The accuracy of radiographic peer review in identifying primary hip and knee arthroplasties at higher risk of early aseptic revision: an observational cohort study from the JointCare registry of South Africa.

BACKGROUND AND PURPOSE: Healthcare value might be improved with radiographic peer review to evaluate surgical quality. While numerous radiographic recommendations are available, peer review differs by including expert appraisal. We aimed to assess the predictive value of radiographic peer review in identifying primary hip and knee arthroplasties at higher risk of aseptic revision. METHODS: A retrospective observational cohort study was conducted to compare aseptic revision rates with peer review scores, using data from the JointCare registry of South Africa. Matching and weighting was applied among cohorts, and revision rates were compared using Cox proportional hazards regression. RESULTS: The sample consisted of 7,218 arthroplasties performed between January 2020 and June 2023. 5,126 arthroplasties were scored as "good," 1,931 as "acceptable," and 161 as "suboptimal." "Suboptimal" arthroplasties had a significantly higher hazard ratio (HR) for aseptic revision than "good" arthroplasties (HR 9.4, 95% confidence interval [CI] 3.6-24.5). Furthermore, "suboptimal" total hip (HR 11.4, CI 2.6-49.7), total knee (HR 5.7, CI 1.3-25.6), and medial unicompartmental knee arthroplasties (HR 9.8, CI 2.0-47.4) had significantly higher HRs than "good" ones. Overall and for each procedure, "acceptable" arthroplasties had higher HRs than "good" arthroplasties; however, these results were not statistically significant. CONCLUSION: Radiographic review by experienced arthroplasty surgeons was able to identify arthroplasties at higher risk of aseptic revision. As such, radiographic review could be used for quality improvement by identifying directions for surgical improvement.