Citation
Role of Health Literacy in Total Knee Arthroplasty in Older Adults with Knee Osteoarthritis

Material Information

Title:
Role of Health Literacy in Total Knee Arthroplasty in Older Adults with Knee Osteoarthritis
Creator:
Bautista, Miho
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (39 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences
Clinical Investigation (IDP)
Committee Chair:
Garvan, Cynthia W.
Committee Members:
Beyth, Rebecca
Graduation Date:
8/9/2008

Subjects

Subjects / Keywords:
Arthroplasty ( jstor )
Cognitive models ( jstor )
Health care industry ( jstor )
Health literacy ( jstor )
Knee replacement arthroplasty ( jstor )
Knees ( jstor )
Logistic regression ( jstor )
Older adults ( jstor )
Osteoarthritis ( jstor )
Pain ( jstor )
Clinical Investigation (IDP) -- Dissertations, Academic -- UF
adult, aging, arthritis, arthroplasty, health, knee, literacy, old, older, osteoarthritis, replacement, surgery, utilization
City of Daytona Beach ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Medical Sciences thesis, M.S.

Notes

Abstract:
Health literacy is an important determinant of healthcare utilization. This is thought to be related to inability for an individual with limited health literacy to advocate for his/her healthcare needs. We sought to determine whether health literacy is associated with the use of total knee arthroplasty (TKA)-an effective procedure to palliate an important cause of disability in older adults-osteoarthritis of the knee. Our specific aim was to investigate the association of health literacy with the utilization of TKA among older adults with knee OA. We hypothesize that older adults with limited health literacy have a decreased utilization of TKA for treatment of knee OA. The study included 889 black and white community-dwelling men and women aged 70-79 years participating in the Health, Aging and Body Composition Study (Health ABC Study)-a 10-year longitudinal cohort study. Participants were recruited in either Memphis, TN, or Pittsburgh, PA. They were all high functioning at baseline without dementia, functional difficulties, or life-limiting cancer. Participants took health literacy (HL) assessment during their clinical visits in year 3, using the Rapid Estimate of Adult Literacy of Medicine (REALM). They were excluded from the study if they had prior history of TKA or having no knee pain, aches or stiffness in either knee for most days of the week in the past 30 days. The REALM score was dichotomized into limited HL (=REALM score 0-60) and adequate HL (= REALM score 61-66). The incidence of the first TKA surgery was assessed from year 3 through year 9 of the Health ABC Study. Descriptive statistics, chi-square test, Wilcoxon rank test and logistic regression were used to analyze the relationship of HL with TKA. Overall, participants (n=889) were 40% male, 48% lived in Memphis, 49% had annual household income less than $25,000, 37% reported their health status to be very good or excellent, 56% Caucasians, 24% had less than high school equivalency, 97% reported having a usual source of healthcare, and advice, 24% had limited health literacy(HL). Their mean age was 73.5+/-2.9 years (Mean+/-SD) and body mass index of 28.5+/-5.1 kg/m2. The overall incidence of TKA was 6.7%. The incidence of TKA varied significantly by HL: 8.1% among persons with adequate HL; and 1.9% among persons with limited HL (p=0.0014). We removed body mass index, annual household income, health status, age, depression and usual source of healthcare and advice from the logistic regression model because they had no significant relationship with TKA (p=0.1). Cognitive function was also removed from the model because it was highly collinear with HL and TKA, and had a suppressive effect on the relationship of HL with TKA. The significant association of HL with TKA persisted after adjusting for knee pain, race, education and clinical site (p=0.03). In this cohort of well-functioning older adults, limited health literacy was associated with a decreased utilization of TKA. Future interventions to prevent disability in older adults may be more successful if the role of health literacy is better understood. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.S.)--University of Florida, 2008.
Local:
Adviser: Garvan, Cynthia W.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-08-31
Statement of Responsibility:
by Miho Bautista.

Record Information

Source Institution:
UFRGP
Rights Management:
Copyright Bautista, Miho. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
8/31/2010
Classification:
LD1780 2008 ( lcc )

Downloads

This item has the following downloads:


Full Text






Demographic Characteristics

Demographics were obtained at baseline in year 1. The demographic characteristics were

age, gender, race/ethnicity (Caucasian or African American), education (
equivalency, >High school equivalency, > College Graduation with a bachelor' s degree), annual

household income (< $25,000, >$25,000 or Missing) and clinical site (Memphis, Tennessee, or

Pittsburgh, Pennsylvania), and whether they had a usual source of healthcare and advice. There

was a large number of missing observations in the annual family income. Because income is an

important socioeconomic variable that may affect the utilization of TKA, we included missing

observations as a level in the income variable. Since the demographics were obtained from

participants only once in year 1, we assumed that these demographics were constant throughout

this study.

Clinical Characteristics

Participants' clinical characteristics were obtained in year 3 and included-1) body mass

index (BMI) calculated as participants' body weight divided by their body height squared

(kilogram/meter2), 2) self-report of whether participants had a usual access to healthcare and

advice, 2) knee pain measured by a short version of the Western Ontario McMasters Universities

(WOMAC) Osteoarthritis Index, 3) depression measured by the Centers for Epidemiologic

Studies Depression 10-Item (CESD-10), and 4) cognitive function measured by the Teng' s

Modified Mini-Mental Status Exam (3MS). Knee pain, Depression and Cognitive function

variables are described below in more detail.

Knee pain

The short version of the Western Ontario McMasters Universities (WOMAC)

Osteoarthritis Index is a valid, reliable, and responsive measure of pain and function in knee

OA(38). The original long version of this index included 17 items addressing the degree of









and follow-up questionnaire. Results of this study showed that before the educational

intervention, African American patients had a significantly lower expectation of pain and

functional improvement after TKA and less willingness to consider TKA than Caucasian

patients. After receiving the educational intervention, African American patients had significant

improvement in their expectation of pain (p=0.04) and marginal improvement in their

expectation of function (p=0.09) after TKA. Caucasian patients had no significant changes in

their expectation of pain or function after TKA. Their willingness to consider TKA did not

change before and after the educational intervention in either the African American or Caucasian

group. Future studies are needed to investigate how a decision aid such as this will influence the

process of decision making for TKA in patients with low health literacy.

In this study, we found that 24% of the study participants had limited health literacy-the

REALM score of 60 and lower. According to the National Assessment of Adult Literacy

(NAAL) in 2003, as many as 43% of the U. S. adult population have a low reading proficiency,

and difficulty reading and understanding most health education materials(45). While

overwhelming evidence indicated that health literacy is a predictor of healthcare use and health

outcomes(22;23), recent evidence suggests that the problem of limited health literacy is often

unrecognized because patients are often ashamed and tend to hide their problem by avoiding

situations that could expose their lack of understanding(22). Wolf and his colleagues have

recently surveyed 313 patients in a general internal medicine clinic and found that limited health

literacy was associated with self-reporting difficulty taking medications, need for help with

health-related reading tasks and difficulty understanding and following instructions on

appointment slips (p<0.001). Among these patients, nearly half (47.8%) of patients reading at

the 3rd-grade level acknowledged having felt shame or embarrassment about their difficulties









Utilization of Total Knee Arthroplasty

Among treatment options available for knee osteoarthritis (OA), total knee arthroplasty

(TKA) is a surgical treatment to replace a diseased native knee j oint with a prosthetic j oint.

Total knee arthroplasty (TKA) ranks at or near the top among medical and surgical interventions

in its cost-effectiveness as it alleviates knee pain, restores physical function and improves the

quality of life in patients with an advanced stage of knee OA(8). Previous studies reported that

the effect sizes of short-term and long-term pain and functional outcomes of TKA ranged from

1.27 to 3.91(9). The complication rates are low-the mortality rate 0.5%, hospital readmission

rate 0.9%, pulmonary embolism 0.8%, wound infection 0.4%, pneumonia 1.4% and myocardial

infarction 0.8%(8).

In spite of its clinical benefits, TKA also has significant tradeoffs. The failure rate of

prosthetic joints is moderately high-10% at 10 years and 20% at 20 years after initial TKA.

Failed prosthetic j points will require revision surgery which has higher surgical complication rates

than initial TKA(10;11i). Although TKA is safer than other orthopedic surgeries(8), TKA is still

riskier than any other non-surgical interventions for knee OA. These non-surgical interventions

include physical therapy, analgesics and anti-inflammatory medications that are primarily

prescribed to provide temporary relief of pain and physical function with variable efficacy(12).

In addition, post-surgical care of TKA requires a prolonged period of intensive physical therapy-

from weeks to months-in restoring physical function, which makes TKA an inconvenient

treatment option to patients(12). Because of these tradeoffs of TKA, clinical guidelines

consensus groups agreed that TKA is indicated only at an advanced stage of knee OA in which

patients' function has declined to an unacceptable level and failed to respond to other non-

surgical treatments of knee OA(12;14).









reading, compared with 19.2% of patients reading at the 4th to 6th-grade level, and 6.5% of

patients reading at the 7th to 8th-grade level(46). The embarrassment or reluctance of patients to

seek more information and discuss with their doctors may delay decision making for effective

treatment for knee OA(22), which can manifest as the under-utilization of TKA as we observed

in this study.

Role of Cognitive Function in Health Literacy

We also found that participants' cognitive function-assessed by the Teng' s Mini-Mental

Status Exam (3MS)-had a suppressive effect on the association of health literacy with the

utilization of TKA. When the cognitive function was removed from the multivariate regression

model, the association between health literacy and TKA improved significantly. Previous

literature reported similar findings. For example, in a study of 3,260 community-dwelling adults

aged 65 years and older, their health literacy and cognitive function were partly collinear with

their mortality. Health literacy also predicted their mortality independently from their cognitive

function(47). These results suggest that, while the health literacy assessment shares a common

construct with the cognitive assessment, health literacy may also possess a unique construct that

is distinct from the construct of the cognitive assessment.

Dichotomization of Health Literacy

In this study, we performed multivariate regression analyses using a dichotomized health

literacy variable rather than using numerical scores of the REALM. Most statistical literature on

this topic cautioned against dichotomizing a predictor because it will introduce an extreme form

of rounding with an inevitable loss of information and power(48;49). If a normally distributed

variable is dichotomized at the median, asymptotic efficiency relative to analysis using an

ungrouped variable decreases by 35%(49). This means that dichotomizing at the median is

equivalent to losing a third of the data and serious loss of power to detect real relationships.









excluded from the study if they had: 1) prior history of TKA (n=29), or 2) having no knee pain,

aches or stiffness in either knee for most days of the week for the past 30 days (n=1,594). Eight

hundred and eighty-nine participants were included in the study. None of the participants

(n=889) reported having diagnoses other than osteoarthritis or degenerative joint disease for their

knee pain. The Institutional Review Board of the University of Florida approved this study.

Measures

Health Literacy

The Rapid Estimate of Adult Literacy in Medicine (REALM) was measured during the

clinical visits in year 3 of the Health ABC Study. The REALM is a verbal fluency test that

measures the ability to pronounce 66 common medical words and lay terms that adult primary

care patients are expected to recognize(32). These words are presented in the order of increasing

difficulty and syllable length, with one point given for each word pronounced correctly. These

words were originally chosen from patient education materials and patient intake forms used in

university-based primary care clinics. Item reduction from 125 to 66 words was based on

psychometric estimates of item difficulty and discrimination and the frequency of retained words

in written materials given to patients. It has a high concentration of words at lower difficulty

levels, which increases its discriminatory power when administered to patients with limited

reading ability. Participants are asked to read aloud as many words as they can, beginning with

the first word in the first column. When they encounter a word they cannot read, they asked to

do the best they can or say "blank" and go onto the next word. The raw score is the number of

correctly pronounced words with the dictionary pronunciation taken as the scoring standard(32).

The REALM scores range from 0 to 66 from the lowest to the highest health literacy.

Examples of words in the REALM are 'fat', 'pill', 'exercise', 'arthritis', 'medication' and

'obesity'. The REALM has a high face validity and criterion validity because it correlated well









Literature also suggests that a three-level ordinal variable is a better alternative to a dichotomized

variable in a logistic regression model(50). However, we believe that recommendations from the

statistical literature do not apply to the REALM scores used in this study because the frequency

distribution of the REALM had a significant deviation from a normal distribution with highly

negative skewness and positive kurtosis. To confirm our belief, we performed logistic regression

analyses using three different health literacy variables-numerical, three-level ordinal and

dichotomized health literacy. After performing these analyses we found that the dichotomized

variable had the best result in preserving the relationship of health literacy with TKA with the

most significant p-value (p=0.03). Thus, in this study, we performed logistic regression analyses

using dichotomized health literacy.

Strength of the Study

The strength of this study is that participants were an inception cohort of a large sample

of racially diverse (African Americans vs. Caucasians) and well-functioning older adults at

baseline. This prospective cohort design provided evidence that there is a possible causal

relationship between health literacy and the utilization of TKA. Also, the incidence of TKA was

adjudicated by medical record review, which provided a more accurate account of TKA than

self-reported incidence of TKA.

Limitations of the Study

This study also has limitations. First, the REALM was measured only once at baseline

during this study. The stability of the REALM scores over time has not been previously

reported. However, previous literature suggested that other standardized word recognition tests

similar to the REALM appeared to be fairly stable over time and less dependent on subj ects'

current cognitive function(51). Thus, when we performed the analyses, we assumed that the

REALM score was constant throughout the study. Second, the REALM measures a very narrow

































To my husband, Dad, Mom, and my brother









score was dichotomized into limited HL (=REALM score 0-60) and adequate HL (= REALM

score 61-66). The incidence of the first TKA surgery was assessed from year 3 through year 9 of

the Health ABC Study. Descriptive statistics, chi-square test, Wilcoxon rank test and logistic

regression were used to analyze the relationship of HL with TKA.

