Non-invasive simple predictors of biliary atresia in cholestatic infants - A preliminary report.

Background: Biliary atresia (BA) and other cholestatic liver diseases of early infancy share many similar clinical manifestations. We intended to investigate easy and popular non-invasive parameters to assist the differential diagnosis of BA among cholestatic infants.

Methods: We enrolled 145 consecutive cholestatic infants (43.03 ± 2.00 days) who underwent cholestatic workups at our hospital as the study cohort, and another 27 cholestatic infants (47.78 ± 13.33 days) who received their initial cholestatic workups and laboratory tests at four other 4 hospitals as the validation cohort. The clinical data were surveyed retrospectively using a cross-sectional design.

Results: We demonstrated that the "GGT∗D-bil" was significantly higher in the BA group than in the non-BA group in the study cohort (p < 0.001). The Receiver operating characteristic curve analysis of the study cohort identified the "GGT∗D-bil" cutoff >510 mg U/dL.L for the best prediction of BA among cholestatic infants (92 % area under the curve, p < 0.001). The sensitivity and negative predictive value (NPV) for predicting BA among cholestatic infants are 100 % in the study cohort. In the validation cohort, the "GGT∗D-bil" > 510 mg U/dL.L also achieved good sensitivity and NPV (92.31 % and 90.91 %, respectively) for the prediction of BA.

Conclusions: We demonstrated that an easy parameter, "GGT∗D-bil" > 510 mg U/dL.L, may be used as part of the non-invasive cholestatic workups to assist in the diagnosis of BA among cholestatic infants in our study and validation cohorts. The application of this parameter is easy and not limited by the resources of the hospitals.