Multicenter Analysis of the Learning Curve for Endoscopic Ultrasound‐Guided Gallbladder Drainage: A Cumulative Sum Approach

OBJECTIVES: Endoscopic ultrasound-guided gallbladder drainage (EUS-GBD) is effective and safe for poor surgical candidates; however, it remains technically challenging with potential serious adverse events. This study evaluated the learning curve for EUS-GBD for safe and effective procedures. METHODS: From January 2016 to January 2024, 146 patients underwent EUS-GBD performed by three endosonographers. We analyzed the baseline characteristics, procedural outcomes, and learning curves using cumulative sum (CUSUM) analysis. RESULTS: No significant differences were found among the three operators regarding technical success (96.4% vs. 94% vs. 95.1%; p = 0.85), clinical success (96.4% vs. 92% vs. 95.1%; p = 0.61), mean procedure time (8.0 ± 5.7 vs. 9.9 ± 7.9 vs. 9.9 ± 5.1 min; p = 0.24), or procedural adverse events (12.7% vs. 20% vs. 9.8%; p = 0.35). CUSUM analysis revealed that proficiency in procedure time was achieved after performing 27 procedures. Comparing procedure time before and after achieving technical proficiency, technical success (90.6% vs. 98.8%, p = 0.04), clinical success (89.1% vs. 98.8%, p = 0.02), and procedure time (14.45 ± 6.13 min vs. 5.09 ± 2.23 min, p < 0.01) were improved. Adverse event proficiency was reached in 23 procedures, with notable improvements post-proficiency; technical success (87.5% vs. 99%, p = 0.02), clinical success (85.4% vs. 99%, p = 0.01), and procedure time were also improved (15.04 ± 6.43 min vs. 6.33 ± 3.99 min, p < 0.01). CONCLUSIONS: Based on CUSUM analysis, approximately 23 procedures may be required to achieve technical proficiency in EUS-GBD with regard to minimizing adverse events, while 27 procedures are necessary to reach proficiency in terms of procedure time.