Abstract
IMPORTANCE: Visual inspection is commonly the first-line screening method for hyperbilirubinemia in neonates cared for at home. However, it is unreliable and may delay treatment. OBJECTIVE: To assess the clinical outcomes, costs, and user convenience associated with universal transcutaneous bilirubin (TCB) screening for detecting hyperbilirubinemia in a diverse neonatal population cared for at home. DESIGN, SETTING, AND PARTICIPANTS: This prospective decision analytical model was conducted between July 11, 2021, and June 9, 2023, in 9 midwifery practices across the Netherlands. Neonates (gestational age ≥35 weeks) were eligible if they had their first midwife visit prior to postnatal day 6 and had not yet received phototherapy. EXPOSURE: At each home visit, the midwife first undertook visual inspection and then performed TCB measurements on the neonate's sternum. Bilirubin was measured in blood when indicated by visual inspection and/or elevated TCB reading. MAIN OUTCOMES AND MEASURES: The hypothesis was that TCB could replace visual inspection if (1) more neonates requiring treatment for hyperbilirubinemia were detected and (2) fewer heel pricks were needed. The 2 main outcomes were (1) blood bilirubin level greater than the national guideline's treatment threshold and (2) requiring a heel prick for bilirubin quantification. McNemar tests were used for both hypotheses. Cost-effectiveness was assessed using a decision-tree analytic model. RESULTS: Data from 2314 neonates (median [IQR] gestational age, 39 [39-40] weeks; 1172 [50.6%] male) were analyzed. Overall, 78 (3.4%) had a bilirubin level greater than the treatment threshold. Of these, 28 had been identified through TCB screening but missed by visual inspection (absolute risk difference, 28%; 95% CI, 13%-42%; P < .001). Although TCB missed 7 neonates identified through visual inspection, these were all attributable to either misinterpretation of TCB readings by the midwife or structural overestimation of blood bilirubin levels in one hospital laboratory. TCB screening led to 102 additional heel pricks vs visual inspection (244 vs 142; P < .001). Sensitivity analysis showed that when TCB screening was selectively applied only to neonates with any jaundice, additional heel pricks were reduced to 82 (P < .001), with no loss in diagnostic performance. Universal TCB screening saved €15 ($17) per neonate. CONCLUSIONS AND RELEVANCE: In this decision analytical model study of 2314 (near) full-term neonates, universal TCB screening among neonates cared for at home identified more neonates requiring treatment but required more heel pricks than visual inspection alone. Universal TCB screening was cost-effective compared with visual inspection. The additional required heel pricks decreased by selective TCB screening.