Learning curve-associated safety in minimally invasive esophagectomy using a time-based supervisory takeover strategy: a propensity score-weighted analysis of 603 VATS procedures.

BACKGROUND: Minimally invasive esophagectomy (MIE) is technically demanding with a prolonged learning curve, yet training opportunities must be preserved at high-volume centers without compromising patient safety. Whether supervised trainee participation under a structured, protocol-based curriculum with predefined intraoperative checkpoints can maintain perioperative and long-term oncological outcomes remains uncertain. METHODS: This retrospective cohort study was conducted including 603 consecutive patients undergoing video-assisted thoracoscopic esophagectomy (VATS) between 2015 and 2024 at a high-volume cancer center. Patients were categorized into supervisor-only and trainee-involved groups under a structured, protocol-based training system incorporating predefined intraoperative pacing and supervisor takeover criteria. Nine trainees participated over the study period; five completed ≥3 years of structured training (long-term trainees) and four were in-training (< 3 years). A stabilized inverse probability of treatment weighting (IPTW) approach was applied. The primary outcome was Textbook Outcome (TO). Secondary outcomes included postoperative complications and long-term survival. RESULTS: After IPTW adjustment, baseline characteristics were well balanced. Trainee participation was not associated with inferior TO achievement (adjusted OR 1.15, 95% CI 0.82-1.60). No significant differences were observed in anastomotic leakage, pneumonia, recurrent laryngeal nerve palsy, or 90-day mortality. Overall survival and progression-free survival were equivalent between groups. Intraoperative supervisor takeover occurred in 16.1% of trainee cases (n = 51), all prompted by failure to meet a predefined step-level pacing checkpoint. Among long-term trainees, all five (100%) achieved the 300-min reference level, three of five (60%) achieved the 240-min level, and all obtained national endoscopic surgical skill certification. None of the four in-training trainees met either level. CONCLUSIONS: Structured supervision with predefined intraoperative checkpoints in VATS esophagectomy was not associated with inferior perioperative or long-term outcomes. Long-term trainees within a protocol-based framework achieved operative benchmarks and endoscopic surgical skill certification, supporting safety of a protocol-based supervisory takeover strategy during the learning curve of minimally invasive esophagectomy.