigation of medical residents'

attitudes toward psychosocial issues in patient care, how

components of such attitudes, specifically beliefs and

behavioral intentions, relate to interview skill development

in communication training programs, and whether these

components of attitudes change as a result of communication

training. This study was guided by the findings from and

the framework provided by social psychological research on

the attitude construct.

The first hypothesis of the study positted that

residents who participated in the study would report beliefs

that were more psychosocially oriented than a matched

control group of residents who chose not to participate.

This hypothesis was not supported. The finding indicates

that residents who volunteered to participate in the four

and one-half hour, three session, communication training

program for an honorarium of $100 had a similar level of

psychosocial orientation toward patient care as residents

who did not volunteer to participate in the training

program, but who agreed to fill out a brief questionnaire

assessing level of psychosocial orientation. This suggests

that the sample of residents who participated in the








communication program may not be a sample biased toward a

psychosocial orientation to medical care, but rather may be

representative of the total sample of residents informed of

the opportunity to participate in the study. This implies

that the findings of this study may be generalized further

than if this sample of residents proved to be biased.

The finding that the trained group of residents did not

differ from the matched group is limited by the fact that

both the trained residents and the matched residents

voluntarily agreed with the request to participate by at

least completing a questionnaire. These two groups together

may differ from the group of residents who did not agree to

both the request to take part in the training program and

the request to fill out the questionnaire. Research which

included responses from the group of residents who refused

to volunteer to either participate in the training or to

complete the attitude questionnaire would clarify this

issue. In order for such data to be collected, residents

would have to be in a position where they were required to

complete the questionnaire, a liberty this project did not

have.

The second hypothesis of the study proposed that the

three experimental groups would differ in belief change from

pre- to post-training. It was hypothesized that the content

+ affect group would show the most change, followed by the

content group, and then the control group, which was not

expected to show any change. The hypothesized differences








were accurate for comparisons between each of the training

groups and the control group, but not so for the comparison

of the two training groups. The findings indicate that

components of medical residents' attitudes toward

psychosocial issues in patient care can change as the result

of brief psychosocially-oriented training interventions.

The belief change of residents in both communication

training groups of this study from pre- to post-training was

significantly more psychosocially directed and oriented in

comparison to the control group. Additionally, the

behavioral intentions of residents in the content + affect

group changed significantly in the direction toward carrying

out the behaviors taught by the training programs. These

findings imply that influencing medical residents' attitudes

in a more psychosocially oriented direction, as has been a

goal of medical education training, is indeed attainable.

Although the results indicated that the components of

attitudes did change as a result of the training programs,

less clear is what the factors are that caused the changes.

One explanation of the changes is to emphasize that by

altering the residents' behavior through the practice of

specific behavioral skills which were central components of

the training program, i.e. roleplaying, behavioral

rehearsal, and simulated exercises, modifications in the

beliefs and behavioral intentions were effected. By acting

in a particular manner in accordance with some skill, more

fundamental attitudinal adjustments may be created which








will serve to maintain that type of behavior in the future.

A second approach would suggest that the attitude components

were altered during the training as a result of verbally

transmitted information which was communicated as part of

the didactic lecture format. What seems most likely is that

a combination of these factors produced the changes in the

beliefs and behavioral intentions. The relationships

between the beliefs, behavioral intentions, and behaviors

may be bidirectional in nature. Specifically, changes in

the residents' cognitive structures (beliefs and intentions)

likely influenced their behavior to change, at the same time

that changes in the residents' behaviors influenced

alterations in their cognitive structures. This

interactionist approach maintains that behavior is best

accounted for as an interactive function of external stimuli

and person variables (i.e. cognitions). Which of these

factors may have the most powerful influence on the others

is unclear from the results of this study. The design of

the study does not make it possible to make conclusive

statements regarding the directionality of the relationship

between the cognitive structures and the behaviors as both

training groups used the same teaching format, which

included both didactic and behavioral components.

The attitude change results are consistent with those

of Smith et al. (1991) which found that a much more

intensive, comprehensive month-long training program which

focused on communication and relationship-building skills








with first-year residents resulted in improved attitudes

toward psychosocial issues toward patient care compared to a

control group. Both studies used beliefs as the attitude

component of interest, utilized pre- and post-study

evaluation of the beliefs, and included a control group in

the study design. However, the current study extends the

findings of Smith et al. through the use of an experimental

design with random group assignment, a belief measure with

demonstrated reliability and validity, a wider range of

resident specialties, and a much briefer training program.

The finding that components of psychosocial attitudes

of residents can change as a result of training provides

empirical validation of the idea that the period of

residency training is an important time to influence

physicians. Consider Principle Four of the Four Principles

Regarding Humanistic Qualities of the Internist Adopted by

the American Board of Internal Medicine: The ability to

affect attitudes, behavior patterns, and moral conduct in

medical care should be recognized and utilized during the

residency training a unique experience that is not

available at other times in medical education (Merkel,

Margolis, & Smith, 1990). The present study confirms that

this is indeed possible.

