What is the Learning Curve for Total Ankle Arthroplasty?

Abstract

Category: Ankle Arthritis; Ankle Introduction/Purpose: There is limited data concerning the “learning curve” in Total Ankle Replacement (TAR). This study aimed to examine surgeons’ revision rates over time to determine the number of cases after which revision rates plateaued. Methods: Primary TAR procedures for any diagnosis performed by surgeons whose first procedure was recorded by the AOANJRR since January 2008 were analysed. The time to first revision for procedures performed by surgeons with < 15, 15-29, 30-49, and ≥50 prior TAR procedures was summarised using Kaplan-Meier estimates of implant survivorship. Hazard ratios (HR) were obtained from Cox proportional hazards models adjusting for age, gender and the operating surgeon’s number of prior TAR cases, with the latter modelled using restricted cubic splines. Further adjustment for insert mobility was also performed. Results: There were 2,637 TAR procedures performed by 128 eligible surgeons. Procedures performed by surgeons with < 15 prior cases had a cumulative percent revision (CPR) of 15% at 8 years, compared to 11.8% for procedures where the surgeon had previously performed ≥50 TARs. Procedures where the surgeon had 15 prior cases had approximately 1.5 times the hazard of revision of procedures where the surgeon had ≥50 prior cases (HR = 1.57, p = 0.042). The revision rate of procedures in which the surgeon had performed 25 prior cases did not differ significantly compared to those with ≥50 prior cases (HR = 1.24, p = 0.143). Similar results were obtained when adjusting for insert mobility, although estimated revision rates did not differ significantly with number of procedures. Conclusion: The first 15 TAR undertaken by a surgeon have a significantly higher rate of revision compared to those after 50 TAR. This difference reduces after 25 TAR.