Learning curve and surgical factors influencing the surgical outcomes during the initial experience with laparoscopic pancreaticoduodenectomy.

Abstract

Laparoscopic pancreaticoduodenectomy (LPD) requires sufficient laparoscopic training for optimal outcomes. Our aim is to determine the learning curve and investigate the factors influencing surgical outcomes during the learning curve. We analyzed surgical results of 150 consecutive cases of LPD performed by three hepatopancreatobiliary surgeons during their 50 first cases. Learning curves were constructed by cumulative sum (CUSUM) analysis. Preoperative factors influencing resection time and blood loss were investigated in the introductory and stable periods. RESULTS : The learning curve could be divided into three phases: initial (1-20 cases), plateau (21-30), and stable (31-50). Resection time with lymph node dissection was significantly longer during the introductory period (initial and plateau periods) (P < 0.01) but not the stable phase (P = 0.51). Multivariate analysis revealed that patients with pancreatitis had longer resection times and massive blood loss in both the introductory and stable periods (stable phase). High visceral fat area was also significantly related to massive blood loss in the introductory period (P = 0.04). Hepatopancreatobiliary surgeons need more than 30 cases until LPD becomes stable. Lymph node dissection and patients with high visceral fat area and concomitant pancreatitis should be avoided during the introductory period of the learning curve.