RESUMO
BACKGROUND: Biliary atresia (BA) is the main cause of severe liver damage in infants. Successful surgical treatment is related directly to an early and rapid diagnosis. The aim of this study was to determine specificity, sensitivity, and predictive value of the duodenal tube test (DTT) in the diagnosis of BA in a series of infants with cholestatic jaundice. METHODS: This was a descriptive study of a series of infants with cholestatic jaundice created to validate the sensitivity, specificity, and predictive value of the DTT in the diagnosis of BA. A total of 254 patients were identified from 1988 to 1998. The study cohort included 137 male infants (53.9%), and the mean age on admission was 8.3 weeks +/- 2.47 weeks (standard deviation). Study protocol included liver function tests, liver ultrasound, metabolic screening and serology for viral hepatitis, and toxoplasma, rubella, cytomegalovirus, herpes, and others. A nasoduodenal tube was, placed at the distal duodenum and the fluid was collected for 24 hours. DTT was considered bile positive when yellow biliary fluid was observed; the test was concluded at this time. When no yellow biliary duodenal fluid was observed, the collection was continued for 24 hours and, if negative, was reported as bile negative. The patients with a bile-positive DTT were not explored surgically, and the cholestasis workup was completed. Laparotomy and a surgical cholangiogram were indicated in patients with bile-negative DTT. If BA was verified, portoenterostomy was performed. The gold standard for BA diagnosis was the following: obstruction of the biliary tract confirmed by laparotomy and a surgical cholangiogram, and clinical outcome in patients without laparotomy (followed for a minimum of 18 months). RESULTS: The results are as follows. BA: bile-positive DTT, n = 3; bile-negative DTT, n = 108. No BA: bile- positive DTT, n = 134; bile-negative DTT, n = 9. The following values were also determined: sensitivity, 97.3%; specificity, 93.7%; positive predictive value, 92.3%; and negative predictive value, 98.5%. The final diagnoses were as follows: BA, n = 111 (43.7%); neonatal hepatitis syndrome, n = 103 (40.6%); cholestasis associated with inspissated bile syndrome, n = 13 (5.1%); choledochal cyst, n = 11 (4.3%); galactosemia, n = 9 (3.5%); cirrhosis of unknown etiology, n = 5 (2%), and Alagille syndrome, n = 2 (0.8%). CONCLUSIONS: The data obtained from this series validate the DTT as a useful clinical tool for the differential diagnosis of the infant with cholestasis, particularly for indicating laparotomy and cholangiogram to substantiate BA. This diagnostic test is quick and simple, and offers the clinician valuable information with which to determine whether surgical intervention is necessary.