Triaging infantile cholestasis: Validation of non-invasive diagnostic modalities and a composite scoring system for early detection of biliary atresia
Yang Yang Lee1, Siddhant Chavan2, Alba Bueno2
1Liver Unit, Birmingham Women's and Children's Hospitals NHS Foundation Trust, Steelhouse Lane, Birmingham B4 6NH, UK; Department of Paediatric Surgery, National University Hospital, 1E Kent Ridge Road, Singapore 119228.
Background:
Biliary atresia (BA) is the most common cause of infantile obstructive jaundice, with outcomes critically dependent on timely Kasai portoenterostomy. Differentiating BA from other causes of infantile cholestasis remains challenging. We aimed to evaluate the diagnostic performance of individual investigations and the composite predictive index (the Simple Biliary Atresia Score, SBAS) within a UK national tertiary paediatric liver unit.
Methods:
This retrospective observational cohort study included all infants referred with conjugated hyperbilirubinaemia at less than 100 days of life to the Liver Unit at Birmingham Children's Hospital between March 2022 and March 2025. The primary endpoint was the diagnosis or exclusion of BA, confirmed by intra-operative cholangiography and histopathology. SBAS was calculated from gallbladder length, common bile duct diameter, periportal echogenicity, direct-to-total bilirubin ratio, and GGT, yielding a 0-7 composite score; a cutoff ≥3 defined high risk. Diagnostic performance was assessed using sensitivity, specificity, predictive values, and AUROC.
Results:
Of 143 included infants, 46 (32.2%) were diagnosed with biliary atresia. Direct bilirubin (median 102 vs 85 μmol/L, p = 0.015) and GGT (333 vs 135 U/L, p < 0.001) were higher in biliary atresia, and all sonographic markers differed significantly between groups. At SBAS ≥3 the score achieved sensitivity 97.7% (95% CI 87.9-99.6), specificity 68.5% (58.3-77.2), positive predictive value 60.0% (48.3-70.7) and negative predictive value 98.4% (91.4-99.7); the AUROC of the score as a continuous variable was 0.874 (0.817-0.930), exceeding that of every individual investigation. The single false negative was a cystic biliary atresia misinterpreted on ultrasound. Alagille syndrome was over-represented among false positives (17.9% vs 1.6% of true negatives). Liver biopsy was performed in 55 of 142 infants (38.7%), in 50 cases concurrently with operative cholangiography.
Conclusion:
SBAS is a robust, low-cost, bedside triage tool that retains excellent sensitivity and discriminative performance in a UK tertiary referral cohort, supporting its use as secondary triage to identify infants requiring urgent escalation to definitive management.