Abstract
Objective. Ethnic disparities in the use of total joint arthroplasty (TJA) may be attributed to differences in the clinical appropriateness to undergo TJA. We sought to determine if racial differences in clinical appropriateness for surgery existed among a sample of primary care clinic patients with moderately to severely symptomatic knee or hip osteoarthritis (OA).
Methods. We used the cross-sectional data of 684 patients who are potential candidates for TJA. Using a validated TJA appropriateness algorithm, an appropriateness factor was derived using the following variables: age (50–70 or > 70 yrs), Charlson comorbidity (≤ 1 or > 1), Western Ontario and McMaster Universities OA Index (WOMAC) pain and physical function, and adequacy of previous medical management. We used logistic regression to estimate the association of race with the dichotomous outcome of clinical appropriateness for TJA consideration.
Results. Sample consisted of 425 (62%) whites and 260 (38%) African Americans; 532 (78%) had knee OA and 153 (22%) had hip OA. The mean age was 64 ± 9 years and the mean body mass index was 33.6 ± 8 kg/m2. The mean overall WOMAC score was 56 ± 14 (range 30–96), suggesting moderately severe OA. There were no significant racial group differences (p = 0.3) in the proportions of those deemed clinically appropriate for TJA. After controlling for potential confounders, race was not a predictor of clinical appropriateness for TJA (odds ratio 1.2, 95% confidence interval 0.8–1.8, p = 0.3).
Conclusion. African Americans and whites were equally appropriate to be considered for TJA.
Osteoarthritis (OA) is the most prevalent form of arthritis and is among the most common chronic conditions in the United States1. Nearly 70 million Americans, approximately 1 of every 3, are affected by arthritis or musculoskeletal disease2. Specifically, knee/hip OA is among the leading causes of disability in the US3,4. There is no known cure for knee/hip OA.
Total joint arthroplasty (TJA) is an effective treatment option for endstage knee/hip OA. The evidence base for TJA for endstage knee or hip OA has been the subject of several National Institutes of Health (NIH) consensus statements and Agency for Healthcare Research and Quality-sponsored systematic reviews5–8. The substantial body of evidence of the effectiveness and safety of TJA makes it one of the most commonly performed elective surgeries in the elderly. With the aging of the US population, the use of TJA is expected to increase over the next few decades.
Numerous studies have documented racial differences in the use of knee or hip joint replacement over the past 10–15 years9–14. The reasons for the observed racial disparities remain unclear and cannot be attributed to differences in the prevalence of lower extremity OA15. Indeed, the prevalence of knee/hip OA among older African American patients is at least as high as that reported for whites1,16.
Potential mechanisms of racial differences in the use of TJA are multifactorial and may include patient-level factors (e.g., preference or health beliefs), provider-level factors (e.g., physician-patient communication style), and system-level factors (e.g., access to specialist care)17–19. A relatively less examined factor is potential racial differences in clinical eligibility for TJA at the time of evaluation in a primary care setting. If African Americans were less clinically eligible or appropriate than whites to be considered for surgery, then bridging the racial gap in receipt of TJA would have no merit.
Our objective was to determine whether racial differences in clinical appropriateness for surgery existed among a sample of primary care clinic patients with moderately to severely symptomatic knee or hip OA. We hypothesized that African Americans were less clinically appropriate for TJA than whites. We applied a previously validated algorithm to determine the clinical appropriateness or eligibility of each patient for TJA20–23.
MATERIALS AND METHODS
Patient population
Patients were enrolled from the primary care clinics of the Roudebush Veterans Affairs Medical Center (VAMC) and the Wishard Hospital, Indianapolis, Indiana. The Wishard Hospital is a county hospital with an established primary care network consisting of its primary care centers and 6 community health centers located throughout the city. Close to the county hospital, the Roudebush VAMC serves veterans from Indiana and the surrounding states.
Initial identification of potential participants has been described17. To be eligible for this study, patients had to be ≥ 50 years of age, have radiographic evidence of OA on the symptomatic joint, and have a Western Ontario and McMaster Universities OA Index (WOMAC) summary score = 30. WOMAC summary score ≥ 30 was chosen based on a report that the mean standard deviation (SD) WOMAC summary score for patients with knee OA undergoing preoperative evaluation for total knee arthroplasty was 27.6 ± 2.924. Patients who had already undergone knee/hip TJA for their knee and/or hip pain were excluded.
Between March 1, 2003, and September 30, 2006, 1478 patients were screened; 748 met the study eligibility criteria. Sixty-four patients (8.5%) declined to participate. We present cross-sectional analyses of the baseline data of the 684 patients who met the study entry criteria.
Study procedures, including written informed consent, were approved by the Indiana University Purdue University Indianapolis Institutional Review Board (IUPUI-IRB).
