Does technology increase the risk of stiffness after Total Knee Arthroplasty ? Evaluating the role of the learning curve
摘要
Arthrofibrosis and stiffness are challenging complications after total knee arthroplasty (TKA), often treated with Manipulation Under Anesthesia (MUA). Robotic-assisted TKA (RA-TKA) aims to enhance surgical precision, though its impact on stiffness remains unclear. This study examined the association between surgical technologies and MUA rates using propensity score matching (PSM) and assessed whether surgeon experience affects MUA risk. We conducted a retrospective case-control study of 25,611 primary unilateral TKAs (2016–2024), stratified by manual, computer-assisted (CA), or RA techniques. MUA cases within 90 days were matched 1:2 to controls using PSM. Conditional logistic regression assessed the association between surgical technologies and MUA risk. A learning curve analysis of each surgeon’s first 100 RA-TKAs assessed surgical proficiency progression. Odds ratios were compared before and after the learning phase to evaluate its impact on MUA rates. Both CA-TKA (OR = 1.11, P = 0.33) and RA-TKA (OR = 1.24, P = 0.12) revealed a higher trend in MUA risk compared to manual. Learning curve analysis of 10 surgeons demonstrated a distinct learning breakpoint at case 11.8 (95% CI 9.2–14.4), after which MUA rates declined and stabilized. The pre-learning phase showed a 5.2% MUA rate, compared with 2.9% post-learning (P = 0.042). Adjusted analyses showed that the learning phase was independently associated with a 2.13-fold increase in MUA risk (95% CI 1.02–4.45, P = 0.043), suggesting that early robotic adoption may contribute to postoperative stiffness risk. Finally, after removing these early cases from the full cohort, RA-TKA showed no increased risk (OR 0.98; 95% CI 0.74–1.29), indicating no increased risk once surgeons were beyond the learning phase. Early robotic adoption may contribute to postoperative stiffness risk. Once surgeons progressed beyond the adoption phase, RA-TKA showed similar risk to other techniques, emphasizing the importance of structured training during robotic adoption.