Comprehensive evaluation of the learning curve for submucosal tunneling endoscopic resection

To analyze the learning curve for submucosal tunneling endoscopic resection (STER).

A retrospective analysis of clinical data from patients with submucosal tumors (SMTs) treated with STER at the First Affiliated Hospital of Ningbo University from January 2018 to October 2024. The surgeries were performed by three endoscopists in our hospital with no prior STER experience. Cumulative sum (CUSUM) curves were used to analyze lesion diameter, lesion location, operation time, tunnel length, procedure speed, en bloc resection rate, complete resection rate, and intraoperative and postoperative complications in different phases for each endoscopists.

There was no statistically significant difference (P > 0.05) in age and lesion location among the three endoscopists' cases. Based on the inflection point of the CUSUM curve (16-19 cases), the process can be divided into two phases (phase 1 and phase 2). Each endoscopist showed a significant increase in procedure speed in phase 2 compared to phase 1 [0.21 (IQR: 0.12-0.35) cm³/min vs. 0.08 (IQR: 0.06-0.13) cm³/min, P₁ < 0.001; 0.18 (IQR: 0.13-0.21) cm³/min vs. 0.07 (IQR: 0.05-0.10) cm³/min, P₂ < 0.001; 0.08 (IQR: 0.07-0.16) cm³/min vs. 0.04 (IQR: 0.03-0.05) cm³/min, P₃ < 0.001]. Meanwhile, endoscopist 1 and endoscopist 2 showed a significant increase in the diameter of resected lesions in phase 2 compared to phase 1 [2.35 (IQR: 1.63-3.50) cm vs. 1.50 (IQR: 1.20-2.00) cm, P₁ = 0.02; 2.00 (IQR: 2.00-2.90) cm vs. 1.40 (IQR: 1.05-2.00) cm, P₂ = 0.004]. The en bloc resection rate and complete resection rate for three endoscopists reached the ideal level (100%) in both phases. The incidence of complications in the two phases showed no statistically significant difference (P > 0.05).

For endoscopists inexperienced in STER, approximately 20 cases are required to achieve fundamental proficiency in its operative methods and techniques.