Overall, participants (n=889) were 40% male, 48% lived in Memphis, 49% had annual

household income less than $25,000, 37% reported their health status to be very good or

excellent, 56% Caucasians, 24% had less than high school equivalency, 97% reported having a

usual source of healthcare, and advice, 24% had limited health literacy. Their mean age was

73.5 & 2.9 years and body mass index of 28.515. 1 kg/m2. The overall incidence of TKA was

6.7%. The incidence of TKA varied significantly by HL-8.1% among persons with adequate HL

and 1.9% among persons with limited HL (p=0.0014). We removed body mass index, annual

household income, health status, age, depression and usual source of healthcare and advice from

the logistic regression model because they had no significant relationship with TKA (p>0.1).

Cognitive function was also removed from the model because it was highly collinear with HL

and TKA, and had a suppressive effect on the relationship of HL with TKA. The significant

association of HL with TKA persisted after adjusting for knee pain, race, education and clinical

site (p=0.03).

In this cohort of well-functioning older adults, limited health literacy was associated with

a decreased utilization of TKA. Future interventions to prevent disability in older adults may be

more successful if the role of health literacy is better understood.













5 CONCLUSIONS .............. ...............33....


LIST OF REFERENCES ................. ...............34................


BIOGRAPHICAL SKETCH .............. ...............38....










they had a significant relationship with TKA. Additionally, race, education and cognitive

function were highly collinear with health literacy and TKA. Thus, we performed logistic

regression with and without race, education and/or cognitive function to further investigate the

relationship of these variables with health literacy and TKA.

Table 3 summarizes results of the logistic regression analyses. We performed 4 logistic

regression models to investigate the relationship of race, education and cognitive function with

health literacy and TKA. The cognitive function was most collinear with health literacy because

it suppressed the association of health literacy with TKA in the first three models (Models 1, 2

and 3) with no significant association between health literacy and TKA. When the cognitive

function was removed from the model (Model 4), the association of health literacy and TKA

emerged (p=0.03). The association of knee pain and clinical site with TKA remained significant

in all four models.


Table 3-1. Frequency distribution of the participants' characteristics
Shapiro- Kolomogorov-
Wilk Smirnov
Characteri sti cs Skewness Kurtosis Min Median Max. (p-value) (p-value)
Age 0.35 -0.96 67 73 80 <0.0001 <0.01
Body mass index 0.67 0.95 16.8 27.8 52.1 <0.0001 <0.01
Knee pain 1.1 1.14 1 5 24 <0.0001 <0.01
Cognitive function -1.67 5.33 33 92 100 <0.0001 <0.01
Depression 1.05 0.89 0 4 23 <0.0001 <0.01
Health literacy -3.03 9.64 0 65 66 <0.0001 <0.01
n=889









3 8. Baron G, Dubach F, Raved P, Loge art I, Doodads M. Validation of a short form of the
Western Ontario and McMaster Universities osteoarthritis index function subscale in hip and
knee osteoarthritis. Arthritis Rheum. 2007;57:633-38.

39. Kuptniratsaikul V, Rattanachaiyanont M. Validation of a modified Thai version of the
Western Ontario and McMaster (WOMAC) osteoarthritis index for knee osteoarthritis. Clin
Rheumatol. 2007;26:1641-45.

40. Schootman M, Andresen EM, Wolinsky FD, Malmstrom TK, Miller JP, Miller DK.
Neighborhood environment and the incidence of depressive symptoms among middle-aged
African Americans. J Epidemiol Community Health. 2007;61:527-32.

41. Tombaugh TN. Test-retest reliable coefficients and 5-year change scores for the MMSE and
3MS. Arch Clin Neuropsychol. 2005;20:485-503.

42. Hawker GA, Guan J, Croxford R, Coyte PC, Glazier RH, Harvey BJ, et al. A prospective
population-based study of the predictors of undergoing total j oint arthroplasty. Arthritis
Rheum. 2006;54:3212-20.

43. Weinstein JN, Bronner KK, Morgan TS, Wennberg JE. Trends and geographic variations in
maj or surgery for degenerative diseases of the hip, knee, and spine. Health Aff (Millwood ).
2004; Suppl Web Exclusives:VAR8 1-VAR8 9.

44. Weng HH, Kaplan RM, Boscardin WJ, Maclean CH, Lee IY, Chen W, et al. Development of
a decision aid to address racial disparities in utilization of knee replacement surgery. Arthritis
Rheum. 2007;57:568-75.

45. Kutner, M., Greenberg, E., and Baer, J. A First Look at the literacy of America's adults in the
21st Century. NCES 2006-470. 2005. Jessup, MD, U.S. Department of Education. National
Assessment of Adult Literacy (NAAL). Available at:
http://nces.ed. gov/pub search/pub sinfo. asp?pubid=2006470 (Last accessed July, 2008)

46. Wolf MS, Williams MV, Parker RM, Parikh NS, Nowlan AW, Baker DW. Patients' shame
and attitudes toward discussing the results of literacy screening. J Health Commun.
2007;12:721-32.









CHAPTER 5
CONCLUSIONS

The role of health literacy in healthcare utilization remains an active area of research

because health literacy relates to patients' decision making capacity. In today's consumer-

minded healthcare system in which shared-decision making is important(21), health literacy may

present an important paradigm in understanding the access to TKA. Interventions to prevent

disability in older adults may be more successful if the role of health literacy is better

understood. Thus, further research is needed to investigate the effect of health literacy on

decision making process for TKA among older adults with knee OA-an important cause of

disability.
































O 2008 Miho Kojima Bautista









expenditures, health literacy has a significant implication to health policy and research on the

disparity in healthcare access to TKA.

For these reasons, it is important that health services research be conducted to investigate

the role of health literacy in the utilization of TKA. In this study, we investigated the association

of health literacy in the incidence of TKA among older patients with knee OA using available

data from the Health, Aging and Body Composition Study (Health ABC Study)-10-year

longitudinal cohort study of older adults. We hypothesize that limited health literacy is

associated with the decreased incidence of TKA-an effective procedure to palliate an important

cause of disability in older adults osteoarthritis (OA) of the knee.

Specific Aim and Hypothesis

Our specific aim was to investigate the association of health literacy with the utilization of

TKA among older adults with knee OA. We hypothesize that older adults with limited health

literacy have a decreased utilization of TKA for treatment of knee OA.









difficulty in accomplishing 24 activities of daily life in 3 maj or domains-1) pain (=5 items), 2)

stiffness (=2 items), and 3) function (=17 items)(39). The long-version of the WOMAC

Osteoarthritis Index has been shortened to increase its applicability in epidemiologic studies,

clinical trials and daily clinical practices. The procedures used to derive the short version of the

WOMAC Osteoarthritis Index relied on statistical approaches as well as the perceived

importance of the items in the WOMAC Osteoarthritis Index by patients and physicians(3 8). In

this study, we used the short version of the WOMAC Osteoarthrits Index that consisted of a

questionnaire asking participants whether they had pain, aches or stiffness in either knee on most

days of the week for the past 30 days with the following six activities-1) Walking on a flat

surface, 2) Going up or down stairs, 3) at night while in bed, 4) Standing upright, 5) Getting in or

out of a chair, and 6) Getting in or out of a car. Participants rated their pain with each activity by

5-point Likert scale (0-4: from no pain to extreme pain) with a possible score range from 0 to

24.

Depression

The Center for Epidemiological Studies 10-Item Form (CESD-10) is a shortened version

of a long version of the CESD. The CESD-10 is a valid, reliable and responsive measure of

depression(40). In a 1,206 well-functioning older adults in an health maintenance organization,

the CESD-10 demonstrated good reliability with test-retest correlation coefficients ranged from r

= 0.21 to r = 0.84 with an overall correlation of r = 0.71 at an average time interval of 22 days.

Principal component factor analysis demonstrated that the CESD-10 collapsed into two factors of

positive affect and negative affect(40).

In the CESD-10, a study examiner asked participants about their feelings in the past week

as described in the following 10 items: 1) I was bothered by things that usually don't bother me;

2) I had trouble keeping my mind on what I was doing; 3) I was depressed; 4) I felt that









CHAPTER 3
RESULTS

Descriptive Analyses

Overall, the participants (n=889) were 40 % male, 48 % lived in Memphis, 49% had

annual household income less than $25,000, 37 % reported their health status to be very good or

excellent, 56 % Caucasians, 24 % had less than high school equivalency, 97% reported having a

usual source of healthcare and advice and 24% had limited health literacy. The participants had

a mean age of 73.512.9 years, body mass index of 28.515.1 kg/m2, depression score (CESD-10)

of 5.314.5, cognitive function score (3MS) of 9017.9, maximum knee pain score of 6.014.2 and

health literacy score (REALM) of 60. ~111.6. The overall incidence of total arthroplasty (TKA)

in this cohort was 6.9%.

Next, we performed tests for location and normality to describe the frequency distribution

of the participants' characteristics. Specifically, we used skewness, kurtosis, minimum, median,

maximum, the Shapiro-Wilk test and the Kolomogorov-Smirnov test. Table 3-1 summarizes

results of these analyses. None of the numerical variables were normally distributed as

indicated by the Shapiro-Wilk and Kolomogorov-Smirnov tests (p-value <0.01). Particularly,

the health literacy score (REALM) had a significant deviation from a normal distribution with a

median score of 65. It also had a highly negative skewness and positive kurtosis. Based on these

findings, we performed subsequent statistical analyses using non-parametric tests.

Bivariate Analyses

We performed bivariate analyses using the chi-square test and Wilcoxon rank test to

compare the association of participants' characteristics with health literacy. Results of these

analyses are summarized in Table 2. Many of the participants' characteristics were significantly

associated with health literacy. Female gender (p=0.005), Pittsburgh site (p<0.0001), annual










Independence Center of Excellence in 2006, and completion of the M. S. program with the UF

Advanced Postgraduate Program in Clinical Investigation in 2008.

Upon completion of her M.S. program in 2008, Miho will continue an academic

appointment as a Clinical Assistant Professor with the UF Department of Aging and Geriatrics,

and staff physician with the NF/SGVHS, Geriatric Research, Education and Clinical Center

(GRECC). Miho has been married to Jun Baldoz Bautista for 12 years.









dichotomized health literacy. Because the distribution of the raw REALM scores was

significantly deviated from a normal distribution, we attempted transformation of the REALM

scores using logarithmic and inverse transformation methods. After transformation, the REALM

scores still had a significant deviation from normal. We also performed logistic regression using

the three-level ordinal health literacy, but the model did not meet the conversion criteria and had

a questionable model's validity. When we performed logistic regression using the dichotomized

health literacy, the model met the conversion criteria and had a good model fit. Thus, we used

the dichotomized health literacy in the subsequent analyses.

We used the SAS Statistical Software Version 9.1 (SAS Institute, Inc., Cary, North

Carolina) to perform statistical analyses.









available patient education materials that are mostly written at or above the 9th-grade reading

level. The fourth level (REALM score range: 61-66) approximates a 9th-grade reading level and

above. At this level, participants may be able to read most of the currently available patient

education materials. Because of a small sample size in the first and second levels of the

REALM, we combined them to create a three-level ordinal variable (REALM score range: 0-44,

45-60, and 61-66). The REALM score was also dichotomized into 2 levels by combining the

first 3 levels of the original four-level ordinal variable. We defined the 2 levels of the

dichotomized variable as follows: 1) Limited health literacy if the REALM score is 0 through

60, and 2) Adequate health literacy if the REALM score is 61 through 66.

Total Knee Arthroplasty

The incidence of total knee arthroplasty (TKA) was defined as the participants' first

hospitalization events due to TKA that they had ever had in their life. Data on hospitalization

events were available in the Detailed Hospitalization Listing dataset of the Health ABC Study

that contains information about hospitalization that resulted in at least an over-night stay in a

hospital. The study examiners initially obtained the hospitalization events from participants

during annual in-person follow-up visits and semi-annual phone call contacts with participants.

Information on the hospitalization events such as admission and discharge dates, diagnoses and

procedures performed during the hospitalization were adjudicated by reviewing medical records.

This adjudicated information on the hospitalization was subsequently entered into the Detailed

Hospital Listing dataset. At the time of this study, the hospitalization data were available from

year 1 (1997) at baseline clinical visits through November 20, 2007, when participants were last

contacted(37). In the Detailed Hospital Listing dataset, we censored hospitalization events due

to TKA if they met two criteria-1) osteoarthritis was the primary diagnosis for the

hospitalization, and 2) TKA was the primary procedure performed during hospitalization.










ROLE OF HEALTH LITERACY IN TOTAL KNEE ARTHROPLASTY
IN OLDER ADULTS WITH KNEE OSTEOARTHRITIS


















By

MIHO KOJIMA BAUTISTA


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2008











Table 3-2. Relationship of participants' characteristics with health literacy (HL) and total knee
arthroplasty (TKA)
Limited Adequate
HL HL No TKA Had TKA
Characteri sti cs (n=2 13) (n=676) P-value (n=830) (n=5 9) P-value
Age (years), mean & SD 73.6 & 2.9 73.4 & 2.8 0.3 73.5 & 2.9 73.1 & 2.4 0.42


Gender
Male (%)
Female (%)
Site
Memphis (%)
Pittsburgh (%/)
Income
< $25,000 per year (%)
> $25,000 per year (%)
Missing income (%)
Health status
Excellent/very good (%)
Good (%)
Fair or poor (%)
Race
Caucasian (%)
African American (%)
Education
>High school eq.(%)
>College grad.(%)
Have a usual source of
healthcare and advice

Body mass index,
mean & SD
Knee pain, mean & SD

Cognitive function,
mean & SD
Depression, mean & SD
Health literacy
Adequate (%)
Limited (%)
n=889


0.005



<0.0001




<0.0001




<0.0001


0.36


<0.0003




0.10




0.21


<0.0001




<0.0001


0.05




0.02


76 0.4


6 0.25


29.3 & 5.5
6.5 & 4.8


82.4 & 9.4
6.7 & 4.8


28.2 & 4.9
5.9 & 4.0


92.4 & 5.5
4.8 & 4.3


0.003
0.2


<0.0001
<0.0001


28.4 & 5.1
5.9 & 4.1

89.7 &
8.0
5.3 & 4.5


29.2 & 4.3
8.2 & 4.7


93.9 & 4.2
5.2 & 4.8


0.16
<0.0001


<0.0001
0.46


91.9


1.9 0.014









CHAPTER 1
INTTRODUCTION

Background

Scope of Problems of Knee Osteoarthritis

Osteoarthritis (OA)-also known as degenerative joint disease-is the most common

rheumatologic disease that results in significant disability among middle-aged and older persons.

