Search | Search by Center | Search by Source | Keywords in Title
Hausmann LR, Brandt C, Carroll C, Fenton B, Ibrahim S, Becker W, Burgess D, Wandner L, Bair M, Goulet J. Race/Ethnicity and Total Knee Arthroplasty Rates in the VA Healthcare System: A Prospective Analysis of a Large Cohort. Poster session presented at: AcademyHealth Annual Research Meeting; 2016 Jun 27; Boston, MA.
Research Objective: Long-standing racial and ethnic differences in total knee arthroplasty (TKA) utilization have been reported. However, prior epidemiological studies have not taken into account potential differences in clinical severity, have relied on short observation periods, and have not examined whether differences changed in recent years. We examined racial and ethnic differences in receipt of TKA over a 13-year period in a large cohort of Veterans diagnosed with osteoarthritis (OA) in the Veterans Affairs (VA) Healthcare System, adjusting for clinical covariates. Study Design: Veterans diagnosed with OA between 2001 and 2011 were followed from the first OA diagnosis date until the date of TKA, death, or the end of the observation period (September 30, 2013). We used Cox proportional hazards regression to examine racial and ethnic differences in the rate of TKA by the year of OA diagnosis. Analyses were controlled for potential demographic and clinical confounding factors, including the 0-10 numeric pain intensity rating scale score. Multiple imputation methods were used to account for missing data. Population Studied: The sample was drawn from the VA Musculoskeletal Disorders (MSD) cohort study, which was constructed from VA administrative and electronic clinical data sources. The MSD cohort includes information on over 4.1 million Veterans with one or more MSD diagnosis who received VA care between 2001 and 2011. For this analysis, we included those with an OA diagnosis who were aged 50 years or older at the time of the OA diagnosis; had a race/ethnicity of white (non-Hispanic), black (non-Hispanic), or Hispanic; and had an orthopedic or rheumatology clinic visit after their OA diagnosis but before the TKA (a proxy for clinical severity). We excluded those with a prior diagnosis of rheumatoid arthritis and who had a record of TKA prior to the first-identified OA diagnosis. A total of 148,844 Veterans met these criteria. Principal Findings: There were 123,956 white, 18,865 black, and 6,023 Hispanic Veterans. Women made up 2.9% of the sample. We identified 12,025 TKA procedures (8.1%) over a median observation time of 8 years. There was a significant interaction between race/ethnicity and year of OA diagnosis. Black Veterans had a significantly lower rate of TKA relative to white Veterans in all but two years (2003 and 2005) of the observation period. TKA rates did not differ significantly between Hispanic and white Veterans, or black and Hispanic Veterans. Comorbid conditions, including PTSD and depressive disorders, were associated with a lower rate of TKA. Higher pain scores at time of OA diagnosis were associated with a higher rate of TKA. Conclusions: TKA rates vary significantly by race in a large cohort of Veterans treated for OA in the VA Healthcare System. These differences, which are also found in the private sector, appear to be enduring.