The finding of significant differences between each of

the training groups and the control group is limited by the

significant differences that emerged between each of the

groups on the belief scale pre-training. As random block








assignment was employed in the assignment of residents to

groups, it was not expected that such differences would

exist. Although the statistical analyses which revealed the

differences in belief change between the groups controlled

for the pre-training belief scale differences, it is

worthwhile to understand what may have contributed to this

initial divergence. At least three explanations may account

for this unexpected finding. First, the differences could

be the result of the relatively small sample size of each

group. Second, the differences may be a reflection of

different response sets which characterized each group. The

control group may have been more likely to respond to items

in a positive direction than the two training groups. This

may be a result of being paid an honorarium despite not

being required to participate in the training. As a way of

ensuring minimal cognitive dissonance between beliefs which

accompanied their decision to complete the questionnaire

and their beliefs which were assessed on the questionnaire

the control residents may have been more likely to endorse

items in a positive direction. The content and content +

affect groups may have been less likely to respond to items

in a positive direction as a result of having to fill out an

additional questionnaire (the Behavioral Intention

Questionnaire) which was not required of the content group.

Additionally, the two training groups likely experienced

less cognitive dissonance as they had less need to justify

their beliefs since they had agreed to participate to a much








greater extent in the study. Third, a combination of

these two factors may have resulted in the pre-training

belief scale differences.

The hypothesized difference in belief change between

the two training groups of this study was not confirmed.

Various possible explanations exist for this result. First,

it may be that the factors which resulted in belief change

were present in each of the training programs. Although the

training program sought to directly teach the affect-focused

skills (demonstration of empathy, respect, and genuineness

to the patient, understanding nonverbal communication,

active listening skills ) to only the affect group, it is

possible these skills were indirectly modeled during the

process of teaching the content-oriented skills. In other

words, affect-oriented and content-oriented skills may not

be as easily separated as anticipated prior to training the

residents, and the amount of content overlap between the two

training groups may have prevented detectable differences

between the two groups from emerging. Second, the measure

used to assess beliefs may not be sensitive to components of

affective change taught to the content + affect group as it

focuses on beliefs rather than feelings. The present study

could have included a more direct affective measure,

specifically a subjective judgement of general evaluation or

feeling of favorableness or unfavorableness toward

psychosocial issues in patient care, as proposed by the

Fishbein-Ajzen model discussed earlier. Including such a








measure may have helped to clarify such differences between

the two groups.

The third hypothesis of the study proposed that the

differences between the three experimental groups in belief

change would be maintained at follow-up. Specifically, it

was purported that the content + affect group would continue

to show the most change, followed by the content group, and

then the control group, which was not expected to show any

change. The hypothesized differences were accurate for each

of the training groups and the control group, but not so for

the comparison of the two training groups. The findings

suggest that changes in components of medical residents'

attitudes toward psychosocial issues in patient care which

result from brief psychosocially-oriented training

interventions can be maintained over time. The belief

change of residents in both communication training groups of

this study from pre-training to follow-up was significantly

more psychosocially directed and oriented in comparison to

the control group. Although within each of the training

groups belief change failed to reach significance, this may

have been due to the relatively small sample size of

residents who completed the follow-up questionnaires, as the

differences for each of the training groups approached

significance. Additionally, the behavioral intentions of

residents in the content + affect group changed

significantly as assessed at follow-up in the direction

toward continuing to carry out the behaviors taught by the








training programs. These findings imply that not only is it

possible to influence medical residents' attitudes in a more

psychosocially oriented direction, but that changes in such

attitudes can be maintained over time.

The attitude maintenance results are consistent with

those of the Smith et al. (1991) study mentioned earlier.

That study assessed 13 of 28 trained resident at 15 month

follow-up and found that no significant decline in attitudes

toward psychosocial medicine had occurred from post-training

to follow-up. The current study extends the finding of

Smith et al. through the use of an experimental design with

random group assignment including a control group as part of

the assessment at follow-up.

The finding of significant differences between each of

the training groups and the control group on the

psychosocial belief scale at follow-up is limited by the

number of residents who participated at follow-up (38 out of

67). The sample was smaller than planned and limited the

statistical power of the analysis. Moreover, it is possible

that agreeing to return the study questionnaire at follow-up

is a reflection of a positive attitude toward the study and

may have resulted in a biased sample. However, this issue

appears to have been adequately controlled for by the use of

an experimental design with a control group. Additionally,

the pattern of results at follow-up is similar to that which

emerged immediately post-training, which further suggests

that sample bias was not an issue at follow-up.








The fourth hypothesis of the study proposed positive

relationships between the attitude components, beliefs and

behavioral intentions, and skill level (for the control

group) and skill level following training (for the trained

groups). The results of the Pearson correlations do not

support the hypothesis that there would be a positive

relationship between physicians' beliefs and skill level

within each of the groups and between behavioral intentions

and skill level within the affect + content group. This

suggests that residents' beliefs about psychosocial issues

in patient care (for residents in each of the experimental

groups) and residents' intentions to carry out skills taught

in the training program (for the content + affect group) are

not related to interview skill level (for the control group)

and interview skill level following communication training

(for the two trained groups). These findings indicate that

improved attitudes towards psychosocial information do not

correspond to an increased ability to demonstrate interview

skill during a simulated patient interview.