In 2005, estimated prevalence of OA was 46 million: 1 in 5 adults in the United States(1).

Osteoarthritis (OA) placed third among the leading causes of disability in the U.S. with

approximately 1 million years lived in disability and $13.2 billion in annual job-related costs due

to OA. Because of the considerable impact of OA on public health, the U. S. Public Health

Service Healthy People 2010 included an agenda on decreasing health disparity associated with

OA(2).

The knee is the most common joint affected by OA. The estimated prevalence of

symptomatic knee OA is 16 % among U.S. adults over 45 years old (18.7% female, 13.5% male)

(3), and the annual incidence of symptomatic knee OA is 240 per 100,000 person years(4). Knee

OA also results in destruction of the knee j points, which appears on knee radiographs in 0.9% of

U.S. adults(5). Chronic j oint pain, aching and stiffness in the knee can cause functional

limitation in 40% of knee OA patients in performing one of these activities-walking, stair

climbing, performing heavy home chores, carrying heavy objects, house keeping, cooking and

grocery shopping. Due to the significant pain and immobility, 14% need assistance in personal

care on the daily basis(6). Furthermore, knee OA significantly decreases patients' quality of life

as 40% of them reported either poor or fair health. With the expected growth of older segment

of the U. S. adult population in the future, knee OA will continue to place enormous healthcare

and economic burdens in the U.S(7).









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

ROLE OF HEALTH LITERACY IN TOTAL KNEE ARTHROPLASTY IN OLDER ADULTS
WITH KNEE OSTEOARTHRITIS

By

Miho Kojima Bautista

August 2008

Chair: Cynthia Garvan
Major: Medical Science--Clinical Investigation

Health literacy is an important determinant of healthcare utilization. This is thought to be

related to inability for an individual with limited health literacy to advocate for his/her healthcare

needs. We sought to determine whether health literacy is associated with the use of total knee

arthroplasty (TKA)-an effective procedure to palliate an important cause of disability in older

adults-osteoarthritis of the knee.

Our specific aim was to investigate the association of health literacy with the utilization

of TKA among older adults with knee OA. We hypothesize that older adults with limited health

literacy have a decreased utilization of TKA for treatment of knee OA.

The study included 889 black and white community-dwelling men and women aged 70-

79 years participating in the Health, Aging and Body Composition Study (Health ABC Study)-a

10-year longitudinal cohort study. Participants were recruited in either Memphis, TN, or

Pittsburgh, PA. They were all high functioning at baseline without dementia, functional

difficulties, or life-limiting cancer. Participants took health literacy (HL) assessment during their

clinical visits in year 3, using the Rapid Estimate of Adult Literacy of Medicine (REALM).

They were excluded from the study if they had prior history of TKA or having no knee pain,

aches or stiffness in either knee for most days of the week in the past 30 days. The REALM









BIOGRAPHICAL SKETCH

Miho Kojima Bautista grew up in Gifu, Japan. After graduating from high school in

Japan, she moved to the United States in 1988, to pursue her goal of developing a career in

medicine. She earned her A.A. degree in nutrition from Broward Community College in Fort

Lauderdale, Florida, in 1990, and B.S. degree in nutrition from the University of Florida, in

Gainesville, Florida, in 1992. Upon her graduation from the University of Florida, she took a

position as a clinical nutritionist with the Halifax Medical Center in Daytona Beach, Florida. In

her role as a clinical nutritionist, she provided nutritional therapy and counseling to acutely ill

patients in the hospital. While she worked full-time at the Halifax Medical Center, she took

courses at the Daytona Beach Community College in Daytona Beach, Florida, to prepare for her

application to medical school.

In August, 1997, Miho entered medical school at Meharry Medical College in Nashville,

Tennessee. Upon earning her M.D. in 2001, she entered residency training in internal medicine

with the University of Florida in Gainesville, Florida. During her internal medicine residency,

she earned an Excellence in Outpatient Care award from the University of Florida Internal

Medicine Residency Program, for her achievement in providing outstanding patient care. On

completion of her internal medicine residency in 2004, she entered fellowship in geriatrics

medicine with the University of Florida. During her geriatrics medicine fellowship, Miho has

successfully competed for and obtained a Geriatric Academic Career Award from the U.S.

Department of Health and Human Services. Such an outstanding achievement afforded her

many wonderful opportunities including an academic appointment with the UF Department of

Medicine and North Florida/South Georgia Veterans Affairs Health System (NF/SGVHS) in

2005, the junior research scholar award from the UF Claude Pepper Older American









with three other standardized reading recognition tests- 1) the revised Wide Range Achievement

Test-Third Edition (WRAT-3) (r=0.88), 2) the Slosson Oral Reading Test-Revised (SORT-R)

(r-0.96), and 3) the Peabody Individual Achievement Test-Revised (PIAT-R) (r-0.97)(33). The

REALM also correlated well with other standardized health literacy tests such as the Test of

Functional Health Literacy in Adults (TOHFLA) (r=0.84)(34) and the short version of the

TOHFLA(35). The REALM has a high test-retest reliability (r-0.99) and inter-rater reliability

(r-0.99) among 100 adults(33). The 66-item version of the REALM takes 1-2 minutes to

complete by personnel with minimal training. Uniform administration and scoring of the

REALM was achieved by standardized training and direct observation of a subset of test

admini stations.

The raw REALM scores were used to derive U.S. high school grade range estimates(32).

The grade level was determined by linear regression analysis using the REALM raw scores to

predict scores on the Slossan Oral Reading Test-Revised (SORT-R), which is a widely used

national standardized test in the U.S. The REALM scores were categorized into four levels to

estimate the reading ability(33;36). The first level (REALM score range: 0 to 18) approximates

a reading level of 3rd grade or less. At this level, participants may not be able to read materials

written even at or below the 3rd-grade reading level and may need repeated oral instructions in

order to understand most patient education materials that were written at a higher reading level.

The second level (REALM score range: 19 to 44) approximates a 4th- to 6th-grade reading level.

At this level, participants may be able to read low-literacy materials, but may not be able to read

instructions written on a prescription medication label. The third level (REALM score range:

45-60) approximates a 6th to 8th-grade reading level. At this level, participants may be able to

read instructions written on a prescription medication label, but may struggle with currently










25. Scott TL, Gazmararian JA, Williams MV, Baker DW. Health literacy and preventive health
care use among Medicare enrollees in a managed care organization. Med Care. 2002;40:395-
404.

26. Polacek GN, Ramos MC, Ferrer RL. Breast cancer disparities and decision-making among
U.S. women. Patient Educ Couns. 2007;65:158-65.

27. Guerra CE, Krumholz M, Shea JA. Literacy and knowledge, attitudes and behavior about
mammography in Latinas. J Health Care Poor Underserved. 2005;16:152-66.

28. Kim SP, Knight SJ, Tomori C, Colella KM, Schoor RA, Shih L, et al. Health literacy and
shared decision making for prostate cancer patients with low socioeconomic status. Cancer
Invest. 2001;19:684-91.

29. Guerra CE, Jacobs SE, Holmes JH, Shea JA. Are physicians discussing prostate cancer
screening with their patients and why or why not? A pilot study. J Gen Intern Med.
2007;22:901-7.

30. Skinner J, Weinstein JN, Sporer SM, Wennberg JE. Racial, ethnic, and geographic disparities
in rates of knee arthroplasty among Medicare patients. N Engl J Med. 2003;349: 13 50-13 59.

31. National Institute on Aging. Health ABC Study. 2-16-2008. Available at
http://www. nia.nih. gov/ResearchInformation/ScientificResource/elhB~srpinh
m (Last Accessed July, 2008)

32. Buchbinder R, Hall S, Youd JM. Functional health literacy of patients with rheumatoid
arthritis attending a community-based rheumatology practice. J Rheumatol. 2006;33:879-86.

33. Davis TC, Michielutte R, Askov EN, Williams MV, Weiss BD. Practical assessment of adult
literacy in health care. Health Educ Behav. 1998;25:613-24.

34. Parker RM, Baker DW, Williams MV, Nurss JR. The test of functional health literacy in
adults: A new instrument for measuring patients' literacy skills. J Gen Intern Med.
1995;10:537-41.

3 5. Baker DW, Williams MV, Parker RM, Gazmararian JA, Nurss J. Development of a brief test
to measure functional health literacy. Patient Educ Couns. 1999;38:33-42.

36. Arozullah AM, Yarnold PR, Bennett CL, Soltysik RC, Wolf MS, Ferreira RM, et al.
Development and validation of a short-form, rapid estimate of adult literacy in medicine.
Med Care. 2007;45:1026-33.

37. Health, Aging and Body Composition Study Group. Hospital Prevalent Disease File
(HPREVDIS.SD2) Analysis file documentation. Version 2.2, 1-8. 11-30-2007. Available at
http ://keeptrack.ucsf.edu/ (Last accessed July, 2008)












TABLE OF CONTENTS


Page

ACKNOWLEDGMENT S .........__.. ..... .__. ...............4....


LIST OF TABLES ........._.___..... .__. ...............7....


AB S TRAC T ......_ ................. ............_........8


CHAPTER


1 INTTRODUCTION ................. ...............10.......... ......


Background ................. ... .......... ........... ...... .........10
Scope of Problems of Knee Osteoarthritis ................. ...............10...............
Utilization of Total Knee Arthroplasty ................. ...............11........... ...
Role of Health Literacy in Healthcare Utilization ................. .............................12
Statement of the Problem ................. ...............13................

Specific Aim and Hypothesis .............. ...............14....

2 M ETHODS ................. ...............15.......... .....


Participants .............. ...............15....
M measures ................. ...............16.......... .....
Health Literacy ................. ...............16.......... .....
Total Knee Arthroplasty ................. ...............18................
Demographic Characteristics............... ............1
Clinical Characteristics............... ............1
Knee pain............... ...............19..
Depression ................. ...............20.................
Cognitive function............... ...............21
Statistical Analy ses ................. ...............21.................

3 RE SULT S .............. ...............23....


Descriptive Analyses .............. ...............23....
Bivariate Analyses ................. ...............23.................
Logistic Regression Analyses ................. ...............24........... ....

4 DI SCUS SSION ................. ...............28................


Significance of the Study ................... ............ ...............28......
Role of Cognitive Function in Health Literacy .............. ...............30....
Dichotomization of Health Literacy .........._...__........_ ....._._....... ...............30
Strength of the Study ................. ...............3.. 1.............
Limitations of the Study .............. ...............3 1....









Despite of these tradeoffs of TKA, the number of TKA performed in the United States has

steadily increased over the past 15 years. Between 1993 and 2005, TKA was the second fastest

growing reason for orthopedic-related hospital admissions among Medicare enrollees, with 89%

growth in the number of hospital admissions due to TKA(15). It also is the most frequently

performed elective orthopedic surgeries in the U.S. with approximately 478,000 TKA surgeries

were performed annually, accounting for one-quarter of all orthopedic-related surgeries(16). In

2000, the Medicare spent a total of $3.2 billion on TKA(17). As the aging population grows in

the U.S. and the indication for TKA broadens with improved performance of new prosthetic

joints and advances in surgical techniques of TKA, the Medicare expenditure for TKA is

proj ected to increase by 6-fold in year 2030(18).

Because TKA is a common, elective surgery that contributes to a substantial Medicare

expenditure, the utilization of TKA has come under increasing scrutiny. Numerous studies

documented considerable racial/ethnic and regional inequality in the utilization of TKA(19).

The reason behind the socio-cultural disparity in the utilization of TKA is not well known.

However, previous literature implicated a complex interaction of patient-level factors (e.g.

treatment preference), provider factors (e.g. surgeons' practice style and physician-patient

communication style) and system-level factors (e.g. access to specialist care) as a cause of such

disparity(20;21).

Role of Health Literacy in Healthcare Utilization

During the past 15 years, health literacy has emerged as an important patient-level factor

that explains the disparity in healthcare utilization(21-23). Health literacy refers to individuals'

capacity to obtain, process, and understand basic health information and services needed to make

appropriate healthcare decisions(23). A substantial body of evidence indicates that inadequate

health literacy significantly impairs patients' knowledge, understanding and decision making









CHAPTER 2
METHOD S

Participants

Data were available from the Health ABC Study-a 10-year longitudinal cohort study of

older adults from 1997 through 2006. This study-funded by the National Institute of Aging with

the National Institute of Health-aims to investigate the epidemiology, demographics and

biometry associated with the decline in functional status and change in body composition of

older persons in transition from wellness to frailty(31). In 1997, investigators from the

University of Pittsburgh (Pittsburgh, Pennsylvania) and the University of Tennessee (Memphis,

Tennessee) recruited 3,075 English-speaking participants aged 70 to 79 years from a random

sample of white Medicare beneficiaries and all age-eligible African American community

residents to participate in the study. All participants were well functioning at baseline. The

eligibility criteria included self-reporting no difficulty in walking a distance of 400 meters or

climbing at least 10 stairs, free of life-threatening cancers, independently performing activities of

daily living and plans to living in the area for the next 3 years. Of baseline participants, 46% of

the women and 37% of the men were African Americans. Residents within designated ZIP codes

were mailed study brochures and then called on the phone to request study participation and

assess their functional status.

During the annual clinical visits in year 3 of the Health ABC Study (1999/2000), an in-

person clinical assessment of health literacy was performed in 2,5 12 participants. Of the original

3,075 participants, 563 subj ects did not complete health literacy assessment because of lack of an

in-person clinic interview (n=418), death (n=107), poor eye sight (n=14), refusal (n=13),

withdrawal from the study (n=6), and missing data (n=5)(24). Only the participants who

completed health literacy assessment (n=2512) were included in the study. Participants were









everything I did was an effort; 5) I felt hopeful about the future; 6) I felt fearful; 7) My sleep was

restless; 8) I was happy; 9) I felt lonely; and 10) I could not get going. For each of the 10 items,

the participants were asked how often they had these feeling in the past week in a 4-point Likert

scale from 0 to 3-from rarely or none of the time (<1 day) to most or all of the time (3-4 days).

