Background/Purpose: U.S. blacks have a higher risk of revision total knee replacement (TKR) than whites, but whether this is mediated by poverty is unknown. The goal of this study was to evaluate racial disparities in TKR failure and determine whether community poverty modifies this risk.

Methods:  All black and white New York (NY) State residents enrolled in a prospective single-institution TKR registry 2007-2011 were included. Institutional registry patients were linked to the NY Statewide Planning and Research Cooperative System (SPARCS) database (1/1/07-12/31/14) to capture patients who underwent revision TKR at another institution, and the reasons for revision. Patients were linked by geocoded addresses to their residential census tract (CT). Cox regression was used to assess predictors of TKR revision.  Next, multivariable logistic regression was used to analyze predictors of TKR failure, defined in two ways: 1. TKR revision in NY State 2 years or less after the initial surgery or; 2. KOOS Quality of Life (QOL) that worsened or improved less than or equal to 7.5 points, or HSS Satisfaction Survey QOL rating of no improvement or worsening 2 years after surgery.  Logistic regression was used to estimate the interaction between percent of the CT under the poverty line (CT poverty) and race on TKR failure.

Results: 4529 TKR in 4263 patients were included in the study. 137 (3.0%) required TKR revision, and 330 (17%) experienced TKR failure.  Mean age was 68.0 ± 9.8, 64.1% were female, 8.4% black. Cases came from 1687 unique census tracts and mean CT poverty was 7.7% ± 8.1% (a poverty area is defined as 20% or higher). Median follow-up in SPARCS was 63 months. TKR revisions occurred a median of 15.3 months [IQR 8.0- 28.1] after the index surgery. Causes of TKR revision were septic in 26 (19%) and aseptic in 111 (81%) including mechanical failure (n=104, 75.9%), fracture (n=5, 3.7%), other causes (n=2,1.5%). Compared to aseptic revisions, septic revision cases were older (67.5 vs. 62.4 years, p=0.028), had lower volume surgeons (p=0.019), and a shorter time to revision (6.3 vs. 17.6 months; p=0.018). In multivariable analysis, factors influencing the risk of TKR revision were younger age (HR 0.80 per 5 years; 95% CI 0.74 – 0.86) and disruption of the operative wound during the index surgery admission (HR 8.05; 95% CI 1.12 – 57.7). Three hundred thirty out of 1943 (17%) cases resulted in TKR failure at 2 years. In univariate analysis, risk factors for TKR failure included race, diabetes, and upper respiratory tract infection during the index surgery admission, TKR laterality, patient expectations score, and CT poverty, but in multivariable analysis only black race remained statistically significant (OR 2.18; 95% CI 1.26 – 3.79). There was no interaction between CT poverty and race on the risk of TKR failure.

Conclusion: Blacks are at higher risk than whites of TKR failure, defined as no improvement after TKR or the need for revision, and community level poverty does not modify this risk.