Differences in Success Rate of Two-Stage Revision for Periprosthetic Joint Infection of the Knee Depending on the Applied Definition.

Background: Success rates of two-stage exchange arthroplasty are subject to substantial variance in the literature. An explanation is the lack of a universally accepted definition of what constitutes treatment success for periprosthetic joint infection. Therefore, the main objective of this study was to assess success rates, applying four definitions to identify the one definition that best captures "true" success from both a clinical and patient-centered perspective.

Methods: This was a retrospective study investigating the outcome of a consecutive series of 78 chronic knee periprosthetic joint infections that were treated with a two-stage revision at a single center between 2017 and 2020. Kaplan-Meier survival analyses were conducted to estimate success rates.

Results: After a mean follow-up of 36 months (range, 25 to 60), 64 patients (82%) were reimplanted after a mean interim period of 90 days (range, 22 to 201). There was one patient (1%) lost to follow-up, and the overall mortality was 15% (n = 12). Kaplan-Meier survival analyses were used to estimate treatment success at two and three years: implant survival was 93% (95% confidence interval [CI]: 82 to 97) and 87% (95% CI: 73 to 94), infection control was 90% (95% CI: 79 to 95) and 87% (95% CI: 73 to 93), success according to Delphi criteria was 88% (95% CI: 78 to 94), and success according to Musculoskeletal Infection Society criteria was 65% (95% CI: 53 to 74) at both timepoints.

Conclusions: Success rates were ranging between 65 and 93% depending on the applied definition, highlighting the need for universal success criteria. This variability mainly results from the differing approaches to patients who did not undergo reimplantation or passed away within a year of surgery. These patients exemplify cases of fatal treatment failures. Consequently, the Musculoskeletal Infection Society criteria should be routinely employed to accurately report the "true" success rate.