Racial/ethnic differences in preferences for total knee replacement surgery

OBJECTIVES: To determine whether there are ethnic differences in preferences for surgery vs. medical treatment of knee osteoarthritis (OA). STUDY DESIGN AND SETTING: Cross-sectional in-person interviews using conjoint analysis methodology, a technique often used in marketing, involved individuals making choices between alternative hypothetical scenarios for medical or surgical treatment of knee OA. One hundred ninety-three individuals over the age of 20 were recruited through random digit dialing in Harris County, TX, and 198 individuals with knee OA were recruited from a large outpatient health care provider in Houston, TX. RESULTS: African Americans were significantly less likely to chose surgery than whites (odds ratio 0.63 [0.42, 0.93]). Women and older individuals were also less likely to choose surgery (0.69 [0.51, 0.94], 0.98 [0.97, 0.99]). Larger reductions in negative symptoms with surgery significantly increased the likelihood of choosing surgery. There was no difference between the public and patients, and no effect of income level. CONCLUSIONS: Disparities in knee replacement rates among ethnic groups may be partly due to differences in preferences for surgery. Conjoint analysis was shown to be a feasible methodology for collecting preferences in health research. This methodology has great promise in contributing to our knowledge of drivers of health care decision making in individuals.