The role of peritoneal lavage in severe acute pancreatitis.

Encouraged by reports of the therapeutic efficacy of peritoneal lavage in small series of five or six patients with acute pancreatitis, we have evaluated this treatment in 24 patients with "severe" pancreatitis. One hundred and three patients with "severe" pancreatitis (28% mortality) were separated from 347 with "mild" pancreatitis (0.9% mortality) by previously described early objective signs. Early treatment (Day 0-7) of "severe" pancreatitis included peritoneal lavage through catheters placed nonoperatively in 18 (Group A) and by catheters placed at laparotomy in six (Group C). Early treatment of nonlavaged patients with "severe" pancreatitis was by standard nonoperative measures in 61 (Group B) and included early operation in 18 (Group D). Lavage was continued for 48-96 hours, usually using 36-48 L/24 hours of balanced isotonic dialysate fluid, and was uncomplicated. Lavage led to striking immediate clinical improvement and no lavaged patient (Groups A and C) died during the first 10 days of treatment of pancreatitis. By contrast, 45% of deaths in nonlavaged patients (Group B and D) occurred during this early period, usually from cardiovascular or respiratory failure. Although lavage reduced mortality in subgroups of patients, ultimate overall survival was no affected (Group A, 83%; B, 84%; C, 33%; D,33%). Late peripancreatic abscesses caused most deaths in lavaged patients. These data show that peritoneal complications of severe acute pancreatitis and dramatically reduces early mortality. Lavage does not prevent the late local sequelae of peripancreatic necrosis.