The finding of the lack of a relationship between

components of psychosocial attitudes and interview skill

level and interview skill level following communication

training is consistent with research previously carried out

at the medical student level. Dwyer et al. (1988) examined

the relationship between third year medical students'

psychosocial attitudes and the ability to elicit and use

effectively psychosocial information during a medical








interview conducted during a family practice rotation. By

the use of a correlational design they found that medical

students' psychosocial attitudes were not related to their

ability to accurately assess patient reliance on social

networks, patient compliance, and patient income. The

current study extends these previous findings by its use of

an experimental design with random assignment and the

incorporation of a control group, the assessment of resident

physicians at a level of training more advanced than medical

students, the inclusion of a range of specialties, and a

direct assessment of interview skill rather than inferring

skill through information communicated during the interview.

The finding of the current study that residents'

components of psychosocial attitudes do not relate to

interview skill level and the finding of the Dwyer et al.

(1988) study taken together have implications for medical

education programs. As these findings suggest that improved

attitudes towards psychosocial issues in patient care do not

necessarily correspond to an increased skill in effectively

conducting a medical interview, medical educators may need

to reassess the content and focus of psychosocial

interventions in medical training. Whereas much previous

research and training has focused on characteristics

affecting attitudes toward psychosocial issues, inferring

that such attitudes are acceptable surrogates for

demonstrated ability, the present findings indicate that








measures of psychosocial attitudes are not an acceptable

substitute for proven measures of physician skill.

Although other findings of the present study

(specifically that beliefs and intentions change from pre-

to post-training) support the idea that demonstrating to

residents the valid diagnostic use of psychosocial

interviewing skills does have an effect and should be

emphasized in residency training, such a focus on attitude

components does not appear to be sufficient. Communication

training programs at the residency level should be

encouraged to focus on specific communication skills which

are aimed at increasing interviewing proficiency. It is

clear from present and previous research findings that

attitudes do not necessarily reflect ability and that

research and training programs should progress beyond

equating these two discrete variables. Residency

communication training programs should aim toward procedures

that teach residents simultaneously the value of

psychosocial issues in patient care and how to effectively

acquire such information efficiently and accurately through

specific interviewing skills. Williamson, Smith, Kern,

Lipkin, Barker, Hoppe, & Florek (1992) have provided

recommendations for training residents in the psychosocial

aspects of medicine and the medical interview which takes

into consideration the importance of distinguishing between

specific skills and attitudes. Indeed, the current study

exemplifies how psychosocial attitudes and behavioral








interviewing skills can each be influenced through clearly

defined techniques.

The present finding may be limited by measurement

problems. Specifically, attitudes (beliefs and intentions

in the present study) and behaviors (reflected in the

interview skill level assessment) can only be expected to be

related when the underlying disposition of which they are

expressions is measured adequately. This study emphasized

beliefs based on the assumption that cognitive information,

rather than affective information, would best define

attitude for the present study. Such an assumption was made

for the current study a priori based on the recommendation

by Zanna and Rempel (1988) that beliefs best predicted

behavior when the behavior was undertaken for instrumental

purposes, i.e., gathering information during a medical

interview. These authors also note that when behavior was

performed for consumatory, noninstrumental purposes

effectivelyy driven), it was better predicted by attitudes

based primarily on feelings. What the findings of the

present study may suggest is that the distinction Zanna and

Rempel make between instrumental and noninstrumental

purposes may not be so clear-cut. As a result the present

study, by not including a measure that directly assessed the

affective component of attitudes toward psychosocial issues,

may not have adequately assessed the underlying disposition

that would allow for a correlation between attitudes toward

psychosocial issues and interview skill level. Indeed, the








present findings may reflect that the information gathering

process of the medical interview is much more effectively

driven than cognitively driven. Specific research needed to

clarify this issue would require that the three postulated

sources of attitudes (cognitions, affects, and behaviors) be

measured separately and each included in research directed

at understanding psychosocial attitudes of medical

residents. The present study, consistent with previous

research of physicians and physicians in training, did not

assess the affective component of attitudes adequately.

Moreover, the lack of emphasis on the affective component of

psychosocial attitudes in patient care is evident in that

existing measures focus exclusively on beliefs. Measurement

techniques for the affective component of attitudes, which

could include asking physicians in training to directly rate

relevant feelings, have not been developed or utilized in

this body of research.

The fifth hypothesis of the study positted that there

would be differences in the relationship between attitude

components and skill level following training between each

of the three groups. No such differences emerged from the

statistical analyses. This finding indicates that an

interaction effect between training program and the pre-

training levels of beliefs and intentions did not exist for

any of the three groups. This finding is limited by the

lack of any significant relationships between the attitude

components and skill within any of the three groups. Given








the findings of the previous hypothesis, specifically, that

for none of the groups did a positive relationship emerge

between attitude components and skill level, the finding of

the present hypothesis is not unexpected. The likelihood

that differences will emerge between groups becomes less

when significant relationships within any groups do not

emerge. Future research into the issue of interactions

between residents' psychosocial attitudes and types of

training programs would be best carried out once

relationships between these variables have been established.

