Marked heterogeneity exists in the investigation of prolonged jaundice within the UK. We aimed to audit the safety and clinical implications of a rationalised investigation plan.

From August 2006 to April 2008 infants from two neonatal units were enrolled. Initial investigation protocol included full blood count, Coombs test, thyroid function, direct bilirubin, liver function tests, urine culture, urinary reducing substances and glucose-6-phosphate deydrogenase (G6PD) levels. After interim analysis, the investigation protocol was altered to include rationalised investigations; direct bilirubin, urinary reducing substances and G6PD levels.

105 infants were assessed by initial protocol and 65 by the rationalised protocol. Rationalisation resulted in a significant reduction in repeat investigations and return appointments (table). No infants had serious pathology identified or presented to medical services with undetected pathology during the study.

Rationalised investigation of prolonged jaundice is safe and effective in detecting pathology. €2368–3910 in laboratory costs alone could have been saved in the first half of audit. Future service design should focus on maximising appropriate and swift investigation of infants with prolonged jaundice.