The CESD-10 has a possible score range from 0 to 30 from no depression to severe depression.

Cognitive function

Teng' s Modified Mini-Mental Status Exam (3MS) is an extended version of the

Folstein's Mini-Mental Status Exam-the most widely used cognitive assessment in clinical and

research settings. The 3MS contains four additional subtests date of birth, place of birth, word

fluency and delayed recall of words. The score of the 3MS ranges from 0 to 100 points from

severe impairment to no impairment, and a modified scoring procedure permitted assignment of

partial credits on some items(41).

Statistical Analyses

First, we performed univariate analyses using summary statistics such as frequency plots,

histogram, tests for location and tests for normality to describe the characteristics of each

variable. Second, we performed bivariate analyses using the chi-square test for categorical

variables and the Wilcoxon rank test or student t-test for numerical variables to compare group

differences by the incidence of TKA. Because none of the continuous variables were normally

distributed, we performed the Wilcoxon rank test for these analyses. We also performed

bivariate analyses to compare the association of participants' characteristics with health literacy.

Third, we performed multivariate analyses using logistic regression to test the relationship of

health literacy with the incidence of TKA.

We performed separate logistic regression analyses using three different health literacy

variables- 1) numerical REALM scores, 2) three-level ordinal health literacy, and 3)









construct of health literacy-the ability to pronounce common medical words correctly.

However, health literacy will require a set of skills that are much more complex than mere word

pronunciation. The REALM has been compared with the Test of Functional Health Literacy in

Adults (TOFHLA)-another common assessment of health literacy. The TOFHLA assesses both

reading comprehension and numeracy skills related to healthcare issues. It consists of a 50-item

reading comprehension test that uses the modified Cloze procedure in which every fifth to

seventh word in a healthcare-related passage is omitted and 4 multiple-choice options are

provided. One of these 4 choices is correct and 3 of them are similar but grammatically or

contextually incorrect. The TOFHLA also contains a 17-item numeracy section that tests a

reader' s ability to comprehend directions for taking medicines, monitoring blood glucose,

keeping medical appointments and obtaining financial assistance(35). While there was a

correlation between the REALM and the TOHFLA(34;3 5), each of these health literacy

measures possesses a unique construct that is distinct from one another(32).

Our study is also limited because the results are not applicable to the general population.

Participants in this study were all healthy and well-functioning older adults at baseline, who

lived in one of two locations-Memphis, Tennessee, or Pittsburgh, Pennsylvania. Furthermore,

participants in this study had a much lower prevalence of limited health literacy (24%) than the

prevalence of limited health literacy in the general U.S. population (43%)(45). High health

literacy among this study cohort could contribute to inaccurate estimation of the incidence of

TKA.









about their own healthcare(22). However, previous studies have primarily investigated the

association of inadequate health literacy with the decreased utilization of preventive healthcare

services such as influenza vaccination and cancer screening tests for breast, prostate and colon

cancers(24-29). Evidence is currently lacking on whether there is an association between health

literacy and the utilization of TKA.

Statement of the Problem

Although the preponderance of evidence documents that inadequate health literacy impairs

patients' knowledge, understanding and decision making about their own healthcare, little is

known about health literacy's role in the utilization of total knee arthroplasty (TKA). Research in

this topic is justified at several levels of relevance to the current health care. First, examining

health literacy in regards to the utilization of TKA provides a unique health services research

model because TKA is a preference-sensitive healthcare. According to Wennberg, et al, a

preference-sensitive healthcare entails at least two valid alternative treatment options and the

choice of treatment involves trade-offs. Thus, decision making about preference-sensitive

healthcare should be based on patient' s opinions and values(20). Because having adequate

knowledge and understanding about TKA is a pre-requisite for patients to make an informed

decision, health literacy may play a critical role in the utilization of TKA. In today's consumer-

minded healthcare system in which shared-decision making is important(21), health literacy may

present an important paradigm in understanding the access to TKA. Secondly, health literacy is

clinically important because it presents an opportunity for an innovative strategy to improve

patients' knowledge, understanding and decision making for TKA. Lastly, research is needed to

understand the access to TKA in patients with limited health literacy because the maj ority of

TKA surgeries are performed on older patients who commonly use their Medicare benefits to

pay for TKA(21;30). Since the disparity in the utilization of TKA is related to federal healthcare









LIST OF REFERENCES


1. Hootman, J, Bolen, J, Helmick, C, Langmaid, G. Prevalence of doctor-diagnosed arthritis
and arthritis-attributable activitylimitation. Centers for Disease Control and Prevention-
United States, 2003-2005. 55 (40), 1089-1092. 2006. Atlanta, Centers for Disease Control
and Prevention. MMWR. Available at
http ://www.cdc.gov/mmwr/preview/mmwrhtml/mm5 540a2.htm (Last accessed July, 2008)

2. Centers for Disease Control and Prevention and National Institute of Health. Healthy People
2010. Objectives for improving health. 2008. Available at
http://www. cdc.gov/arthriti s/pub s_docs/healthy_people. htm (Last accessed July, 2008)

3. Jordan JM, Helmick CG, Renner JB, Luta G, Dragomir AD, Woodard J, et al. Prevalence of
knee symptoms and radiographic and symptomatic knee osteoarthritis in African Americans
and Caucasians: The Johnston County Osteoarthritis Project. J Rheumatol. 2007;34:172-80.

4. Oliveria SA, Felson DT, Reed JI, Cirillo PA, Walker AM. Incidence of symptomatic hand,
hip, and knee osteoarthritis among patients in a health maintenance organization. Arthritis
Rheum. 1995;38:1134-41.

5. Dillon CF, Rasch EK, Gu Q, Hirsch R. Prevalence of knee osteoarthritis in the United States:
Arthritis data from the Third National Health and Nutrition Examination Survey 1991-94. J
Rheumatol. 2006;33:2271-79.

6. Guccione AA, Felson DT, Anderson JJ, Anthony JM, Zhang Y, Wilson PW et al. The effects
of specific medical conditions on the functional limitations of elders in the Framingham
study. Am J Public Health. 1994;84:351-58.

7. Buckwalter JA, Saltzman C, Brown T. The impact of osteoarthritis: Implications for
research. Clin Orthop Relat Res. 2004; S6-15.

8. Mahomed NN, Barrett J, Katz JN, Baron JA, Wright J, Losina E. Epidemiology of total knee
replacement in the United States Medicare population. J Bone Joint Surg Am. 2005;87:1222-
28.

9. Kane RL, Saleh KJ, Wilt TJ, Bershadsky B. The functional outcomes of total knee
arthroplasty. J Bone Joint Surg Am. 2005;87:1719-24.

10. Khan F, Ng L, Gonzalez S, Hale T, Turner-Stokes L. Multidisciplinary rehabilitation
programmes following joint replacement at the hip and knee in chronic arthropathy.
Cochrane Database Syst Rev. 2008;CDOO4957.

1 1. Minns Lowe CJ, Barker KL, Dewey M, Sackley CM. Effectiveness of physiotherapy
exercise after knee arthroplasty for osteoarthritis: Systematic review and meta-analysis of
randomised controlled trials. BMJ. 2007;335:812.









household income of $25,000 or higher (p<0.0001), higher health status, (p<0.0001), Caucasian

race (p<0.0001), higher education (p<0.0001), lower body max index (p=0.0003), higher

cognitive function (p<0.0001) and less severe depression (p<0.0001) were more likely to be

associated with adequate health literacy. Age (p=0.3), having a usual source of healthcare and

advice (p=0.4), and knee pain (p=0.2) were not associated with health literacy.

Next, we performed bivariate analyses to compare the association of participants'

characteristics with total knee arthroplasty (TKA). Table 2 summarizes results of these analyses.

The incidence of TKA varied significantly by the levels of health literacy 8.1% among

participants with adequate health literacy and 1.9% among participants with limited health

literacy (p=0.014). Participants' characteristics such as Pittsburgh site (p=0.0003), Caucasian

race (p=0.05), higher education (p<0.02), higher knee pain (p<0.0001) and higher cognitive

function (p<0.0001) were also significantly associated with the incidence of TKA. Age (p=0.42),

gender (p=0.36), annual household income (p=0.1), health status (p=0.21), having a usual source

of healthcare and advice (p=0.25), body mass index (p=0. 16) and depression (p=0.46) were not

associated with the incidence of TKA.

Logistic Regression Analyses

We performed logistic regression analyses to measure the relationship of health literacy

with the incidence of TKA. Initially, we entered all the participants' characteristics into the

model to test this relationship. According to results of this model, we decided on which

variables would be included in the final model. If the variable had a significant relationship with

TKA (p<0. 1), we included the variable in the final model. We removed body mass index, annual

household income, health status, age, depression and usual source of healthcare and advice from

the final model because they had no significant relationship with TKA. Knee pain, gender, race,

education, school and cognitive function were included in the final logistic regression because










ACKNOWLEDGMENTS


I express my most sincere appreciation to my supervisory committee chair, Cynthia

Garvan, Ph.D., with the University of Florida (UF), College of Education; and my mentors

Rebecca Beyth, M.D., M. Sc., and Ron Shorr, M.D., M. S., with the University of Florida, College

of Medicine, for their support and guidance. My study would not have been possible without

their guidance and support. I would like to thank Marian Limacher, M.D., Program Director of

the Advanced Post-Graduate Program in Clinical Investigation (APPCI) with the University of

Florida, for sponsoring my education and being a role model as a successful clinician, educator,

and researcher. I am indebted to Marco Pahor, M.D., Chair of the Department of Aging and

Geriatrics, UF College of Medicine, for providing infrastructure and resources necessary in

completing this thesis. I truly value Elena Andresen, Ph.D., with the UF College of Public

Health, and Constance Uphold, Ph.D., with the UF College of Nursing, for their time and

incredible support to my career development in health services research. I thank Philip Scarpace,

Ph.D., and Nihal Tumer, Ph.D., with the UF College of Medicine; and Nannette Hoffman, M.D.,

and Bradley Bender, M.D., with the North Florida/South Georgia Veterans Affairs Health

System, for their heart-felt support and guidance. I also thank Ms. Eve Johnson, Program

Coordinator of the UF APPCI Program, and Ms. Crystal Quirin with the UF College of

Medicine, for providing administrative support. It is with much love that I thank my parents

Masayuki Kojima and Noriko Kojima; and my brother Kikuhiko Kojima for the strength and

courage they have given me to pursue my goals in my medical career. Finally, I thank my

husband, Jun Bautista, for his continued support.












LIST OF TABLES

Table page

3-1 Frequency distribution of the participants' characteristics ........._.__....... ._. ............25

3-2 Relationship of participants' characteristics with health literacy (HL) and total knee
arthroplasty (TKA)............... ...............26.

3-3 Relationship of health literacy (HL) with total knee arthroplasty--logi stic regression .....27









12. Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, et al. OARSI
recommendations for the management of hip and knee osteoarthritis, Part II: OARSI
evidence-based, expert consensus guidelines. Osteoarthritis Cartilage. 2008;16:137-62.

13. NIH Consensus Statement on total knee replacement. NIH Consensus State of Sci
Statements. 2003;20:1-34.

14. Hawker G, Wright J, Coyte P, Paul J, Dittus R, Croxford R, et al. Health-related quality of
life after knee replacement. J Bone Joint Surg Am. 1998;80: 163-73.

15. Levit, K, Ryan, K, and Elixhauser, A. HCUP facts and figures: Statistics on hospital-based
care in the united states. 2007. Rockville, M.D., Agency for Healthcare Research and
Quality, 2007. Available at http://www.hcup-us. ahrq.gov/reports.j sp (Last Accessed July,
2008)

16. DeFrances CJ, Podgornik MN. 2004 National hospital discharge survey. Adv Data. 2006; 1-
19.

17. Kane, RL, Saleh, KJ, and Wilt, TJ. Total Knee Replacement: Evidence report/technology
assessment. 04-E0006-2, 1-150. 2003. Rockville, M.D., Agency for Healthcare Research and
Quality. Available at http://www. ahrq.gov/downloads/pub/evidence/pdf/knee/ke~d (Last
Accessed July, 2008)

18. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Proj sections of primary and revision hip and
knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am.
2007;89:780-785.

19. Skinner J, Zhou W, Weinstein J. The influence of income and race on total knee arthroplasty
in the United States. J Bone Joint Surg Am. 2006;88:2159-66.

20. Wennberg JE, O'Connor AM, Collins ED, Weinstein JN. Extending the P4P agenda, part 1:
How Medicare can improve patient decision making and reduce unnecessary care. Health
Aff (Mill wood ). 2007;26: 1564-74.

21. Ibrahim SA. Racial and ethnic disparities in hip and knee j oint replacement: a review of
research in the Veterans Affairs Health Care System. J Am Acad Orthop Surg. 2007;15
Suppl 1:S87-S94.

22. Cutilli CC. Health literacy in geriatric patients: An integrative review of the literature.
Orthop Nurs. 2007;26:43-48.

23. Health literacy: report of the Council on Scientific Affairs. Ad Hoc Committee on Health
Literacy for the Council on Scientific Affairs, American Medical Association. JAMA.
1999;281:552-57.

24. Sudore RL, Mehta KM, Simonsick EM, Harris TB, Newman AB, Satterfield S, et al. Limited
literacy in older people and disparities in health and healthcare access. J Am Geriatr Soc.
2006;54:770-776.