The overall findings of this study have general

implications for theory, research, and practice.

Theoretically, the results indicate that for attitudes

toward psychosocial issues in patient care it is not clear

how the three classes of information postulated by Zanna and

Rempel (1988) to determine an attitude (cognitive

information, affective information, and information about

behavioral intentions) combine to develop this particular

type of attitude for medical residents. This study did not

sufficiently address this issue, as it followed suggestions

from previous research which emphasized the importance of

beliefs in the formation and assessment of attitudes. In

retrospect, an approach which made an a priori assumption

that beliefs are the predominant class of information in the

determination of the attitude was premature. Research into

psychosocial attitudes toward patient care has not

demonstrated that such an assumption is accurate. At this








stage of research into this type of attitude, all three

sources of information should be assessed. A more complete

approach to understanding the roles of these three classes

of information would have included measures to assess each

class for each of the three experimental groups.

With respect to research, the current study has

implications in the areas of design and measurement. In

terms of design, the present study included findings which

confirmed results previously found with weaker quasi-

experimental or correlational designs. The results of this

study lobby for the continued use of stronger designs, such

as the experimental design in this case, that yield more

conclusive results. In terms of measurement, the present

study highlights the need for a more comprehensive approach

to the assessment of psychosocial attitudes which includes

measurement of the cognitive, affective, and behavioral

components of attitudes concurrently. To date, only within

the cognitive domain have measures with established

reliability and validity been developed. The findings of

the present study emphasize the need for reliable and valid

assessment of the affective and behavioral domains of

psychosocial attitudes, as well.

With respect to practice, the present findings have

direct implications for medical education involving the

training of interviewing skills. These implications were

discussed in depth earlier in this section, and bare

repeating again. Medical educators may want to pay








attention to these findings as they suggest ways for the

goals of effective physician-patient communication and

improved physician-patient relations to be achieved.

Educators may want to alter their training programs to

incorporate the present suggestions aimed at improving the

teaching of interviewing skills. Specifically, residency

educators are likely to find that communication training

programs with the goal of improving residents' interviewing

skills may be most effective when the emphasis is

concurrently on both teaching residents specific skills

involved in acquiring psychosocial information efficiently

and accurately and emphasizing to residents the value of

psychosocial information in patient care. This suggestion

may also be generalizable to physicians and physicians-in-

training at various levels of training.

A discussion of the findings of the present study would

be incomplete without consideration of general

methodological limitations. For this study these

limitations generally come from decisions made about how to

conduct the study and problems that emerged when conducting

it. In terms of external validity, the disproportionate

number of psychiatrists and pediatricians who participated

in the study (49%) may limit the generalizability of the

findings to other residency specialties, especially as these

two specialties tend to score in a more psychosocially-

oriented direction than other specialties (Appendix B).

Another issue having to do with the external validity of the








findings involved the generalizability of the simulated

patient interview. While the use of this method allowed for

excellent experimental control it does raise the issue of

whether the skill level assessed in this artificial setting

is generalizable to a clinical interview. In terms of

measurement issues, the reliance on self-report measures for

the assessment of the attitude components may limit the

findings. Although this approach is typical for research

into psychosocial issues in patient care, it does emphasize

the need for alternative methods of assessing such attitudes

with this population, such as the "bogus pipeline" method

used in social psychological research which minimizes social

influences on expressed attitudes. Lastly, in terms of

data collection, due to initially low response rates to

participate in the the program by residents and difficulty

matching busy resident schedules with training times, the

number of residents in the study was lower than initially

expected (67 instead of 75), limiting the statistical power

somewhat.

The overall findings of the present study suggest

directions for researchers to pursue in the future. First,

this type of study could be extended to other populations,

including both physicians at other levels of training and

other health professionals with clinical responsibilities

that involve direct communication with patients. Second, a

more comprehensive approach to studying psychosocial

attitudes of residents would involve focusing not only on








the role of the cognitive component of attitudes, beliefs,

and the behavioral component, behavioral intentions in this

study, but also the affective component. Future research

should attempt to assess the role of all three attitude

components and the relationships between them. Third, given

that it is clear from the findings that psychosocial

attitude components do not equate with interview skill,

research should advance to studying these two variables

separately. Medical education programs and research should

focus simultaneously on both how to improve residents

psychosocial attitudes as well as how to most effectively

teach interviewing skills. As future research clarifies

these issues and allows for greater comprehension of the

processes involved in physicians' attitudes toward

psychosocial issues in patient care, ultimately improvements

in the training of physicians and the patient care which

they provide should result.
















