DIAGNOSTIC ACCURACY OF TRANSCUTANEOUS BILIRUBINMEASUREMENTS IN NEONATES ADMITTED AT A TERTIARY TEACHINGAND REFERRAL HOSPITAL IN NAIROBI, KENYA

Abstract

Background: Neonatal jaundice is a serious medical issue affecting newborns and is associated with high mortality, hence the need for fast, simple, and accurate diagnosis.

Objectives: The study investigated the diagnostic accuracy of transcutaneous bilirubin measurements (TcB) using total serum bilirubin (TSB) as the gold standard among admitted neonates.

Methods: This was a prospective diagnostic accuracy(cross-sectional) study involving 158 term neonates with clinically suspected jaundice. Diagnostic performance was assessed using two complementary approaches: (i) application of the 2022 American Academy of Paediatrics (AAP) guideline TSB thresholds, along with Bhutani Nomogram TcB interpretation, and (ii) Receiver Operating Characteristic (ROC) curve analysis to evaluate discriminative performance across different TSB cut-off values (222, 256, and 291 µmol/L).

Results: The mean difference between TcB and TSB was –0.67 µmol/L (95% CI:–6.67 to 5.34), indicating close agreement between the two methods. Across TSB thresholds of at least 222, 256, and 291 µmol/L, a TcB cut-off of 222 µmol/L correctly identified 85.9%, 90.7%, and 100.0% of infants with elevated bilirubin, respectively. TcB showed high sensitivity in the first 24 hours of life (90.7%) and maintained good performance at 48 hours (80.0%). Overall, TcB correctly detected most infants with high serum bilirubin (88.9% sensitivity), correctly ruled out most without high bilirubin (88.5% specificity), and achieved an overall accuracy of 88.6%.

Conclusion: The study revealed a strong TcB-TSB correlation, supporting TcB as a fast, non-invasive screening tool for neonatal hyperbilirubinemia, especially within 24-48 hours of life. A TcB cut-off of 222 µmol/L offers optimal performance.