CHAPTER 4
DISCUSSION

Significance of the Study

To our knowledge, this is the first study to demonstrate that health literacy is a predictor

of the utilization of total knee arthroplasty (TKA)-an effective surgical treatment for

osteoarthritis (OA) of the knee. Previous literature has demonstrated that socioeconomic,

geographic and clinical factors such as race, education, income, clinical site and knee pain were

the explanatory factors of the utilization of TKA(19;21;3 0;42;43). The reason behind these

differences in the utilization of TKA is not well known. Previous literature implicated a complex

interaction of patient-level factors (e.g. treatment preference), provider factors (e.g. surgeons'

practice style and physician-patient communication style) and system-level factors (e.g. access to

specialist care) as a cause of such disparity(21). However, based on our findings from this study,

we conclude that health literacy is a predictor of the utilization of TKA and has more significant

effect on the utilization of TKA than race, income or education.

Unlike race, education and income that are generally regarded as immutable, health

literacy represents a patient-level factor that can be modified or overcome by providing patients

with an innovative strategy that is tailored to their levels of health literacy. For instance, Weng

and his colleagues have recently developed a decision aid for patients with knee OA, which

consists of an educational videotape(44). This 45-minute video contains evidence-based

information on pathogenesis and treatment options for knee OA including TKA. The video also

included interviews of patients and physicians on why they chose particular medical or surgical

treatment for knee OA. These patient and physician commentaries were supplemented by

graphic presentation of data on treatment options for knee OA. Fifty-four African American and

48 Caucasian patients watched the video in a group setting, followed by a focus group meeting











Table 3-3. Relationship of health literacy (HL) with total knee arthroplasty-logistic regression
Model 1 Model 2 Model 3 Model 4
No race No education No race or education No cognitive function
Participants'
character sti cs Estimatei SD p-value Estimatei SD p-value EstimateiSD p-value Estimatei SD p-value
Limited HL -0.4510.29 0.13 -0.4010.29 0.17 -0.4310.29 0.13 -0.6310.29 0.03
Knee pain 0.1310.03 <0.0001 0.1310.03 <0.0001 0.1310.03 <0.0001 0.1310.03 <0.0001
White race --0. 1210. 16 0.46 --0. 1910. 16 0.21
Education
< High school eq. 0.05+0.30 0.36 ---- -0.1210.29
> High school eq. -0.2410.20---- -0.2210.20 0.20
Memphis site -0.4010.16 0.01 -0.4010.16 0.01 -0.3 810. 16 0.016 -0.4710.16 0.004
Cognitive function 0.0810.03 0.009 0.0810.03 0.009 0.0910.03 0.004
-2 log likelihood of
intercept 434 434 434 434
-2 log likelihood
overall 384 385 386 340
C statistics 0.75 0.75 0.74 0.74




Full Text

PAGE 1

ROLE OF HEALTH LITERACY IN TOTAL KNEE ARTHROPLASTY IN OLDER ADULTS WITH KNEE OSTEOARTHRITIS By MIHO KOJIMA BAUTISTA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008 1

PAGE 2

2008 Miho Kojima Bautista 2

PAGE 3

To my husband, Dad, Mom, and my brother 3

PAGE 4

ACKNOWLEDGMENTS I express my most sincere appreciation to my supervisory committee chair, Cynthia Garvan, Ph.D., with the University of Florida (UF), College of Education; and my mentors Rebecca Beyth, M.D., M.Sc., and Ron Shorr, M.D., M. S., with the University of Florida, College of Medicine, for their support and guidance. My study would not have been possible without their guidance and support. I w ould like to thank Marian Limach er, M.D., Program Director of the Advanced Post-Graduate Program in Clinical Investigation (APPCI) w ith the University of Florida, for sponsoring my education and being a role model as a successful clinician, educator, and researcher. I am indebted to Marco Pahor M.D., Chair of the Department of Aging and Geriatrics, UF College of Medi cine, for providing infrastructu re and resources necessary in completing this thesis. I truly value Elena A ndresen, Ph.D., with the UF College of Public Health, and Constance Uphold, Ph.D., with the UF College of Nursing, for their time and incredible support to my career de velopment in health services research. I thank Philip Scarpace, Ph.D., and Nihal Tumer, Ph.D., with the UF Co llege of Medicine; and Nannette Hoffman, M.D., and Bradley Bender, M.D., with the North Flor ida/South Georgia Veterans Affairs Health System, for their heart-felt s upport and guidance. I also thank Ms. Eve Johnson, Program Coordinator of the UF APPCI Program, and Ms. Crystal Quirin with the UF College of Medicine, for providing administrativ e support. It is with much love that I thank my parents Masayuki Kojima and Noriko Kojima; and my brother Kikuhiko Kojima for the strength and courage they have given me to pursue my goals in my medical career Finally, I thank my husband, Jun Bautista, for his continued support. 4

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS ...............................................................................................................4 LIST OF TABLES ...........................................................................................................................7 ABSTRACT .....................................................................................................................................8 CHAPTER 1 INTRODUCTION................................................................................................................. .10 Background .............................................................................................................................10 Scope of Problems of Knee Osteoarthritis ......................................................................10 Utilization of Total Knee Arthroplasty ............................................................................11 Role of Health Literacy in Healthcare Utilization ...........................................................12 Statement of the Problem ........................................................................................................13 Specific Aim and Hypothesis .................................................................................................14 2 METHODS...................................................................................................................... .......15 Participants .............................................................................................................................15 Measures .................................................................................................................................16 Health Literacy ................................................................................................................16 Total Knee Arthroplasty ..................................................................................................18 Demographic Characteristics ...........................................................................................19 Clinical Characteristics ....................................................................................................19 Knee pain ..................................................................................................................19 Depression ................................................................................................................20 Cognitive function ....................................................................................................21 Statistical Analyses .................................................................................................................21 3 RESULTS...................................................................................................................... .........23 Descriptive Analyses ..............................................................................................................23 Bivariate Analyses ..................................................................................................................23 Logistic Regression Analyses .................................................................................................24 4 DISCUSSION................................................................................................................... ......28 Significance of the Study ........................................................................................................28 Role of Cognitive Function in Health Literacy ......................................................................30 Dichotomization of Health Literacy .......................................................................................30 Strength of the Study ..............................................................................................................31 Limitations of the Study .........................................................................................................31 5

PAGE 6

5 CONCLUSIONS.................................................................................................................. ..33 LIST OF REFERENCES ...............................................................................................................34 BIOGRAPHICAL SKETCH .........................................................................................................38 6

PAGE 7

LIST OF TABLES Table page 3-1 Frequency distribution of the participants ch aracteristics................................................25 3-2 Relationship of participants characteristic s with health literacy (HL) and total knee arthroplasty (TKA)............................................................................................................. 26 3-3 Relationship of health literacy (HL) with total knee arthroplasty --logistic regression.....27 7

PAGE 8

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ROLE OF HEALTH LITERACY IN TOTAL KNEE ARTHROPLASTY IN OLDER ADULTS WITH KNEE OSTEOARTHRITIS By Miho Kojima Bautista August 2008 Chair: Cynthia Garvan Major: Medical ScienceClinical Investigation Health literacy is an important determinant of healthcare utilization. This is thought to be related to inability for an indivi dual with limited health literacy to advocate for his/her healthcare needs. We sought to determine whether health li teracy is associated w ith the use of total knee arthroplasty (TKA)an effective procedure to pallia te an important cause of disability in older adultsosteoarthritis of the knee. Our specific aim was to investigate the associat ion of health literacy with the utilization of TKA among older adults with knee OA. We hy pothesize that older adults with limited health literacy have a decreased utilization of TKA for treatment of knee OA. The study included 889 black and white community-dwelling men and women aged 7079 years participating in the Health, Aging and Body Compos ition Study (Health ABC Study)a 10-year longitudinal cohort study. Participants were recruited in either Memphis, TN, or Pittsburgh, PA. They were all high functioni ng at baseline without dementia, functional difficulties, or life-limiting cancer. Participants took health litera cy (HL) assessment during their clinical visits in year 3, using the Rapid Estim ate of Adult Literacy of Medicine (REALM). They were excluded from the study if they ha d prior history of TKA or having no knee pain, aches or stiffness in either knee for most days of the week in the past 30 days. The REALM 8

PAGE 9

score was dichotomized into limited HL (=REALM score 0-60) and adequate HL (= REALM score 61-66). The incidence of the first TKA surg ery was assessed from year 3 through year 9 of the Health ABC Study. Descriptiv e statistics, chi-square test, Wilcoxon rank test and logistic regression were used to analyze th e relationship of HL with TKA. Overall, participants (n=889) were 40% ma le, 48% lived in Memphis, 49% had annual household income less than $25,000, 37% reported th eir health status to be very good or excellent, 56% Caucasians, 24% had less than high school equivalency, 97% reported having a usual source of healthcare, and advice, 24% ha d limited health literacy. Their mean age was 73.5 2.9 years and body mass index of 28.5.1 kg/m 2 The overall incidence of TKA was 6.7%. The incidence of TKA varied significan tly by HL.1% among persons with adequate HL and 1.9% among persons with limited HL (p =0.0014). We removed body mass index, annual household income, health status, age, depression a nd usual source of healthcare and advice from the logistic regression model because they had no significant relationship with TKA (p 0.1). Cognitive function was also removed from the m odel because it was highly collinear with HL and TKA, and had a suppressive effect on the re lationship of HL with TKA. The significant association of HL with TKA pers isted after adjusting for knee pai n, race, education and clinical site (p=0.03). In this cohort of well-functioning older adults limited health literacy was associated with a decreased utilization of TKA. Future interventi ons to prevent disability in older adults may be more successful if the role of hea lth literacy is better understood. 9

PAGE 10

CHAPTER 1 INTRODUCTION Background Scope of Problems of Knee Osteoarthritis Osteoarthritis (OA)also know n as degenerative joint diseaseis the most common rheumatologic disease that results in significant disability among middle-ag ed and older persons. In 2005, estimated prevalence of OA was 46 million: 1 in 5 adults in the United States(1). Osteoarthritis (OA) placed thir d among the leading causes of di sability in the U.S. with approximately 1 million years lived in disability and $13.2 billion in annual job-related costs due to OA. Because of the considerable impact of OA on public health, the U.S. Public Health Service Healthy People 2010 included an agenda on decreasing health disparity associated with OA(2). The knee is the most common joint affect ed by OA. The estimated prevalence of symptomatic knee OA is 16 % among U.S. adults over 45 years old (18.7% female, 13.5% male) (3), and the annual incidence of symptomatic knee OA is 240 per 100,000 person years(4). Knee OA also results in destruction of the knee join ts, which appears on knee radiographs in 0.9% of U.S. adults(5). Chronic joint pain, aching and stiffness in the knee can cause functional limitation in 40% of knee OA patie nts in performing one of th ese activitieswalking, stair climbing, performing heavy home chores, carryin g heavy objects, house keeping, cooking and grocery shopping. Due to the significant pain and immobility, 14% need assistance in personal care on the daily basis(6). Furthe rmore, knee OA significantly decr eases patients quality of life as 40% of them reported either poor or fair hea lth. With the expected growth of older segment of the U.S. adult population in the future, knee OA will continue to place enormous healthcare and economic burdens in the U.S(7). 10

PAGE 11

Utilization of Total Knee Arthroplasty Among treatment options available for knee osteoarthritis (OA), tota l knee arthroplasty (TKA) is a surgical treatment to replace a diseas ed native knee joint with a prosthetic joint. Total knee arthroplasty (TKA) ranks at or near the top among medical and surgical interventions in its cost-effectiveness as it al leviates knee pain, restores ph ysical function and improves the quality of life in patients with an advanced stage of knee OA(8). Previous studies reported that the effect sizes of short-term and long-term pain and functional outcomes of TKA ranged from 1.27 to 3.91(9). The complication rates are low the mortality rate 0.5%, hospital readmission rate 0.9%, pulmonary embolism 0.8%, wound in fection 0.4%, pneumonia 1.4% and myocardial infarction 0.8%(8). In spite of its clinical benefits, TKA also has significant tradeoffs. The failure rate of prosthetic joints is moderately high% at 10 years and 20% at 20 years after initial TKA. Failed prosthetic joints will require revision surg ery which has higher surgical complication rates than initial TKA(10;11). Although TKA is safer than other orthopedic surg eries(8), TKA is still riskier than any other non-surgical interventions for knee OA. These non-surgical interventions include physical therapy, analge sics and anti-inflammatory me dications that are primarily prescribed to provide temporary re lief of pain and physical functi on with variable efficacy(12). In addition, post-surgical care of TKA requires a prolonged period of intensive physical therapy from weeks to monthsin restoring physical function, which makes TKA an inconvenient treatment option to patients(12). Because of these tradeoffs of TKA, clinical guidelines consensus groups agreed that TKA is indicated on ly at an advanced stage of knee OA in which patients function has declined to an unaccep table level and failed to respond to other nonsurgical treatments of knee OA(12;14). 11

PAGE 12

Despite of these tradeoffs of TKA, the numbe r of TKA performed in the United States has steadily increased over the past 15 years. Between 1993 and 2005, TKA was the second fastest growing reason for orthopedic-related hospital ad missions among Medicare enrollees, with 89% growth in the number of hospital admissions due to TKA(15). It also is the most frequently performed elective orthopedic surgeries in the U.S. with approximately 478,000 TKA surgeries were performed annually, accounting for one-quarter of all orthopedic-related surgeries(16). In 2000, the Medicare spent a total of $3.2 billion on TKA(17). As the aging population grows in the U.S. and the indication for TKA broadens with improved performance of new prosthetic joints and advances in surgical techniques of TKA, the Medicare expenditure for TKA is projected to increase by 6fold in year 2030(18). Because TKA is a common, elective surgery th at contributes to a substantial Medicare expenditure, the utilization of TKA has come under increasing scruti ny. Numerous studies documented considerable racial/ethnic and regional inequality in the utilization of TKA(19). The reason behind the socio-cultural disparity in the utilizatio n of TKA is not well known. However, previous literature implicated a complex interactio n of patient-level factors (e.g. treatment preference), provider factors (e.g. su rgeons practice style and physician-patient communication style) and system-level factors (e. g. access to specialist care ) as a cause of such disparity(20;21). Role of Health Literacy in Healthcare Utilization During the past 15 years, health literacy has emerged as an important patient-level factor that explains the disparity in healthcare utilizat ion(21-23). Health literac y refers to individuals capacity to obtain, process, and understand basic h ealth information and services needed to make appropriate healthcare decisions( 23). A substantial body of eviden ce indicates that inadequate health literacy significantly impairs patients knowledge, understandi ng and decision making 12

