Characteristics of learning curve in minimally invasive ileal pouch-anal anastomosis in a single institution.

Abstract

Previous work from our institution has characterized the learning curve for open ileal pouch-anal anastomosis (IPAA). The purpose of the present study was to assess the learning curve of minimally invasive IPAA. Perioperative outcomes of 372 minimally invasive IPAA by 20 surgeons (10 high-volume vs. 10 low-volume surgeons) during 2002-2013, included in a prospectively maintained database, were assessed. Predicted outcome models were constructed using perioperative variables selected by stepwise logistic regression, using Akaike's information criterion. Cumulative sums (CUSUM) of differences between observed and predicted outcomes were graphed over time to identify possible improvement patterns. Institutional pelvic sepsis and other pouch morbidity rates (hemorrhage, anastomotic separation, pouch failure, fistula) significantly decreased (18.2 vs. 7.0%, CUSUM peak after 143 cases, p=0.001; 18.4 vs. 5.3%, CUSUM peak after 239 cases, respectively, p<0.001). Institutional total proctocolectomy mean operative times significantly decreased (307min vs. 253min, CUSUM peak after 84 cases, p<0.001), unlike completion proctectomy (p=0.093) or conversion rates (10 vs. 5.4%, p=0.235). Similar learning curves were identified among high-volume surgeons but not among low-volume surgeons. Learning curves were identified in the two busiest individual surgeons for pelvic sepsis (peaks at 47 and 9 cases, p=0.045 and p=0.002) and in one surgeon for operative times (CUSUM peak after 16 and 13 cases for both total proctocolectomy and completion proctectomy (p<0.001 and p=0.006) but not for other pouch complications (peak at 49 and 41 cases, p=0.199 and p=0.094). Pouch complications, particularly pelvic sepsis, are the most consistent and relevant learning curve end points in laparoscopic IPAA.