We describe a 5-yr retrospective analysis of the accuracy of the microscopic examination of bile in the detection of
biliary tract disease in patients with episodic upper
abdominal pain who had negative imaging procedures. In 182 patients, 189 studies of bile were performed using duodenal intubation and
sincalide stimulation for gallbladder contraction. The presence of
cholesterol crystals, leukocytes (greater than or equal to 5/hpf) or the absence of "B" bile constituted a "positive" study. Bilirubinate sludge alone, was defined as "suspicious." Eighty-three patients underwent
cholecystectomy. Among the acalculous patients who underwent
cholecystectomy, 28/28 with bilirubinate sludge had symptomatic improvement as compared with the negative group of which only five of 10 improved (p less than 0.005). The sensitivity of this test for the presence of
gallstones in these imaging-negative patients was 87%, while the specificity was 16%. We conclude that a single microscopic examination of bile cannot accurately predict the pathological findings or the presence of
gallstones in image-negative patients with biliary
pain. The presence of bilirubinate sludge may predict symptomatic improvement in those patients with acalculous
gallbladder disease undergoing
cholecystectomy.