PAGE 13

about their own healthcare(22). However, prev ious studies have primarily investigated the association of inadequate health literacy with th e decreased utilization of preventive healthcare services such as influenza vaccination and cancer screening tests for breast, prostate and colon cancers(24-29). Evidence is currently lacking on wh ether there is an association between health literacy and the utili zation of TKA. Statement of the Problem Although the preponderance of evid ence documents that inadequate health literacy impairs patients knowledge, understanding and decision making about their own healthcare, little is known about health literacy s role in the utilizati on of total knee arthroplasty (TKA). Research in this topic is justified at several levels of relevance to the current health care. First, examining health literacy in regards to the utilization of TKA provides a unique heal th services research model because TKA is a preference-sensitive healthcare. According to Wennberg, et al, a preference-sensitive healthcare en tails at least two valid altern ative treatment options and the choice of treatment involves trade-offs. T hus, decision making about preference-sensitive healthcare should be based on patients opinions and values(20). Because having adequate knowledge and understanding about TKA is a pre-requ isite for patients to make an informed decision, health literacy may play a critical role in the utilization of TKA. In todays consumerminded healthcare system in which shared-decision making is important(21), health literacy may present an important paradigm in understanding the access to TKA. Secondly, health literacy is clinically important because it presents an oppor tunity for an innovative strategy to improve patients knowledge, understanding a nd decision making for TKA. Lastly, research is needed to understand the access to TKA in pati ents with limited health liter acy because the majority of TKA surgeries are performed on older patients who commonly use their Medicare benefits to pay for TKA(21;30). Since the dispar ity in the utilization of TKA is related to federal healthcare 13

PAGE 14

expenditures, health literacy has a significant implication to health policy and research on the disparity in healthcar e access to TKA. For these reasons, it is important that health services resear ch be conducted to investigate the role of health literacy in the utilization of TKA. In this st udy, we investigated the association of health literacy in the incidence of TKA among older patie nts with knee OA using available data from the Health, Aging and Body Composition Study (Health ABC Study)-year longitudinal cohort study of older adults. We hypothesize that limited health literacy is associated with the decreased incidence of TKAan effective procedure to palliate an important cause of disability in older adults osteoarthritis (OA) of the knee. Specific Aim and Hypothesis Our specific aim was to investigate the associat ion of health literacy w ith the utilization of TKA among older adults with knee OA. We hypothesize that older adults with limited health literacy have a decreased utilization of TKA for treatment of knee OA. 14

PAGE 15

CHAPTER 2 METHODS Participants Data were available from the Health ABC Studya 10-year longitu dinal cohort study of older adults from 1997 through 2006. This studyf unded by the National Institute of Aging with the National Institute of Healthaims to i nvestigate the epidemiology, demographics and biometry associated with the decline in functional status a nd change in body composition of older persons in transi tion from wellness to frailty(31). In 1997, investigators from the University of Pittsburgh (Pittsburgh, Pennsylvania) and the University of Tennessee (Memphis, Tennessee) recruited 3,075 English-speaking part icipants aged 70 to 79 years from a random sample of white Medicare beneficiaries and all age-eligible African American community residents to participate in the study. All participants were we ll functioning at baseline. The eligibility criteria incl uded self-reporting no difficulty in walking a distance of 400 meters or climbing at least 10 stairs, free of life-threateni ng cancers, independently performing activities of daily living and plans to living in the area for the next 3 years. Of baseline participants, 46% of the women and 37% of the men were African Amer icans. Residents within designated ZIP codes were mailed study brochures and then called on the phone to request st udy participation and assess their functional status. During the annual clinical vis its in year 3 of the Health ABC Study (1999/2000), an inperson clinical assessment of hea lth literacy was performed in 2,512 participants. Of the original 3,075 participants, 563 subjects did not complete health literacy assessment because of lack of an in-person clinic interview (n =418), death (n=107), poor eye sight (n=14), refusal (n=13), withdrawal from the study (n=6), and missing da ta (n=5)(24). Only the participants who completed health literacy assessment (n=2512) were included in the study. Participants were 15

PAGE 16

excluded from the study if they ha d: 1) prior history of TKA (n=29), or 2) having no knee pain, aches or stiffness in either knee for most days of the week for the past 30 days (n=1,594). Eight hundred and eighty-nine participan ts were included in the stu dy. None of the participants (n=889) reported having diagnoses other than osteoarthritis or de generative joint disease for their knee pain. The Institutional Review Board of th e University of Florid a approved this study. Measures Health Literacy The Rapid Estimate of Adult Literacy in Medicine (REALM) was measured during the clinical visits in year 3 of the Health ABC St udy. The REALM is a verbal fluency test that measures the ability to pronounce 66 common medical words and lay terms that adult primary care patients are expected to rec ognize(32). These words are presented in the order of increasing difficulty and syllable length, with one point gi ven for each word pronounced correctly. These words were originally chosen from patient education materials and patient intake forms used in university-based primary care clinics. Item reduction from 125 to 66 words was based on psychometric estimates of item difficulty and disc rimination and the frequency of retained words in written materials given to patients. It has a high concentration of words at lower difficulty levels, which increases its discriminatory power when administered to patients with limited reading ability. Participants are asked to read aloud as many words as they can, beginning with the first word in the first column. When they encounter a word they cannot read, they asked to do the best they can or say blank and go onto th e next word. The raw score is the number of correctly pronounced words with th e dictionary pronunciation taken as the scoring standard(32). The REALM scores range from 0 to 66 from th e lowest to the highest health literacy. Examples of words in the REALM are fat, pil l, exercise, arthritis, medication and obesity. The REALM has a high face validity a nd criterion validity because it correlated well 16

PAGE 17

with three other standardized reading recognition tests 1) the revised Wide Range Achievement Test-Third Edition (WRAT-3) (r= 0.88), 2) the Slosson Oral Read ing Test-Revised (SORT-R) (r=0.96), and 3) the Peabody Indivi dual Achievement Test-Revised (PIAT-R) (r=0.97)(33). The REALM also correlated well with other standardized health literacy tests such as the Test of Functional Health Literacy in Adults (TOHFLA) (r=0.84)(34) a nd the short version of the TOHFLA(35). The REALM has a high test-retest reliability (r=0.99) and in ter-rater reliability (r=0.99) among 100 adults(33). The 66-item ve rsion of the REALM takes 1-2 minutes to complete by personnel with minimal training. Uniform administration and scoring of the REALM was achieved by standardized training a nd direct observation of a subset of test administrations. The raw REALM scores were used to derive U.S. high school grade range estimates(32). The grade level was determined by linear regres sion analysis using the REALM raw scores to predict scores on the Slossan Oral Reading Test -Revised (SORT-R), which is a widely used national standardized test in the U.S. The REAL M scores were categorized into four levels to estimate the reading ability(33;36). The first level (REALM score range: 0 to 18) approximates a reading level of 3 rd grade or less. At this level, particip ants may not be able to read materials written even at or below the 3 rd -grade reading level and may need repeated oral instructions in order to understand most patient education materials that were written at a higher reading level. The second level (REALM score range: 19 to 44) approximates a 4 th to 6 th -grade reading level. At this level, participants may be able to read low-literacy materials, but may not be able to read instructions written on a prescr iption medication label. The th ird level (REALM score range: 45-60) approximates a 6 th to 8 th -grade reading level. At this level, participants may be able to read instructions written on a prescription medi cation label, but may struggle with currently 17

PAGE 18

available patient education materials that are mostly written at or above the 9 th -grade reading level. The fourth level (REALM scor e range: 61-66) approximates a 9 th -grade reading level and above. At this level, participants may be able to read most of the cu rrently available patient education materials. Because of a small samp le size in the first and second levels of the REALM, we combined them to create a threelevel ordinal variable (REALM score range: 0-44, 45-60, and 61-66). The REALM score was also di chotomized into 2 levels by combining the first 3 levels of the original four-level ordi nal variable. We defined the 2 levels of the dichotomized variable as follows: 1) Limited health literacy if the REALM score is 0 through 60, and 2) Adequate health literacy if the REALM score is 61 through 66. Total Knee Arthroplasty The incidence of total knee arthroplasty (T KA) was defined as the participants first hospitalization events due to TKA that they had ev er had in their life. Data on hospitalization events were available in the Detailed Hospitali zation Listing dataset of the Health ABC Study that contains information about hospitalization that resulted in at least an over-night stay in a hospital. The study examiners initially obtained the hos pitalization events from participants during annual in-person follow-up visits and semi-annual phone call contacts with participants. Information on the hospitalization events such as admission and discharge dates, diagnoses and procedures performed during the hospitalization were adjudicated by reviewing medical records. This adjudicated information on the hospitalization was subsequently entered into the Detailed Hospital Listing dataset. At the time of this study, the hospitalization da ta were available from year 1 (1997) at baseline clini cal visits through November 20, 2007, when participants were last contacted(37). In the Detailed Hospital Listi ng dataset, we censored hospitalization events due to TKA if they met two criteria) osteoa rthritis was the primary diagnosis for the hospitalization, and 2) TKA was the primary procedure performed during hospitalization. 18

PAGE 19

Demographic Characteristics Demographics were obtained at baseline in ye ar 1. The demographic characteristics were age, gender, race/ethnicity (Caucasian or African American), e ducation ($25,000 or Missing) a nd clinical site (Memphis, Tennessee, or Pittsburgh, Pennsylvania), and whether they had a usual source of healthcare and advice. There was a large number of missing observations in the annual family income. Because income is an important socioeconomic variable that may aff ect the utilization of TKA, we included missing observations as a level in the income variable Since the demographics were obtained from participants only once in year 1, we assumed that these demographices were constant throughout this study. Clinical Characteristics Participants clinical charact eristics were obtained in ye ar 3 and included) body mass index (BMI) calculated as participants body we ight divided by their body height squared (kilogram/meter 2 ), 2) self-report of whether participants had a usual access to healthcare and advice, 2) knee pain measured by a short versio n of the Western Ontari o McMasters Universities (WOMAC) Osteoarthritis Index, 3) depression measured by the Centers for Epidemiologic Studies Depression 10-Item (CESD-10), and 4) cognitive function measured by the Tengs Modified Mini-Mental Status Exam (3MS). Knee pain, Depression and Cognitive function variables are described below in more detail. Knee pain The short version of the Western Onta rio McMasters Universities (WOMAC) Osteoarthritis Index is a valid, reliable, and responsive measure of pain and function in knee OA(38). The original long version of this i ndex included 17 items addressing the degree of 19

PAGE 20

difficulty in accomplishing 24 activities of daily li fe in 3 major domains) pain (=5 items), 2) stiffness (=2 items), and 3) function (=17 ite ms)(39). The long-version of the WOMAC Osteoarthritis Index has been s hortened to increase its applicability in epidemiologic studies, clinical trials and daily clinical practices. The procedures used to derive the short version of the WOMAC Osteoarthritis Index relied on statistical approaches as well as the perceived importance of the items in the WOMAC Osteoarthr itis Index by patients and physicians(38). In this study, we used the short version of the WO MAC Osteoarthrits Index that consisted of a questionnaire asking participants whether they had pain, aches or stiffness in either knee on most days of the week for the past 30 days with the following six activities) Walking on a flat surface, 2) Going up or down stairs 3) at night while in bed, 4) Standing upright, 5) Getting in or out of a chair, and 6) Getting in or out of a car. Participants rated their pain with each activity by 5-point Likert scale (0-4: from no pain to extr eme pain) with a possible score range from 0 to 24. Depression The Center for Epidemiological Studies 10-It em Form (CESD-10) is a shortened version of a long version of the CESD. The CESD-10 is a valid, reliable and responsive measure of depression(40). In a 1,206 well-functioning older adults in an health maintenance organization, the CESD-10 demonstrated good reliab ility with test-retest correlati on coefficients ranged from r = 0.21 to r = 0.84 with an overall correlation of r = 0.71 at an average time interval of 22 days. Principal component factor analys is demonstrated that the CESD10 collapsed into two factors of positive affect and negative affect(40). In the CESD-10, a study examiner asked participan ts about their feelings in the past week as described in the following 10 items: 1) I was bothered by things that usually dont bother me; 2) I had trouble keeping my mi nd on what I was doing; 3) I wa s depressed; 4) I felt that 20

PAGE 21

everything I did was an effort; 5) I felt hopeful ab out the future; 6) I felt fearful; 7) My sleep was restless; 8) I was happy; 9) I fe lt lonely; and 10) I could not get going. For each of the 10 items, the participants were asked how often they had thes e feeling in the past week in a 4-point Likert scale from 0 to 3from rarely or none of the time (< 1 day) to most or all of the time (3-4 days). The CESD-10 has a possible score range from 0 to 30 from no depression to severe depression. Cognitive function Tengs Modified Mini-Mental Status Exam (3MS) is an extended version of the Folsteins Mini-Mental Status Examthe most widely used cognitive assessment in clinical and research settings. The 3MS contains four additiona l subtests date of birth, place of birth, word fluency and delayed recall of words. The scor e of the 3MS ranges from 0 to 100 points from severe impairment to no impairment, and a modi fied scoring procedure permitted assignment of partial credits on some items(41). Statistical Analyses First, we performed univariate analyses usi ng summary statistics such as frequency plots, histogram, tests for location and tests for normality to describe the characteristics of each variable. Second, we performed bivariate analys es using the chi-square test for categorical variables and the Wilcoxon rank test or student t-test for numerical variables to compare group differences by the incidence of TKA. Because none of the continuous variables were normally distributed, we performed the Wilcoxon rank test for these analyses. We also performed bivariate analyses to compare the association of pa rticipants characteristics with health literacy. Third, we performed multivariate analyses using l ogistic regression to test the relationship of health literacy with the incidence of TKA. We performed separate logistic regression analyses using thre e different health literacy variables 1) numerical REAL M scores, 2) three-level ordi nal health literacy, and 3) 21