APPENDIX A
CORRELATION MATRIX
BY RATER

Correlation Matrix for Beliefs. Intentions. and Skill Level
Following Training By Rater



Skill Level Following Training (PRS)
Rater 1 Rater 2 Average


Beliefs

Group 1
(Control)
PBS-pre -.09 -.14 -.12
PBS-post -.13 -.11 -.13
PBS-difference -.09 -.14 -.06

Group 2
(Content)
PBS-pre .20 .38 .31
PBS-post .12 .29 .18
PBS-difference -.10 -.15 -.14

Group 3
(Affect)
PBS-pre -.12 -.23 -.19
PBS-post -.19 -.27 -.25
PBS-difference -.10 -.05 -.08

Total
PBS-pre .19 .13 .17
PBS-post -.06 -.06 -.07
PBS-difference -.36** -.27* -.34**


*R<.05
**R<.01
















APPENDIX B
BELIEF SCORES BY SPECIALTY

Physician Belief Scale Scores by Residency Specialty



Specialty PBS-pre PBS-post


Surgical 82.5 78.5
N=2

Medical 74.4 74.5
N=32

Pediatric 69.7 65.5
N=19

Psychiatric 55.5 56.2
N=14















REFERENCES

Accreditation Council for Graduate Medical Education.
(1986). Essentials of accredited residencies in
graduate medical education: 6 special requirements
for residency training in family practice. In
Directory of Residency Training Programs (pp. 23-28).
New York: American Medical Association.

Almy, T.P. (1989). The gastroenterologist and the
graying of America. American Journal of
Gastroenterology, 84, 464-468.

Altchek, E.D. (1988). Informed consent. Plastic and
Reconstructive Surgery, J2, 920.

American Board of Internal Medicine, Subcommittee on
Evaluation of Humanistic Qualities in the Internist.
(1983). Evaluation of humanistic qualities in the
internist. Annals of Internal Medicine, 99, 720-724.

American Board of Pediatrics, Medical Ethics Subcommitee.
(1987). Teaching and evaluation of interpersonal
skills and ethical decision making in pediatrics.
Pediatrics, Z7, 829-833.

Ashworth, C.D., Williamson, P., & Montano, D. (1984).
A scale to measure physician beliefs about psycho-
social aspects of patient care. Social Science in
Medicine, 19, 1235-1238.

Azjen, I. (1984). Attitudes. In R.J. Corsini (Ed.),
Wiley encyclopedia of psychology, volume I (pp. 99-
100). New York: Wiley.

Bartlett, E.E., Grayson, M., & Barker, R. (1984). The
effects of physician communications skills on patient
satisfaction, recall, and adherence. Journal of
Chronic Disease, 32, 755.

Bass, D.M., Garland, T.N., & Otto, M.E. (1986).
Characteristics of hospice patients and their care-
givers. Omega Journal of Death and Dying, 1i, 51-68.









Becker, H.S. & Geer, B. (1958). The fate of idealism in
medical school. American Sociological Review, 21, 50-
56.

Beckman, H.B., & Frankel, R.M. (1984). The effect of
physician behavior on the collection of data. Annals
of Internal Medicine, In1, 692-696.

Bem, D.J. (1972). Self-perception theory. In L.
Berkowitz (Ed.), Advances in experimental social
psvcholoav. vol. 6 (pp. 2-57). New York: Academic
Press.

Bird, J. & Cohen-Cole, S.A. (1990). The three function
model of the medical interview: An educational
device. In Hale, M. (Ed.). Models of Consultation-
Liaison Psychiatry (pp. 65-88). Switzerland: Karger.

Blizek, W.L. & Finkler, D. (1977). Teaching human
values in medicine. Journal of Medical Education, 52,
858-859.

Burack, R.C. & Carpenter, R.R. (1983). The predictive
value of the presenting complaint. Journal of Family
Practice, 1&, 749-754.

Carter, W.B., Inui, T.S., & Kukull, W.A. (1982).
Outcome based doctor-patient interaction analysis: II.
Identifying effective provider and patient behavior.
Medical Care, 20, 550-559.

Christie, R. & Merton, R.K. (1958). Procedures for the
sociological study of the values climate of medical
school. Journal of Medical Education, 33, 125-153.

Cope, D.W., Linn, L.S., Leake, B.D., & Barrett, P.A.
(1986). Modifications of residents' behavior by
preceptor feedback of patient satisfaction. Journal
of General Internal Medicine, 1, 394-398.

Cox, A., Holbrook, D., & Rutter, M. (1981). Psychiatric
interviewing techniques. VI. Experimental study:
eliciting feelings. British Journal of Psychiatry,
139, 144-152.

Cox, A., Hopkinson, K. & Rutter, M. (1981). Psychiatric
interviewing technique. II. Naturalistic study:
eliciting factual information. British Journal of
Psychiatry, 18, 283-291.








Cox, A., Rutter, M., & Holbrook, D. (1981). Psychiatric
interviewing techniques. V. Experimental study:
eliciting factual information. British Journal of
Psychiatry, 12, 29-37.

Davis, D. & Ostrom, T.M. (1984). Attitude measurement.
In R.J. Corsini (Ed.), Wiley encyclopedia of
Dsvchologv. vol. 1 (pp. 100-102). New York: Wiley.

