Background: Indicators for cholangiography were originally designed to select patients at risk for common bile duct (CBD) stones for intraoperative cholangiography.
Objective: To refine these criteria to apply to the much more invasive procedure of preoperative endoscopic retrograde cholangiopancreatography (ERCP).
Design: Retrospective review of selection criteria for ERCP in consecutive patients referred over 18 months following the introduction of laparoscopic cholecystectomy.
Setting: Two ERCP units in adjacent teaching hospitals.
Patients: Three hundred seventeen patients with gallstones and in situ gallbladders.
Intervention: Common bile duct imaging at ERCP.
Main outcome measures: Abnormalities justifying ERCP.
Results: Abnormalities justifying ERCP were found in 66% of patients. This group differed significantly from those with normal ducts, with more being referred with abnormal results of all liver function tests (P < .001), jaundice (P < = .001), a dilated CBD on ultrasound (P < .001), or CBD stones on ultrasound (P < .001). On the other hand, patients with normal ducts were significantly more likely to have been referred with pancreatitis (P = .003) or elevated results of individual liver function tests (P < .001). A logistic regression model using age, presence of jaundice at ERCP, levels of alkaline phosphatase and albumin, and ultrasonography showing dilated ducts or visible CBD stones was found to have a specificity of 75% and a sensitivity of 89%. Past pancreatitis or elevated results of individual liver function tests were not predictive factors.
Conclusion: The use of such a model rather than individual criteria would improve the selection of patients for preoperative ERCP, optimizing its role in the laparoscopic era.