Correlation of Fellowship Training With Clinical Outcomes following Total Knee Arthroplasty: Complications, Readmissions, and Implant Survival.

INTRODUCTION

Total knee arthroplasty (TKA) is a common procedure for advanced knee osteoarthritis. Although TKA outcomes have improved, variability remains, influenced by factors like patient comorbidities, age, and sex. One potential contributor to outcome variability is the surgeon's fellowship training. This study examines the relationship between orthopaedic fellowship types and clinical outcomes following TKA, focusing on complications, readmissions, and implant survival.

METHODS

This retrospective cohort study used the PearlDiver Database (2010 to 2022) to identify TKA patients. Surgeons were classified based on fellowship training in arthroplasty, trauma, or nonarthroplasty/nontrauma specialties. Exclusion criteria included patients younger than 50, with concurrent trauma, neoplasms, infections, or lacking 90 days of follow-up. Ninety-day adverse events, including any, serious, and minor complications, and readmissions, were assessed. Multivariable analyses controlled for age, sex, and comorbidities. Five-year implant survival was evaluated using Kaplan-Meier analysis and compared with a log-rank test.

RESULTS

A total of 1,180,690 TKA patients were analyzed. Procedures were performed by arthroplasty-trained surgeons (66,654 [5.7%]), trauma-trained surgeons (3,857 [0.3%]), and nonarthroplasty/nontrauma surgeons (1,110,179 [94.0%]). Surgeries by arthroplasty-trained surgeons had markedly lower rates of any (odds ratio [OR]: 0.77 ), serious (OR 0.74), and minor (OR 0.81) adverse events compared with nonarthroplasty/nontrauma surgeons (P < 0.001). Five-year implant survival was lower in the trauma cohort (94.83%) compared with the arthroplasty (96.71%) and nonarthroplasty/nontrauma cohorts (96.82%; P < 0.001).

DISCUSSION

Fellowship training in arthroplasty was associated with lower complication rates, suggesting that surgeon specialization and/or associated care algorithms may be relevant variables. Further investigation is needed to optimize patient outcomes and validate these findings.