DeGroot, L. (1987). Pliable but not receptive:
concerning the marginal influence of a medical
psychology course on the socialization process of
doctors. Medical Education, 21, 419-425.

DeMonchy, C., Richardson, R., Brown, R.A., & Harden, R.M.
(1988). Measuring attitudes of doctors: the doctor-
patient rating. Medical Education, 22, 231-239.

DiMatteo, M.R. & Hays, R. (1980). The significance of
patients' perceptions of physician conduct: a study
of patient satisfaction in a family practice center.
Journal of Community Health, 6, 18-34.

DiMatteo, M.R., Hays, R., & Prince, L.M. (1986).
Relationship of physicians' nonverbal communication
skill to patient satisfaction, appointment
noncompliance, and physician workload. Health
Psycholoav, I, 581-594.

Dornbush, R.L., & Singer, P., Brownstein, E.J., &
Freedman, A.M. (1984). Measurement of psychosocial
attitudes in medical students. Journal of Medical
Education, 59, 59-61.

Dornbush, R.L., Singer, P., Brownstein, E.J., &
Freedman, A.M. (1985). Maintenance of psychosocial
attitudes in medical students. Social Science in
Medicine, 20, 107-109.

Dwyer, J.W., Detweiler, N.L., & Kosch, S.G. (1988).
Medical students and comprehensive patient care:
attitudes, perceived competence and demonstrated
ability. Medical Education, 22, 19-26.

Eron, L.D. (1955). Effect of medical education on
medical students' attitudes. Journal of Medical
Education, 30, 559-565.








Evans, B.J., Kiellerup, F.D., Stanley, R.O., Burrows, G.
& Sweet, B. (1987). A communication skills program
for increasing patients' satisfaction with general
practice consultations. British Journal of Medical
Psychology, 60, 373-378.

Ewan, C.E. (1987). Attitudes to social issues in
medicine: a comparison of first-year medical students
with first-year students in non-medical faculties.
Medical Education, 21, 25-31.

Fishbein, M. & Ajzen, I. (1975). Belief, attitude.
intention, and behavior: an introduction to theory
and research. Reading, MA: Addison-Wesley.

Fletcher, C. (1980). Listening and talking to patients.
I. The problem. British Medical Journal, 2i,
845-847.

Fox, R.C. (1957). Training for uncertainty. In R.K.
Merton, G. Reader, & P.L. Kendall (Eds.), The student
physician (pp. 207-241). Cambridge: Harvard
University Press.

Frankel, R.M. (1984). From sentence to sequence:
understanding the medical encounter through micro-
interactional analysis. Discourse Process, 2,
135-170.

Gerber, L.A. (1987). Experiences of forgiveness in
physicians whose medical treatment was not successful.
Psychological Reports, 51, 236.

Glaser, R. (1951). Predicting achievement in medical
school. Journal of Applied Psychology, 35, 272-276.

Goldberg, D.P. (1980). Training family doctors to
recognize psychiatric illness with increased accuracy.
Lancet, II, 521-523.

Goldberg, R.J., Novack, D.H., Fulton, J.P. &
Wartman, S.A. (1985). A survey of psychiatry and
behavioral science curricula in primary care residency
training. The Journal of Psychiatric Education, 33,
3-11.

Goldberg, D.P., Steele, J.J., Johnson, A., & Smith, C.
(1982). Ability of primary care physicians to make
accurate ratings of psychiatric symptoms. Archives of
General Psychiatry, 32, 829-833.








Gordon, L.V. & Mensh, I.N. (1962). Values of medical
students at different levels of training. Journal of
Educational Psvchologv, 53, 48-51.

Greenfield, S., Kaplan, S., & Ware, J.E. (1985).
Expanding patient involvement in care: effects on
patient outcomes. Annals of Internal Medicine, 102,
520-528.

Hall, J.A., Roter, D.L., & Katz, N.R. (1988). Meta-
analysis of correlates of provider behavior in medical
encounters. Medical Care, 26, 657-675.

Hammond, K.R., & Kern, F., Jr. (1959). Teaching
comprehensive medical care. Cambridge: Harvard
University Press.

Heszen-Klemens, I. & Lapinska, E. (1984). Doctor-
patient interaction, patients' health behavior and
effects of treatment. Social Science in Medicine, 1,
9-16.

Kaplan, S.H., Greenfield, S., & Ware, J. (1989). Impact
of the doctor-patient relationship on the outcomes of
chronic disease. In Stewart, M. & Roter, D. (Eds.)
Communicating with medical patients. Newbury Park:
Sage.

Kaplan, C., Lipkin, M., & Gordon, G.H. (1988).
Somatization in primary care: patients with
unexplained and vexing medical complaints. Journal of
General Internal Medicine, 1, 177-190.

Kern, D.E., Grayson, M., Barker, L.R., Roca, R.P.,
Cole, K.A., Roter, D., & Golden, A.S. (1989).
Residency training in interviewing skills and the
psychosocial domain of medical practice. Journal of
General Internal Medicine, 4, 421-431.

Larsen, K.M., & Smith, C.K. (1981). Assessment of
nonverbal communication inthe patient-physician
interview. Family Practice, 12, 481-485.

