A Multicenter Study of Irrigation and Debridement in Total Knee Arthroplasty Periprosthetic Joint Infection: Treatment Failure Is High

Abstract

Background: In total knee arthroplasty (TKA) periprosthetic joint infection (PJI), irrigation and debridement (I&D) with component retention is a treatment option with a wide variation in reported failure rates. The purpose of this study was to determine failure rates, outcomes, and factors that predict failure in I&D for TKA PJI.

Methods: A multicenter observational study of patients with a TKA PJI and subsequently undergoing an I&D with retention of components was conducted. The primary outcome was failure rate of I&D, where failure was defined as any subsequent surgical procedures.

Results: Two hundred sixteen cases of I&D with retention of components performed on 206 patients met inclusion criteria. The estimated long-term failure rate at 4 years was 57.4%. Time-to-event analyses revealed that the median survival time was 14.32 months. Five-year mortality was 19.9%. Multivariable modeling revealed that time symptomatic and organism were independent predictors of I&D failure. Culture-negative status had a higher hazard for failure than culture-positive patients. When primary organism and time symptomatic were selected to produce an optimized scenario for an I&D, the estimated failure rate was 39.6%.

Conclusion: I&D with retention of components has a high failure rate, and there is a high incidence of more complex procedures after this option is chosen. The patient comorbidities we investigated did not predict I&D success. Our results suggest that I&D has a limited ability to control infection in TKA and should be used selectively under optimum conditions.

Keywords: DAIR (debridement, antibiotics, and implant retention); Level III; cohort study; irrigation and debridement; outcomes; periprosthetic joint infection; total knee arthroplasty.

Copyright © 2017 Elsevier Inc. All rights reserved.

Figures

Figure 1

The overall vs optimized estimated survival probability of I&D in TKA PJI. The overall failure rate was 57.4%. (A) When the optimized case is defined as less than one week of symptoms and a non S. aureus cultured organism, the failure rate was 39.6%. The 95% confidence intervals of overall vs optimized survival probabilities do not overlap indicating a significant difference. (B) Typically, the organism is not known at the time of surgery. When the optimized case is defined as only less than one week of symptoms and organism is excluded, the failure rate was 46.8%. Confidence intervals of overall vs optimized survival probabilities do not sufficiently overlap, suggesting no difference in outcomes.