Preoperative versus postoperative endoscopic retrograde cholangiopancreatography in mild to moderate gallstone pancreatitis: a prospective randomized trial

OBJECTIVE: To determine whether endoscopic retrograde cholangiopancreatography (ERCP) and common bile duct (CBD) stone extraction should be performed routinely before surgery or'selectively after surgery in patients with mild to moderate gallstone pancreatitis. SUMMARY BACKGROUND DATA: The role and timing of ERCP in mild to moderate gallstone pancreatitis remains controversial. Routine preoperative ERCP identifies persisting CBD stones but carries risks of complications and may delay definitive care. Selective postoperative ERCP, performed only if a CBD stone is seen on intraoperative cholangiography (IOC), avoids unnecessary ERCP but risks unsuccessful stone extraction. METHODS: A prospective, randomized study of consecutive patients with gallstone pancreatitis was conducted. Using previously determined criteria, patients with acute cholangitis or necrotizing pancreatitis were excluded. Patients considered at high risk for persisting CBD stones (CBD size > or =8 mm on admission ultrasound, serum total bilirubin > or = 1.7 mg/dL, or serum amylase > or = 150 U/L on hospital day 4) were randomly assigned to routine preoperative ERCP followed by laparoscopic cholecystectomy, or laparoscopic cholecystectomy with selective postoperative ERCP and endoscopic sphincterotomy only if a CBD stone was present on IOC. Primary end points were costs, length of hospital stay, and the combined treatment failure rates (failure of diagnostic ERCP and IOC, complications of ERCP and endoscopic sphincterotomy, and complications of surgery). RESULTS: One hundred fifty-four consecutive patients with gallstone pancreatitis were evaluated prospectively for study eligibility. Sixty patients met the randomization criteria. Thirty patients were randomized to routine preoperative ERCP and 29 patients to selective postoperative ERCP (1 patient refused). Age, admission laboratory values, and APACHE II and Imrie scores were similar in both groups. By protocol, ERCP was performed in all patients in the preoperative ERCP group. In the postoperative ERCP group, ERCP was necessary in only 7 of 29 patients (24%). Mean hospital stay was significantly longer in the routine preoperative ERCP group (11.7 days) than in the selective postoperative ERCP group (9.0 days). Mean total cost was higher in the preoperative ERCP group ($9,426) than in the postoperative ERCP group ($7,798). The combined treatment failure rate was 10% in both groups. CONCLUSIONS: In patients with mild to moderate gallstone pancreatitis without cholangitis, selective postoperative ERCP and CBD stone extraction is associated with a shorter hospital stay, less cost, no increase in combined treatment failure rate, and significant reduction in ERCP use compared with routine preoperative ERCP.