RESUMO
BACKGROUND: Transanal endoscopic microsurgery (TEM) is a technically demanding key technique in minimally invasive rectal surgery. We investigated the learning curve of colorectal surgeons commencing with TEM. METHODS: All TEM procedures of four colorectal surgeons were analyzed. Procedures were ranked chronologically per surgeon. Outcomes included conversion, postoperative complications, procedure time, and recurrence. Backward multivariable regression analysis identified learning curve effects and other predictors. RESULTS: Four surgeons resected 693 rectal lesions [69.9 % adenoma/25.5 % carcinoma; median size 20 cm(2); interquartile range (IQR) 11-35; 7 ± 4 cm ab ano]. A total of 555 resections (80.1 %) were histopathologically radical (R0). Conversion (4.3 %) was influenced by a learning curve [odds ratio (OR) 0.991 per additional procedure; 95 % confidence interval (CI) 0.984-0.998] and by lesion size. Postoperative complications depended only on the individual surgeon and lesion size in benign lesions (10.4 % complications). A learning curve (OR 0.99; 95 % CI 0.988-0.998) and peritoneal entrance affected complications in malignant lesions (13.3 %). Procedure time [median 55 min (IQR 30-90)] was influenced by a learning curve [B -0.11 (95 % CI -0.14 to -0.09)], individual surgeon, single-piece resection, peritoneal entrance, lesion size, and rectal quadrant. Recurrence of benign lesions (4.5 %) depended on lesion size, R0 resection, and prior resection attempts. Recurrence of malignant lesions (8.9 %) depended on 3D stereoscopic view, lesion size, full-thickness resection, and length of follow-up. Recurrence-free survival of patients operated during the 36th through 80th procedure per surgeon was significantly shorter than in patients operated during procedures 1-35 and 81 onwards. CONCLUSIONS: A surgical learning curve affected conversion rate, procedure time, and complication rate. It did not influence recurrence rates, possibly due to evolving patient populations. This first insight into the learning curve of TEM stresses the importance of quality monitoring and centralisation of care.