Race and socioeconomic status as predictors of utilization and need for total knee arthroplasty for knee osteoarthritis: Data from the OsteoArthritis Initiative study

Abstract

Objective: Previous studies consistently report reduced rates of utilization of total knee arthroplasty (TKA) among black U.S. adults as compared with whites. This study assessed whether differences in TKA utilization rates between blacks and whites persist after including estimates of socioeconomic status and theoretical candidacy for joint replacement surgery. This study also examined whether blacks and whites differ in rate of reaching candidacy for TKA, and whether socioeconomic factors are related to reaching VTKA. Methods: This study employed data from the OsteoArthritis Initiative study. Study participants were black and white adults enrolled in the OAI study between the ages of 49 and 79 at baseline. Study aims were achieved using a discrete survival approach. Cox-analogue proportional hazards models were employed using a log minus log link to produce hazard ratios specific to respective intervals between time points. Models were fit using General Estimating Equations. Results: Results showed that blacks were significantly less likely to undergo total knee arthroplasty than whites, even after including estimates of baseline differences in BMI and number of comorbidities, baseline need for TKA, and socioeconomic status (education and income) (e^(β)=0.50, p=0.0016). Individuals who met need for TKA criteria at baseline were significantly more likely to undergoing TKA than those who did not (e^(β)=8.25, p<0.0001). Results also revealed race not to be a significant predictor of reaching need for TKA after including estimates of baseline differences in BMI and number of comorbidities. These findings persisted even with the inclusion of socioeconomic variables. Conclusions: Findings confirmed that substantial racial differences in utilization of TKA exist. The inclusion of socioeconomic status measures accounted for only a small proportion of the difference between blacks and whites in terms of TKA utilization (e^(β)=0.41 versus e^(β)=0.50). Racial differences in progression of knee OA to virtual TKA were also found, although race became non-significant after accounting for baseline differences in BMI categories (overweight and obese). Results suggest that other factors not captured in this study differentially influence the rate of TKA utilization among black and white U.S. adults.