Clinical validation and accuracy testing of a radiographic decision aid for unicondylar knee arthroplasty patient selection in midterm follow-up

Purpose

The purpose of this study was to examine the predictive value concerning clinical outcome and implant survival, as well as the accuracy of individual tests of a recently published radiographic decision aid for unicondylar knee arthroplasty indication findings.

Methods

In the retrospective part of the study, 98 consecutive patients who had undergone unicondylar knee arthroplasty (Phase 3 Oxford medial UKA) were included, using revision questionnaires, as well as the Forgotten Joint Score-12 (FJS-12) and Knee Osteoarthritis Outcome Score (KOOS) and analysed for suitability of the radiographic decision aid. Inappropriate and appropriate indications were then compared concerning the clinical outcome and implant survival. The prospective part of the study assessed the accuracy of the decision aid’s radiographic tests (varus and valgus stress views, true lateral view and skyline view), and included 90 patients. Definition as appropriate for UKA procedure included medial bone-on-bone situation in varus stress views, full-thickness lateral cartilage and functional medial collateral ligament in valgus stress views, functional anterior cruciate ligament (ACL) in true lateral views and absence of lateral facet osteoarthritis with bone loss in skyline views. Pre-operative radiographic assessment with respect to the decision aid was then compared with intraoperative articular conditions. The clinical outcome was analysed using non-parametric tests (Mann–Whitney U), and revision rates were compared using the Fisher’s exact test. Accuracy assessment included calculations of the sensitivity, specificity, negative predictive value and positive predictive value. A p value < 0.05 was considered statistically significant.

Results

Appropriate unicondylar knee arthroplasty with respect to the decision aid showed a significantly lower revision rate compared to inappropriate unicondylar knee arthroplasty (7.3% vs. 50.0%, p < 0.0001), as well as higher clinical outcome scores (FJS-12: 53.13 vs. 31.25, p = 0.041 and KOOS-QDL: 68.75 vs. 50.0, p = 0.036). The overall sensitivity (70.1%) and specificity (76.2%) for the radiographic decision aid was comparably low, which was essentially based on false negative cases (22.7%) regarding medial bone-to-bone conditions.

Conclusion

The radiographic decision aid is a helpful tool to predict clinical outcome and implant survival of mobile-bearing unicondylar knee arthroplasty. Strict use of the radiographic decision aid may lead to increased exclusion of appropriate patients with unicondylar knee arthroplasty implantation.