PAGE 22

dichotomized health literacy. Because th e distribution of the raw REALM scores was significantly deviated from a normal distribution, we attempte d transformation of the REALM scores using logarithmic and inverse transfor mation methods. After transformation, the REALM scores still had a significant deviation from normal. We also performed logistic regression using the three-level ordinal health literacy, but the m odel did not meet the conv ersion criteria and had a questionable models validity. When we performed logistic regression using the dichotomized health literacy, the model met the conversion cr iteria and had a good model fit. Thus, we used the dichotomized health literacy in the subsequent analyses. We used the SAS Statistical Software Version 9.1 (SAS Institute, Inc., Cary, North Carolina) to perform statistical analyses. 22

PAGE 23

CHAPTER 3 RESULTS Descriptive Analyses Overall, the participants (n=889) were 40 % male, 48 % lived in Memphis, 49% had annual household income less than $25,000, 37 % reporte d their health status to be very good or excellent, 56 % Caucasians, 24 % had less than high school equi valency, 97% reported having a usual source of healthcare and advice and 24% ha d limited health literacy. The participants had a mean age of 73.5.9 years, body mass index of 28.5.1 kg/m 2 depression score (CESD-10) of 5.3.5, cognitive function score (3MS) of 90.9, maximum knee pain score of 6.0.2 and health literacy score (REALM) of 60.1.6. The overall incidence of total arthroplasty (TKA) in this cohort was 6.9%. Next, we performed tests for location and normality to describe the frequency distribution of the participants characteristics. Specificall y, we used skewness, kurtosis, minimum, median, maximum, the Shapiro-Wilk test and the Kolomogorov-Smirnov test. Table 3-1 summarizes results of these analyses. None of the num erical variables were normally distributed as indicated by the Shapiro-Wilk and Kolomogorov -Smirnov tests (p-value <0.01). Particularly, the health literacy score (REALM) had a signifi cant deviation from a normal distribution with a median score of 65. It also had a highly nega tive skewness and positive kurtosis. Based on these findings, we performed subsequent statistical analyses using non-parametric tests. Bivariate Analyses We performed bivariate analyses using the chi-square test and Wilcoxon rank test to compare the association of partic ipants characteristics with health literacy. Results of these analyses are summarized in Table 2. Many of the participants characteristics were significantly associated with health literacy. Female gender (p=0.005), Pittsburgh site (p<0.0001), annual 23

PAGE 24

household income of $25,000 or higher (p<0.0001) higher health status, (p<0.0001), Caucasian race (p<0.0001), higher education (p<0.0001) lower body max index (p=0.0003), higher cognitive function (p<0.0001) and less severe depression (p<0.0001) were more likely to be associated with adequate health literacy. Age (p=0.3), having a usual source of healthcare and advice (p=0.4), and knee pain (p=0.2) were not associated with health literacy. Next, we performed bivariate analyses to compare the association of participants characteristics with total knee arthroplasty (TKA). Table 2 summarizes results of these analyses. The incidence of TKA varied significantly by the levels of health literacy 8.1% among participants with adequate he alth literacy and 1.9% among part icipants with limited health literacy (p=0.014). Participants characteristics such as Pittsburgh site (p=0.0003), Caucasian race (p=0.05), higher education (p<0.02), higher knee pain (p<0.0001) and higher cognitive function (p<0.0001) were also significantly associated with the in cidence of TKA. Age (p=0.42), gender (p=0.36), annual household income (p=0.1), h ealth status (p=0.21), having a usual source of healthcare and advice (p=0.25), body mass i ndex (p=0.16) and depression (p=0.46) were not associated with the incidence of TKA. Logistic Regression Analyses We performed logistic regres sion analyses to measure the re lationship of health literacy with the incidence of TKA. Initially, we ente red all the participants characteristics into the model to test this relationship. According to results of this model, we decided on which variables would be included in the final model. If the variable had a significant relationship with TKA (p<0.1), we included the variable in the final model. We removed body mass index, annual household income, health status, age, depression a nd usual source of healthcare and advice from the final model because they had no significant re lationship with TKA. Knee pain, gender, race, education, school and cognitive function were incl uded in the final logistic regression because 24

PAGE 25

they had a significant relationship with TKA. Additionally, race, education and cognitive function were highly collinear wi th health literacy and TKA. Thus, we performed logistic regression with and without race education and/or cognitive func tion to further investigate the relationship of these variables with health literacy and TKA. Table 3 summarizes results of the logistic re gression analyses. We performed 4 logistic regression models to investigate the relationshi p of race, education a nd cognitive function with health literacy and TKA. The cognitive function was most collinear with h ealth literacy because it suppressed the association of health literacy wi th TKA in the first three models (Models 1, 2 and 3) with no significant association between health literacy and TKA. When the cognitive function was removed from the model (Model 4), the association of health literacy and TKA emerged (p=0.03). The association of knee pain a nd clinical site with TKA remained significant in all four models. Table 3-1. Frequency di stribution of the participants characteristics Characteristics Skewness Kurtosis Min Median Max. ShapiroWilk (p-value) KolomogorovSmirnov (p-value) Age 0.35 -0.96 67 73 80 <0.0001 <0.01 Body mass index 0.67 0.95 16.8 27.8 52.1 <0.0001 <0.01 Knee pain 1.1 1.14 1 5 24 <0.0001 <0.01 Cognitive function -1.67 5.33 33 92 100 <0.0001 <0.01 Depression 1.05 0.89 0 4 23 <0.0001 <0.01 Health literacy -3.03 9.64 0 65 66 <0.0001 <0.01 n=889 25

PAGE 26

26 Table 3-2. Relationship of partic ipants characteristics with hea lth literacy (HL) and total knee arthroplasty (TKA) Characteristics Limited HL (n=213) Adequate HL (n=676) P-value No TKA (n=830) Had TKA (n=59) P-value Age (years), mean SD 73.6 2.9 73.4 2.8 0.3 73.5 2.9 73.1 2.4 0.42 Gender Male (%) 29 71 94 6 Female (%) 21 79 0.005 92 8 0.36 Site Memphis (%) 34 66 97 3 Pittsburgh (%) 15 85 <0.0001 90 10 <0.0003 Income < $25,000 per year (%) 34 66 95 5 $25,000 per year (%) 10 90 92 8 Missing income (%) 30 70 <0.0001 91 9 0.10 Health status Excellent/very good (%) 19 81 Good (%) 21 79 92 8 Fair or poor (%) 37 63 <0.0001 95 5 0.21 Race Caucasian (%) 10 90 92 8 African American (%) 41 59 <0.0001 95 5 0.05 Education
PAGE 27

27 Table 3-3. Relationship of health literacy (HL) with total knee arthroplastylogistic regression Model 1 Model 2 Model 3 Model 4 No race No education No race or education No cognitive function Participants characteristics EstimateSD p-value EstimateSD p-value EstimateSD p-value EstimateSD p-value Limited HL -0.45.29 0.13 -0.40.29 0.17 -0.43.29 0.13 -0.63.29 0.03 Knee pain 0.13.03 <0.0001 0.13.03 <0.0001 0.13.03 <0.0001 0.13.03 <0.0001 White race 0.12.16 0.46 0.19.16 0.21 Education < High school eq. 0.05.30 0.36 -0.12.29 High school eq. -0.24.20 -0.22.20 0.20 Memphis site -0.40.16 0.01 -0.40.16 0.01 -0.38.16 0.016 -0.47.16 0.004 Cognitive function 0.08.03 0.009 0.08.03 0.009 0.09.03 0.004 -2 log likelihood of intercept 434 434 434 434 -2 log likelihood overall 384 385 386 340 C statistics 0.75 0.75 0.74 0.74

PAGE 28

CHAPTER 4 DISCUSSION Significance of the Study To our knowledge, this is the first study to demo nstrate that health li teracy is a predictor of the utilization of total knee arthroplasty (TKA) an effective surgical treatment for osteoarthritis (OA) of the knee. Previous li terature has demonstrated that socioeconomic, geographic and clinical factors such as race, education, income, c linical site and knee pain were the explanatory factors of the utilization of TKA(19;21;30;42;43). The reason behind these differences in the utilization of TKA is not well known. Previous literatur e implicated a complex interaction of patient-level factors (e.g. treatm ent preference), provider factors (e.g. surgeons practice style and physician-patient communication st yle) and system-level factors (e.g. access to specialist care) as a cause of such disparity(21). However, based on our findings from this study, we conclude that health literacy is a predictor of the utilizatio n of TKA and has more significant effect on the utilization of TKA than race, income or education. Unlike race, education and income that are generally regarded as immutable, health literacy represents a patient-level factor that can be modified or overc ome by providing patients with an innovative strategy that is tailored to th eir levels of health lite racy. For instance, Weng and his colleagues have recently developed a decision aid for patients with knee OA, which consists of an educational videotape(44). Th is 45-minute video contains evidence-based information on pathogenesis and treatment options for knee OA including TKA. The video also included interviews of patients and physicians on why they chose particular medical or surgical treatment for knee OA. These patient and physician commentaries were supplemented by graphic presentation of data on treatment options for knee OA. Fifty-four African American and 48 Caucasian patients watched the video in a group setting, followed by a focus group meeting 28

PAGE 29

and follow-up questionnaire. Results of this study showed that before the educational intervention, African American patients had a significantly lower expectation of pain and functional improvement after TKA and less wi llingness to consider TKA than Caucasian patients. After receiving the educational intervention, African American patients had significant improvement in their expecta tion of pain (p=0.04) and marginal improvement in their expectation of function (p=0.09) after TKA. Caucasian patients had no significant changes in their expectation of pain or function after TKA. Their wil lingness to consider TKA did not change before and after the educational interventi on in either the African American or Caucasian group. Future studies are needed to investigate how a d ecision aid such as this will influence the process of decision making for TKA in pa tients with low health literacy. In this study, we found that 24% of the study participants had limited health literacythe REALM score of 60 and lower. According to the National Assessment of Adult Literacy (NAAL) in 2003, as many as 43% of the U.S. adult population have a low reading proficiency, and difficulty reading and understanding most health education materials(45). While overwhelming evidence indicated that health literacy is a predicto r of healthcare use and health outcomes(22;23), recent evidence suggests that the problem of limited health literacy is often unrecognized because patients are often ashamed and tend to hide their problem by avoiding situations that could expose their lack of understanding(22). Wolf and his colleagues have recently surveyed 313 patients in a general internal medicine clinic and found that limited health literacy was associated with self-reporting diffi culty taking medications, need for help with health-related reading tasks and difficulty understanding and following instructions on appointment slips (p<0.001). Among these patient s, nearly half (47.8%) of patients reading at the 3 rd -grade level acknowledged having felt shame or embarrassment about their difficulties 29

PAGE 30

reading, compared with 19.2% of patients reading at the 4 th to 6th-grade level, and 6.5% of patients reading at the 7 th to 8 th -grade level(46). The embarrassment or reluctance of patients to seek more information and discuss with their doctors may delay decision making for effective treatment for knee OA(22), which can manifest as the under-utilization of TKA as we observed in this study. Role of Cognitive Function in Health Literacy We also found that participants cognitive functionassessed by the Tengs Mini-Mental Status Exam (3MS)had a suppressive effect on the association of hea lth literacy with the utilization of TKA. When the cognitive functi on was removed from the multivariate regression model, the association between health liter acy and TKA improved significantly. Previous literature reported similar findings. For exampl e, in a study of 3,260 community-dwelling adults aged 65 years and older, their health literacy and cognitive function were partly collinear with their mortality. Health literacy also predicted their mortality i ndependently from their cognitive function(47). These results suggest that, while the health literacy assessment shares a common construct with the cognitive assessment, health literacy may also possess a unique construct that is distinct from the construct of the cognitive assessment. Dichotomization of Health Literacy In this study, we performed multivariate regression analyses using a dichotomized health literacy variable rather than using numerical scores of the REALM. Most statistical literature on this topic cautioned against dichotomizing a pred ictor because it will introduce an extreme form of rounding with an inevitable loss of informa tion and power(48;49). If a normally distributed variable is dichotomized at the median, asymptotic efficiency relative to analysis using an ungrouped variable decreases by 35 %(49). This means that dichotomizing at the median is equivalent to losing a third of th e data and serious loss of power to detect real relationships. 30

PAGE 31

Literature also suggests that a three-level ordinal variable is a better altern ative to a dichotomized variable in a logistic regressi on model(50). However, we believe that recommendations from the statistical literature do not apply to the REALM scores used in this study because the frequency distribution of the REALM had a significant deviation from a normal distribution with highly negative skewness and positive kurtosis. To confirm our belief, we performed logistic regression analyses using three different health literacy variablesnumerical, three-level ordinal and dichotomized health literacy. After performing these analyses we found that the dichotomized variable had the best result in preserving the rela tionship of health literacy with TKA with the most significant p-value (p=0.03). Thus, in this study, we performed logistic regression analyses using dichotomized health literacy. Strength of the Study The strength of this study is that participants were an inception cohort of a large sample of racially diverse (African Americans vs. Ca ucasians) and well-functioning older adults at baseline. This prospective cohort design provi ded evidence that there is a possible causal relationship between health liter acy and the utilization of TKA. Also, the incidence of TKA was adjudicated by medical record review, which provided a more accurate account of TKA than self-reported incide nce of TKA. Limitations of the Study This study also has limitations. First, th e REALM was measured only once at baseline during this study. The stability of the REALM scores over time has not been previously reported. However, previous literature suggested that other standardized word recognition tests similar to the REALM appeared to be fairly stable over time and less dependent on subjects current cognitive function(51). Thus, when we performed the analyses, we assumed that the REALM score was constant throughout the study. Second, the REALM measures a very narrow 31

