Timing and Factors Associated With Revision for Infection After Primary Total Knee Arthroplasty Based on American Joint Replacement Registry Data.

Abstract

Infection following total knee arthroplasty (TKA) remains a challenging clinical problem. Using American Joint Replacement Registry data, this study examined factors related to the incidence and timing of infection. Primary TKAs performed from January 2012 through December 2018 among patients ≥65 years of age at surgery were queried from the American Joint Replacement Registry and merged with Medicare data to enhance capture of revisions for infection. Multivariate Cox regressions incorporating patient, surgical, and institutional factors were used to produce hazard ratios (HRs) associated with revision for infection and mortality after revision for infection. Among 525,887 TKAs, 2,821 (0.54%) were revised for infection. Men had an increased risk of revision for infection at all-time intervals (≤90 days, HR = 2.06, 95% CI: 1.75-2.43, P < .0001; >90 days to 1 year, HR = 1.90, 95% CI: 1.58-2.28, P < .0001; >1 year, HR = 1.57, 95% CI: 1.37-1.79, P < .0001). TKAs performed for osteoarthritis had an increased risk of revision for infection at ≤90 days (HR = 2.01, 95% CI: 1.45-2.78, P < .0001) but not at later times. Mortality was more likely among patients who had a Charlson Comorbidity Index (CCI) ≥ 5 compared to those who had a CCI ≤ 2 (HR = 3.21, 95% CI: 1.35-7.63, P = .008). Mortality was also more likely among older patients (HR = 1.61 for each decade, 95% CI: 1.04-2.49, P = .03). Based on primary TKAs performed in the United States, men were found to have a persistently higher risk of revision for infection, while a diagnosis of osteoarthritis was associated with a significantly higher risk only during the first 90 days after surgery.