Residents' Learning Curve of Total Knee Arthroplasty Based on Radiological Outcome Parameters: A Retrospective Comparative Study.

BACKGROUND: This study aimed to plot the impact of a learning curve for a resident's first 103 total knee arthroplasties (TKAs) based on radiological deviations and incision to closure time (ICT), and to compare it to 103 matched TKAs performed by a senior surgeon. METHODS: This is a retrospective comparative study comprising a total of 206 TKAs and evaluated the results based on radiographic outcome and ICT. Radiological evaluation was performed according to a predefined implemented radiological grading system (RGS). t-Tests compared ICT and RGS; data of mechanical axis were observed with Man-Whitey U-tests and Wilcoxon signed-rank-tests. RESULTS: The study included 206 patients (mean age 73 years, mean body mass index of 30). Determining all the deviation points (DP) with the RGS, the deviation ratio for resident vs senior surgeon was 0.96:0.5 DP (P = .0002). The learning curve based on DP showed a decrease over time with statistical significance in the first (26 DP, P = .0001), second (21 DP, P = .0059), and fourth (20 DP, P = .0187) quintiles of implanted cases. The ICT of the resident showed a decrease within the quintiles from 79.45 minutes (first quintile) to 65.17 minutes (fifth quintile), for an improvement of 14.28 minutes. When the quintiles are viewed in relation to the mean operation time of the senior surgeon (mean ICT 66.04 minutes), the mean values of the first and the second quintiles remain statistically significant. CONCLUSION: Supervised TKA showed statistical significance in the learning curves according to deviations documented with a predefined radiological outcome assessment system as well as to ICTs. LEVEL OF EVIDENCE: Level III, retrospective comparative study.