Lazare, A. (1989). Three functions of the interview.
In Lazare, A. (Ed.), Outpatient psychiatry: diagnosis
and treatment. Baltimore: Williams & Wilkins,
153-157.








Lazare, A., Eisenthal, S., Frank, A., & Stoeckle, J.D.
(1976). Studies on a negotiated approach to
patienthood. In Gallagher, E.B. (Ed.), The doctor-
Datient relationship in the chanaina health scene (pp.
78-183). Washington D.C.: U.S. Department of Health,
Education, and Welfare, Government Printing Office.

Lazare, A., Lipkin, M., & Putnam, S.M. (1990). The
functions of the medical interview. In Lipkin, M.,
Putnam, S.M. & Lazare, A. (Eds.) The medical
interview. New York: Springer-Verlag.

Levinson, W., Kaufman, K., & Dunn, P.M. (1990). House
Staff attitudes toward psychosocial aspects of patient
care. Teaching and Learnina in Medicine, 2, 89-94.

Lipkin, M., Quill, T.E., & Napodano, R. (1984). The
medical interview: A core curriculum for residencies
in internal medicine. Annals of Internal Medicine,
iQ2, 277-284.

Maisiak, R., Meredith, R.L., Boker, J., Rider-Gordon, B.,
& Scott, L.K. (1980). Change in medical student's
attitudes toward social issues after a course in
behavioral science. The Journal of Psychiatric
Education, A, 309-312.

Markham, B. (1979). Can a behavioral sciences course
change medical students' attitudes? Journal of
Psychiatric Education, 2, 44-54.

Marteau, T.M., Humphrey, C., Matoon, G., Kidd, J.,
Lloyd, M., & Horder, J. (1991). Factors influencing
the communication skills first-year clinical medical
students. Medical Education, 25, 127-134.

McGuire, W.J. (1985). Attitudes and attitude change.
In G. Lindzey & E. Aronson (Eds.), Handbook of social
psvcholocv. third edition (pp. 234-311). Reading, MA:
Addison-Wesley.

Merkel, W.T., Margolis, R.B., & Smith, R.C. (1990).
Teaching humanistic and psychosocial aspects of care:
Current practices and attitudes. Journal of General
Internal Medicine, 34-41.

Merton, R., Reader, G.G., & Kendall, P.L. (1957). The
student physician. Cambridge: Harvard University
Press.








Orth, J.E., Stiles, W.B., Scherwitz, L., & Hennrikus, D.
(1987). Patient exposition and provider explanation
in routine interviews and hypertensive patients'
blood pressure control. Health Psychology, I, 29-42.

Ostrom, T.M. (1984). Attitude theory. In R.J. Corsini
(Ed.), Wilev Encyclopedia of psvchologv. Volume I
(pp. ). New York: Wiley.
Panel on the General Professional Education of the
Physician and College Preparation for Medicine.
(1984). Physicians for the twenty-first century, the
GPEP report. Washington, DC: Association for
American Medical Colleges.

Parlow, J. & Rothman, A. (1974). Attitudes towards
social issues in medicine of five health science
faculties. Social Science in Medicine, ., 351-358.

Platt, F.W., & McMath, J.C. (1979). Clinical
hypocompetence: The interview. Annals of Internal
Medicine, 21, 898-902.

Reader, G.G., & Goss, M.E. (1967). Comprehensive
medical care and teaching. Ithaca: Cornell
University Press.

Reinhardt, A.M. & Gray, R.M. (1972). A social
psychological study of attitude change in physicians.
Journal of Medical Education, 42, 112-117.

Rezler, A.G. (1974). Attitude changes during medical
school: a review of the literature. Journal of
Medical Education, 4A, 1023-1030.

Rich, E.C., Crowson, T.W., & Harris, I.B. (1987). The
Diagnostic Value of the Medical History: perceptions
of internal medicine physicians. Archives of Internal
Medicine, 147, 1957-1960.

Richards, T. (1990). Chasms in communication. British
Medical Journal, 231, 1407-1408.

Rosenberg, J.I. (1965). Attitude changes in dental and
medical students during professional education.
Journal of Dental Education, 29, 399.








Rosenberg, M.J. & Hovland, C.I. (1960). Cognitive,
affective, and behavioral components of attitude. In
M.J. Rosenberg et al. (Eds.), Attitude organization
and change: an analysis of consistency among
attitude components (pp. 72-97). New Haven, CT: Yale
University Press.

Roskin, G., Carsen, M.L., Rabiner, C.J., & Marell, S.K.
(1988). Attitudes toward patients among different
mental health professional groups. Comprehensive
Psychiatry, 29, 188-194.

Roskin, G. & Marell, S.K. (1988). Differences in
attitudes toward patients among medical specialties.
International Journal of Psychiatry in Medicine, 1.,
223-233.

Roter, D.L., Cole, K.A., Kern, D.E., Barker, L.R., &
Grayson, M. (1990). An evaluation of residency
training in interviewing skills and the psychosocial
domain of medical practice. Journal of General
Internal Medicine, 5, 347-354.

