Analysis of clinical parameters that contribute to the misdiagnosis of biliary atresia.

BACKGROUND

Biliary atresia (BA) is one of the most common and perplexing causes of neonatal cholestasis. Each year many cases of neonatal cholestasis are misdiagnosed as BA, resulting in unnecessary exploratory surgery. Therefore, the aim of our study was to analyze the clinical features and parameters that contribute to the misdiagnosis of BA. We used a retrospective study to compare BA and similar neonatal cholestatic diseases, which were confirmed by intraoperative cholangiography.

METHODS

Six hundred and two infants that were suspected to have BA were recruited for the study. All cases were divided into a non-BA group and a BA group according to intraoperative cholangiography. In addition, each group was divided into three subgroups according to the patients age at surgery (group i, < 60 d; group ii, 60-90 d; and group iii > 90 d). The annual misdiagnosis rate of non-BA patients was calculated. Age at onset of jaundice and the liver function and ultrasound results were compared between the two groups and subgroups. Moreover, the positive predictive value and false positive rate of hepatobiliary scintigraphy in the diagnosis of BA were calculated. Finally, the disease spectrum of the non-BA group was analyzed.

RESULTS

Of the 602 cases, 83 patients were diagnosed as non-BA. The remaining 519 cases were confirmed to have BA. There was no significant decline in the misdiagnosis rate of suspected BA cases by year. The age at onset of jaundice, total bilirubin (TBIL), direct bilirubin (DBIL), DBIL/TBIL and alanine aminotransferase (ALT) values before the exploratory operation showed no statistically significant difference (P > 0.05) in the non-BA versus BA groups. However, the mean level of gamma-glutamyl transpeptidase (γ-GT) was 263.2 mmol/l in the non-BA group and 902.7 mmol/l in the BA group (P < 0.01). The length of the liver below the ribs was detected with ultrasound and found to be smaller in the non-BA group than that of the BA group (2.99 cm ± 1.62 vs. 3.61 cm ± 1.26, respectively; P < 0.05). Among the 498 infants who received hepatobiliary scintigraphy examination, the false positive rate was 13.3% (66/498) and the positive predictive value was 86.7% (432/498). In the non-BA group, 58 infants suffered from hepatitis syndrome, 16 cases were biliary dysplasia, 5 cases were TPN related cholestasis, 2 cases were bile duct perforation and 2 were bile-plug syndrome.

CONCLUSION

The similarity of liver function tests and excessive dependence on hepatobiliary scintigraphy may contribute to the misdiagnosis of infants with jaundice. The age at onset of jaundice, the level of γ-GT and the liver length below the ribs may be helpful in the differential diagnosis of jaundice in infants.