2795 The Effect of an Invasive Approach for Infected Necrotizing Pancreatitis on Mortality: A Retrospective Large Multi-Center Cohort Study

Abstract

INTRODUCTION: Infection of acute pancreatic necrosis is associated with a high mortality rate. The relative merits of invasive therapies for preventing death among patients with this condition are unclear. METHODS: This retrospective multicenter cohort study was conducted from 2009 to 2013 at 44 treatment centers in Japan. Patients who underwent invasive treatment for suspected infection of pancreatic necrosis were enrolled into three treatments groups: the primary open necrosectomy before drainage, secondary open necrosectomy after drainage, and minimally invasive treatment (laparoscopic, percutaneous, and endoscopic) groups. The association of each treatment with mortality was assessed using multivariate analysis after adjusting for age, severity according to the revised Atlanta criteria, and the Charlson index. RESULTS: Of 1159 patients with acute pancreatitis, 118 patients underwent invasive treatment for infected pancreatic necrosis. Of these, nine, 20, and 89 patients underwent a primary open necrosectomy, secondary open necrosectomy, and minimally invasive treatment (three laparoscopies, 49 percutaneous treatments, 37 endoscopies), respectively. The mortality rate was 33.9% (40/118). Multivariate analysis showed no significant association between the type of intervention or Charlson index and mortality (P = 0.35 and P = 0.12, respectively). The mortality odds ratio (OR) for the secondary open necrosectomy versus minimally invasive treatment was 1.25 (95% confidence interval [CI]: 0.40–3.84, P = 0.69) while that for the primary open necrosectomy versus minimally invasive treatment was 3.13 (95% CI 0.66–17.5, P = 0.15). However, age (OR 1.04, 95% CI [1.01–1.08], P = 0.0067) and severity (OR 7.57, 95%CI [2.27–25.2], P = 0.0002) were significantly associated with mortality. CONCLUSION: In patients with infection of acute pancreatic necrosis, age and severity based on the revised Atlanta criteria, but not intervention type, were significantly associated with mortality. For patients undergoing an open necrosectomy, performing drainage after rather than before the procedure may produce better outcomes.