WOMAC disease severity
Lower extremity OA disease severity was assessed using the WOMAC index, which includes 24 items that probe pain (score range 0–20), stiffness (score range 0–8), and functional limitation (score range 0–68). The WOMAC index has been extensively validated and shown to be a reliable and responsive instrument25. Depending on the WOMAC scores, participants were classified as having mild, moderate, or severe level of symptoms. For WOMAC pain, scores of 0–8, 9–14, and 15–20 correspond to mild, moderate, and severe pain, respectively. For WOMAC functional limitation, the severity levels are 0–22 (mild), 23–45 (moderate), and 46–68 (severe).
Adequacy of medical management
Drugs (acetaminophen, nonsteroidal antiinflammatory drugs, tramadol, and narcotic analgesics) and referrals to allied health specialists (physical therapy, occupational therapy, and nutrition) were abstracted from the electronic medical record databases. Previous medical management was considered adequate if any of the following conditions were met within the 24-month period prior to study entry: (1) filled prescriptions for at least 3 different drugs (≥ 30 day supply) for hip or knee pain; or (2) filled prescriptions for 2 different drugs (≥ 30 day supply) and received a referral to one allied health specialist for hip/knee pain or weight reduction; or (3) filled a prescription for 1 drug (≥ 30 day supply) and received referrals to 2 or more allied health specialists.
If none of the conditions was met, previous medical management was considered inadequate20–23.
Medical comorbidity
Medical comorbidity, an indicator of surgical risk, was assessed using the modified Deyo-Charlson Comorbidity Index26,27. The index assigned nonzero weights to 19 conditions based on their risk of mortality28,29. The weights can take on values of 1, 2, 3, or 6 and are then summed for each patient. The modified Deyo-Charlson Comorbidity Index has been shown to predict mortality in a cohort of community-dwelling older adults attending a large primary care practice30. Depending on the total comorbidity score, participants in the study were classified either in the low (≤ 1) or high (≥ 2) surgical risk category at study entry. The relevant diagnoses were electronically abstracted from the medical record database.
Demographic/socioeconomic status (SES)
We also collected data on demographics (sex, education, employment status, income, number of insurance coverage), body mass index (BMI), symptomatic joint (hip or knee), and site of recruitment (county hospital or VAMC). Demographic information was obtained through an interviewer administered questionnaire.
Primary outcome measure
The primary outcome measure was the clinical appropriateness of TJA. Clinical appropriateness for TJA, a dichotomous outcome variable, was obtained for each patient using an algorithm20–23. The algorithm classified each patient as either “appropriate” or “not appropriate” based on the following variables: age, adequacy of previous medical management, severity of pain, and functional limitation and surgical risk. For age, the algorithm categorized patients into 2 groups: 50–70 years of age and > 7020–23. See Figure 1 for examples of appropriate and inappropriate scenarios. Appropriate characterizes someone who is severely symptomatic (pain and functional limitation) despite adequate medical management, and is also healthy enough to withstand the stress of surgery. A validation study supports the use of the appropriateness algorithm for clinical guidelines or evaluation purposes22.
Statistical analysis
Mean values and SD were calculated for continuous variables, frequencies and percentages for categorical variables. We compared baseline characteristics of African Americans and whites using chi-squared test for categorical variables and t-test for normally distributed continuous variables. Logistic regression models were fit to estimate the relationship between (self-reported) race, the primary independent variable, and clinical appropriateness for TJA. We adjusted for potential confounders including demographics, BMI, symptomatic joint, and recruitment site. We assessed all possible interactions between (self-reported) race, the primary independent variable, and covariates. The analyses were performed using SAS version 9.1.
RESULTS
Baseline characteristics of the sample
As summarized in Table 1, the sample consisted of 425 (62%) whites and 260 (38%) African Americans; 532 (78%) had knee OA and 153 (22%) had hip OA. The mean age for the entire sample was 64.4 ± 9 years; 59% were men, 64% had less than high school education, 20% were employed (either full or part-time), and 55% reported annual household income < $15,000. The mean overall WOMAC score was 56 ± 14, suggesting at least a moderately severe symptomatic OA. The 2 groups were similar with respect to educational attainment, employment status, number of insurance coverage, and joint type. For WOMAC, there were no differences in the proportions of African Americans and whites within each level of severity of pain and function. Compared to whites, the African American group had more female subjects, had more participants who reported an annual household income < $15,000, had higher BMI, and were more likely to be recruited from the county hospital clinics. A higher proportion of African Americans than whites was in the 50–70 year age bracket. Similarly, more African Americans than whites were classified by the Deyo-Charlson Comorbidity Index as belonging to the low surgical risk category. Interestingly, a smaller proportion of African Americans than whites met the preset definition of having received “adequate” medical management.
Relationship between race and the appropriateness factor
We found no significant ethnic group differences in the proportions of those deemed appropriate for TJA and those deemed inappropriate for TJA (appropriate: 25.5% African Americans vs 29.4% whites; inappropriate: 74.5% African Americans and 70.6% whites; p = 0.3). As shown in Table 2, the only factors associated with appropriateness were the SES variables (i.e., education, income, and employment).
