Abstract
The diagnostic value of sonography was assessed in 176 patients with jaundice who had a final diagnosis on the basis of surgery, autopsy, liver biopsy, endoscopic retrograde cholangiopancreatography, and serum tests positive for hepatitis A and B. Obstructive (n = 113) versus nonobstructive jaundice (n = 63) was diagnosed with a sensitivity of 91% and a specificity of 95%. Nine of the 10 false negatives had choledocholithiasis. Application of stricter criteria for common duct (CD) dilatation than the one used of CD ± 8 mm to ± 12 mm would have lowered the sensitivity from 91% to 84%. On the basis of these criteria the predicted level of obstruction would have shifted from distal to proximal in 36% of the patients. Choledocholithiasis was more often associated with decreasing bilirubin values than malignancy: 55% versus 9% (p < 0.05). The present study indicates that intermittent obstruction. usually associated with choledocholithiasis, constitutes the main problem in detection of obstructive jaundice. Diameter criteria used for diagnosing dilatation of the CD may greatly influence sonographic accuracy.