Roter, D.L. & Hall, J.A. (1987). Physicians'
interviewing styles and medical information obtained
from patients. Journal of General Internal Medicine,
2, 325-329.

Roter, D.L., Hall, J.A., & Katz, N.R. (1987). Relations
between physicians' behaviors and analogue patients'
satisfaction, recall, and impressions. Medical Care,
25, 437-445.

Roter, D.L., Hall, J.A., & Katz, N.R. (1988). Patient-
physician communication: a descriptive summary of the
literature. Patient Education and Counseling, 12,
99-119.

Schon, D.A. (1983). Reflective practitioner, how
Professionals think in action. New York: Basic
Books.

Shapiro, J. (1991). Does a behavioral science rotation
change residents' self-assessed psychosocial attitudes
and behavior. Family Practice Research Journal, i,
107-119.

Simpson, M.A. (1980). Clinical psvcholinguistics: The
languages of illness and healing. New York:
Irvington.








Simpson, M., Buckman, R., Stewart, M., Maguire, P.,
Lipkin, M., Novack, D., & Till, J. (1991). Doctor-
patient communication: The Toronto consensus
statement. British Medical Journal, 303, 1385-1387.

Smith, R.C. & Hoppe, R.B. (1991). The patient's story:
Integrating the patient- and physician-centered
approaches to interviewing. Annals of Internal
Medicine, 11, 470-477.

Smith, R.C., Osborn, G, Hoppe, R.B., Lyles, J.S., Van
Egeren, L., Henry, R., Sego, D., Alguire, P., &
Stoffelmayr, B. (1991). Efficacy of a one-month
training block in psychosocial medicine for residents.
Journal of General Internal Medicine, &, 535-543.

Starfield, B., Steinwachs, D., Morris, I., Bause, G.,
Siebert, S., & Westin, C. (1979). Patient provider
agreement about problems: influence on outcome of
care. JAMA, 242, 344-346.

Stoeckle, J.D., & Billings, J.A. (1987). The history of
history-taking: The medical interview. Journal of
General Internal Medicine, 2, 119-127.

Streit, U. (1980). Attitudes towards psycho-social
factors in medicine: an appraisal of the ATSIM scale.
Medical Education, 14, 259-266.

Thurstone, L.L. & Chave, E.J. (1929). The measurement
of attitude. Chicago: University of Chicago Press.

Tinker, L.J. (1991). Attitude formation and attitude
change. Dundas, Canada: Hanna Newcombe.

Waitzkin, H. (1984). Doctor-patient communication:
Clinical implications of social scientific research.
Journal of the American Medical Association, 252,
2441-2446.

Williamson, P.R., Smith, R.C., Kern, D.K., Lipkin, M.,
Barker, L.R., Hoppe, R.B., & Florek, J. (1992). The
Medical Interview and Psychosocial Aspects of
Medicine: Block Curricula for Residents. Journal of
General Internal Medicine, 2, 235-242.

Zajonc, R.B. (1980). Feeling and thinking: preferences
need no inferences. American Psychologist, 35, 151-
175.






81

Zanna, M.P. & Rempel, J.K. (1988). Attitudes: a new
look at an old concept. In D. Bar-Tal & A.W.
Kruglanski_(Eds.), The social psvcholog of knowledge
(pp.315-334). Cambridge: Cambridge University Press.

Zeldow, P.B. & Daugherty, S.R. (1987a). The stability
and attitudinal correlates or warmth and caring in
medical students. Medical Education, 21, 353-357.

Zeldow, P.B. & Daugherty, S.R. (1987b). A brief scale to
measure humanistic attitudes toward doctor-patient
relations. Unpublished manuscript.

Zimbardo, P.G. & Leippe, M.R. (1991). The psvcholorv of
attitude change and social influence. Philadelphia:
Temple University Press.














BIOGRAPHICAL SKETCH


Joseph Matthew Behen was born and raised in Detroit,

Michigan. He attended St. Marys of Redford Grade School and

High School. He received his B.A. in psychology from the

University of Michigan in 1989. He obtained his M.S. in

clinical psychology from the University of Florida in 1992.

He is scheduled to complete his internship at Northwestern

University Medical School in August, 1994.















I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Ph osophy.


Nathan Perry, Chair
Professor of Clinical and
Health Psychology

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Phi phy.


Stephen Boggs, Cochadt
Associate Professor of
Clinical and Health
Psychology

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of ct Philosophy.


a es Rodri e
ociate P fessor of
clinical and Health
Psychology

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.


Cynthia\Belar
Profess r of Clinical and
Health Psychology














I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Phi osophy.


Lyn4 \Jtrell
Prof s r of Anatomy and Cell
Biolgy

This dissertation was submitted to the Graduate Faculty
of the College of Health Related Professions and to the
Graduate School and was accepted as partial fulfillment of
the requirements for the degree of Doctor of Philosophy.

December, 1994 "_ k______
Dean, College of Education



Dean, Graduate School
















































li 126 0855I lli ill II
3 1262 08554 864968