PAGE 32

construct of health literacythe ability to pronounce comm on medical words correctly. However, health literacy will require a set of skil ls that are much more complex than mere word pronunciation. The REALM has been compared with the Test of Functional Health Literacy in Adults (TOFHLA)another common assessment of health literacy. The TOFHLA assesses both reading comprehension and numeracy skills related to healthcare issu es. It consists of a 50-item reading comprehension test that uses the modified Cloze proc edure in which every fifth to seventh word in a healthcare-related passage is omitted and 4 multiple-choice options are provided. One of these 4 choices is correct a nd 3 of them are similar but grammatically or contextually incorrect. The TOFHLA also cont ains a 17-item numeracy section that tests a readers ability to comprehe nd directions for taking medici nes, monitoring blood glucose, keeping medical appointments a nd obtaining financial assistan ce(35). While there was a correlation between the REALM and the TOHFLA (34;35), each of thes e health literacy measures possesses a unique construct that is distinct from one another(32). Our study is also limited because the results are not applicable to the general population. Participants in this study were all healthy a nd well-functioning older adults at baseline, who lived in one of two locationsM emphis, Tennessee, or Pittsburgh, Pennsylvania. Furthermore, participants in this study had a much lower prev alence of limited health literacy (24%) than the prevalence of limited health literacy in the ge neral U.S. population (43%)(45). High health literacy among this study cohort could contribute to inaccurate estimation of the incidence of TKA. 32

PAGE 33

CHAPTER 5 CONCLUSIONS The role of health literacy in healthcare ut ilization remains an active area of research because health literacy relate s to patients decision making capacity. In todays consumerminded healthcare system in which shared-decision making is important(21), health literacy may present an important paradigm in understanding the access to TKA. In terventions to prevent disability in older adults may be more successful if the role of health literacy is better understood. Thus, further research is needed to investigate the effect of health literacy on decision making process for TKA among older a dults with knee OAan important cause of disability. 33

PAGE 34

LIST OF REFERENCES 1. Hootman, J, Bolen, J, Helmick, C, Langmaid, G. Prevalence of doc tor-diagnosed arthritis and arthritis-attributable ac tivitylimitation. Centers for Dis ease Control and Prevention United States, 2003-2005. 55 (40), 1089-1092. 2006. Atlanta, Centers for Disease Control and Prevention. MMWR. Available at http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5540a2.htm (Last accessed July, 2008) 2. Centers for Disease Control and Prevention and National Institute of Health. Healthy People 2010. Objectives for improving health. 2008. Available at http://www.cdc.gov/arthritis/ pubs_docs/healthy_people.htm (Last accessed July, 2008) 3. Jordan JM, Helmick CG, Renner JB, Luta G, Drag omir AD, Woodard J, et al. Prevalence of knee symptoms and radiographic and symptomatic knee osteoarthritis in African Americans and Caucasians: The Johnston County Osteoart hritis Project. J Rheumatol. 2007;34:172-80. 4. Oliveria SA, Felson DT, Reed JI, Cirillo PA Walker AM. Incidence of symptomatic hand, hip, and knee osteoarthritis among patients in a health maintenance organization. Arthritis Rheum. 1995;38:1134-41. 5. Dillon CF, Rasch EK, Gu Q, Hirsch R. Prevalen ce of knee osteoarthritis in the United States: Arthritis data from the Third National Hea lth and Nutrition Examination Survey 1991-94. J Rheumatol. 2006;33:2271-79. 6. Guccione AA, Felson DT, Anderson JJ, Anthony JM, Zhang Y, Wilson PW et al. The effects of specific medical conditions on the functional limitations of elders in the Framingham study. Am J Public Health. 1994;84:351-58. 7. Buckwalter JA, Saltzman C, Brown T. The impa ct of osteoarthritis: Implications for research. Clin Orthop Relat Res. 2004;S6-15. 8. Mahomed NN, Barrett J, Katz JN, Baron JA, Wright J, Losina E. Epidemiology of total knee replacement in the United States Medicar e population. J Bone Joint Surg Am. 2005;87:122228. 9. Kane RL, Saleh KJ, Wilt TJ, Bershadsky B. The functional outcomes of total knee arthroplasty. J Bone Joint Surg Am. 2005;87:1719-24. 10. Khan F, Ng L, Gonzalez S, Hale T, Turn er-Stokes L. Multidisciplinary rehabilitation programmes following joint replacement at the hip and knee in chronic arthropathy. Cochrane Database Syst Rev. 2008;CD004957. 11. Minns Lowe CJ, Barker KL, Dewey M, S ackley CM. Effectiveness of physiotherapy exercise after knee arthroplasty for osteoarthr itis: Systematic review and meta-analysis of randomised controlled tr ials. BMJ. 2007;335:812. 34

PAGE 35

12. Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, et al. OARSI recommendations for the management of hi p and knee osteoarthritis, Part II: OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage. 2008;16:137-62. 13. NIH Consensus Statement on total knee replacement. NIH Consensus State of Sci Statements. 2003;20:1-34. 14. Hawker G, Wright J, Coyte P, Paul J, Dittus R, Croxford R, et al. Health-related quality of life after knee replacement. J Bone Joint Surg Am. 1998;80:163-73. 15. Levit, K, Ryan, K, and Elixhauser, A. HCUP facts and figures: Statistics on hospital-based care in the united states. 2007. Rockville, M. D., Agency for Healthcare Research and Quality, 2007. Available at http://www.hcup-us.ahrq.gov/reports.jsp (Last Accessed July, 2008) 16. DeFrances CJ, Podgornik MN. 2004 National ho spital discharge survey. Adv Data. 2006;119. 17. Kane, RL, Saleh, KJ, and Wilt, TJ. Total Knee Replacement: Evidence report/technology assessment. 04-E0006-2, 1-150. 2003. Rockville, M. D., Agency for Healthcare Research and Quality. Available at http://www.ahrq.gov/downloads/pub/ evidence/pdf/knee/knee.pdf (Last Accessed July, 2008) 18. Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am. 2007;89:780-785. 19. Skinner J, Zhou W, Weinstein J. The influence of income and race on total knee arthroplasty in the United States. J Bone Joint Surg Am. 2006;88:2159-66. 20. Wennberg JE, O'Connor AM, Collins ED, Weinstein JN. Extending the P4P agenda, part 1: How Medicare can improve patient decision making and reduce unnecessary care. Health Aff (Millwood ). 2007;26:1564-74. 21. Ibrahim SA. Racial and ethnic disparities in hip and knee joint replacement: a review of research in the Veterans Affairs Health Care System. J Am Acad Orthop Surg. 2007;15 Suppl 1:S87-S94. 22. Cutilli CC. Health literacy in geriatric patients: An integrat ive review of the literature. Orthop Nurs. 2007;26:43-48. 23. Health literacy: report of th e Council on Scientific Affairs. Ad Hoc Committee on Health Literacy for the Council on Sc ientific Affairs, American Medical Association. JAMA. 1999;281:552-57. 24. Sudore RL, Mehta KM, Simonsick EM, Harris TB, Newman AB, Satterfield S, et al. Limited literacy in older people and disp arities in health and health care access. J Am Geriatr Soc. 2006;54:770-776. 35

PAGE 36

25. Scott TL, Gazmararian JA, Williams MV, Baker DW. Health literacy and preventive health care use among Medicare enrollees in a ma naged care organization. Med Care. 2002;40:395404. 26. Polacek GN, Ramos MC, Ferrer RL. Breast canc er disparities and decision-making among U.S. women. Patient Educ Couns. 2007;65:158-65. 27. Guerra CE, Krumholz M, Shea JA. Literacy and knowledge, attitudes and behavior about mammography in Latinas. J Health Ca re Poor Underserved. 2005;16:152-66. 28. Kim SP, Knight SJ, Tomori C, Colella KM, Schoor RA, Shih L, et al. Health literacy and shared decision making for prostate cancer pa tients with low socioeconomic status. Cancer Invest. 2001;19:684-91. 29. Guerra CE, Jacobs SE, Holmes JH, Shea JA. Are physicians discussing prostate cancer screening with their patients and why or why not? A pilot stud y. J Gen Intern Med. 2007;22:901-7. 30. Skinner J, Weinstein JN, Sporer SM, Wennberg JE. Racial, ethnic, and geographic disparities in rates of knee arthroplasty among Medica re patients. N Engl J Med. 2003;349:1350-1359. 31. National Institute on Aging. Health ABC Study. 2-16-2008. Available at http://www.nia.nih.gov/ResearchInformation/ ScientificResources/H ealthABCDescription.ht m (Last Accessed July, 2008) 32. Buchbinder R, Hall S, Youd JM. Functional heal th literacy of patients with rheumatoid arthritis attending a community-based rheumatology practice. J Rheumatol. 2006;33:879-86. 33. Davis TC, Michielutte R, Askov EN, Williams MV Weiss BD. Practical assessment of adult literacy in health care. Health Educ Behav. 1998;25:613-24. 34. Parker RM, Baker DW, Williams MV, Nurss JR. The test of functional health literacy in adults: A new instrument for measuring patients' literacy skills. J Gen Intern Med. 1995;10:537-41. 35. Baker DW, Williams MV, Parker RM, Gazmararian JA, Nurss J. Development of a brief test to measure functional health literac y. Patient Educ Couns. 1999;38:33-42. 36. Arozullah AM, Yarnold PR, Bennett CL, Solty sik RC, Wolf MS, Ferreira RM, et al. Development and validation of a short-form, ra pid estimate of adult literacy in medicine. Med Care. 2007;45:1026-33. 37. Health, Aging and Body Composition Study Group. Hospital Prevalent Disease File (HPREVDIS.SD2) Analysis file documenta tion. Version 2.2, 1-8. 11-30-2007. Available at http://keeptrack.ucsf.edu/ (Last accessed July, 2008) 36

PAGE 37

38. Baron G, Dubach F, Raved P, Loge art I, D oodads M. Validation of a short form of the Western Ontario and McMaster Universities oste oarthritis index functio n subscale in hip and knee osteoarthritis. Arthritis Rheum. 2007;57:633-38. 39. Kuptniratsaikul V, Rattanachai yanont M. Validation of a modi fied Thai version of the Western Ontario and McMaster (WOMAC) osteoart hritis index for knee osteoarthritis. Clin Rheumatol. 2007;26:1641-45. 40. Schootman M, Andresen EM, Wolinsky FD, Malmstrom TK, Miller JP, Miller DK. Neighbourhood environment and the incidence of depressive symptoms among middle-aged African Americans. J Epidemio l Community Health. 2007;61:527-32. 41. Tombaugh TN. Test-retest reliable coefficients and 5-year change scor es for the MMSE and 3MS. Arch Clin Neuropsychol. 2005;20:485-503. 42. Hawker GA, Guan J, Croxford R, Coyte PC, Glazier RH, Harvey BJ, et al. A prospective population-based study of the predictors of unde rgoing total joint arthroplasty. Arthritis Rheum. 2006;54:3212-20. 43. Weinstein JN, Bronner KK, Morgan TS, Wennberg JE. Trends and geographic variations in major surgery for degenerative diseases of the hip, knee, and spine. Health Aff (Millwood ). 2004;Suppl Web Exclusives:VAR81-VAR89. 44. Weng HH, Kaplan RM, Boscardin WJ, Maclean CH Lee IY, Chen W, et al. Development of a decision aid to address racial disparities in utilization of knee replacement surgery. Arthritis Rheum. 2007;57:568-75. 45. Kutner, M., Greenberg, E., and Baer, J. A First L ook at the literacy of America's adults in the 21st Century. NCES 2006-470. 2005. Jessup, MD, U. S. Department of Education. National Assessment of Adult Literacy (NAAL). Available at: http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2006470 (Last accessed July, 2008) 46. Wolf MS, Williams MV, Parker RM, Parikh NS, Nowlan AW, Baker DW. Patients' shame and attitudes toward discussing the results of literacy screen ing. J Health Commun. 2007;12:721-32. 37

PAGE 38

BIOGRAPHICAL SKETCH Miho Kojima Bautista grew up in Gifu, Japa n. After graduating from high school in Japan, she moved to the United States in 1988, to pursue her goal of developing a career in medicine. She earned her A.A. degree in nutriti on from Broward Community College in Fort Lauderdale, Florida, in 1990, and B.S. degree in nutrition from the University of Florida, in Gainesville, Florida, in 1992. Upon her graduati on from the University of Florida, she took a position as a clinical nutritionist with the Halifax Medical Center in Daytona Beach, Florida. In her role as a clinical nutritionist, she provide d nutritional therapy and counseling to acutely ill patients in the hospital. While she worked fu ll-time at the Halifax Medical Center, she took courses at the Daytona Beach Community College in Daytona Beach, Florida, to prepare for her application to medical school. In August, 1997, Miho entered medical school at Meharry Medical College in Nashville, Tennessee. Upon earning her M.D. in 2001, she entered residency training in internal medicine with the University of Florida in Gainesville, Fl orida. During her internal medicine residency, she earned an Excellence in Outpatient Care aw ard from the University of Florida Internal Medicine Residency Program, for her achievement in providing outstanding patient care. On completion of her internal medicine residency in 2004, she entered fellowship in geriatrics medicine with the University of Florida. Du ring her geriatrics medicine fellowship, Miho has successfully competed for and obtained a Geriatric Academic Career Award from the U.S. Department of Health and Human Services. Such an outstanding achievement afforded her many wonderful opportunities including an academic appointment with the UF Department of Medicine and North Florida/South Georgia Vete rans Affairs Health System (NF/SGVHS) in 2005, the junior research scholar award from the UF Claude Pepper Older American 38

PAGE 39

Independence Center of Excellence in 2006, and co mpletion of the M.S. program with the UF Advanced Postgraduate Program in Clinical Investigation in 2008. Upon completion of her M.S. program in 2008, Miho will continue an academic appointment as a Clinical Assistant Professor wi th the UF Department of Aging and Geriatrics, and staff physician with the NF/SGVHS, Geriatri c Research, Education and Clinical Center (GRECC). Miho has been married to Jun Baldoz Bautista for 12 years. 39