In the multivariate models, race was not significantly associated with clinical appropriateness for TJA (Table 3). Being unemployed remained a significant correlate of appropriateness (odds ratio 2.0, 95% CI 1.2–3.3, p = 0.003). The interactions of race with education, income, and employment were nonsignificant (p > 0.2).
DISCUSSION
In this sample of patients with at least moderately severe OA, we found that African Americans and whites were equally appropriate (or alternatively, inappropriate) to be considered for TJA. The only variable associated with higher likelihood of appropriateness was being unemployed. The relationships between race and clinical appropriateness were not modified by SES variables. On the basis of clinical eligibility alone, there should be no reason why one racial group should be considered more or less appropriate to receive TJA.
Motivated by the concern that the increasing complexity of medical care was resulting in some patients not undergoing procedures that they needed, and others undergoing procedures that they did not need, Brook, et al developed the RAND/UCLA appropriateness method — a method that examines appropriateness of medical care31. Combining the available scientific evidence and expert opinion, the RAND appropriateness methodology has been effectively applied in many fields of medicine including coronary revascularization, carotid endarterectomy, and renal transplantation32–34.
In orthopedic surgery, Quintana, et al developed and validated an appropriateness algorithm for TJA21–23. In contrast to the NIH consensus guidelines5,6, the TJA appropriateness algorithm provides guidance in dealing with more than 100 possible clinical scenarios that a physician may encounter in the clinic. A subsequent validation study of 1500 patients with OA suggested a direct relationship between the TJA appropriateness algorithm and better health-related quality of life outcomes 6 months after TJA22.
In the context of disparities in TJA use, ours is the first study to highlight the importance of clinical appropriateness in understanding racial disparity in the use of arthroplasty. Our results are consistent with studies in cardiovascular surgery that examined clinical appropriateness. Oddone, et al reported that African American and white patients had a similar distribution of appropriateness scores for carotid endarterectomy35. Similarly, Conigliaro, et al noted no ethnic differences in the proportions of those who were appropriate or inappropriate for coronary revascularization procedure36. On the other hand, Epstein, et al have reported that racial differences in the clinical appropriateness were present among patients who were candidates for renal transplantation37. Perhaps the role of clinical appropriateness in procedure use varies depending on the disease state and/or the type of surgical procedure in question.
We have found that those who were unemployed, a marker of low SES, were more likely to be appropriate for TJA. The literature supports the association between low SES and poor self-rated health38–41. Indeed, in our study, participants with lower SES reported worse WOMAC pain and function (data not shown). Because OA patients with greater joint pain and worse physical function are generally more suitable to undergo surgery, we speculate that “symptom severity” (as measured by self-report) stands as a path variable in the observed relationship between unemployment and TJA appropriateness.
Although we did not observe racial difference in the appropriateness to undergo TJA, more African Americans than whites met the preset definition for inadequate previous medical management. Compared to whites, African Americans may be more reluctant to try different medical treatment options (e.g., physical/occupational therapy) at any given level of symptom severity42,43. Racial difference in symptom-reporting to the physician may also explain our finding. Two studies have suggested that African Americans communicate less effectively with their physicians than whites44,45. Physician failure to discuss other medical treatment options with African American patients more than whites may also play a role46,47.
Our study has several limitations. First, the result could only be applied to patients with OA who have at least moderately symptomatic joint and have not had TJA. If the study population included the entire spectrum of OA severity (even those who are mildly symptomatic) or those who have already undergone TJA, the finding may have been completely different. Second, for the set of variables used to define adequacy of previous medical management, we relied on electronic abstraction of the medical record database. Electronic abstraction would not identify medical treatment recommendations that the patient declined for any reason. Third, the RAND methodology itself has a limitation because it is partly based on expert opinion. Nonetheless, there are enough data to support the discriminant and predictive validity of the TJA appropriateness algorithm used in the study5,6,20–23,48. Lastly, the cross-sectional design of our study has its inherent limitation. With OA being a progressive disease, clinical appropriateness by race may diverge over time.
We found no racial differences in patients’ clinical appropriateness for TJA. In this sample of patients with moderate to severe knee/hip OA, African Americans and whites were equally appropriate (or alternatively, inappropriate) for consideration for TJA. Future studies should reexamine the issue of appropriateness of TJA for patients with OA who have already undergone the procedure.
Footnotes
-
Dr. Ang is supported in part by a Mentored Patient-Oriented Career Development Award (RR017725) from the National Center for Research Resources (NCRR), National Institutes of Health. Dr. Ibrahim is supported by grant no. K24AR055259 from the National Institute of Arthritis and Musculoskeletal and Skin Diseases.
- Accepted for publication April 29, 2009.