Comparability of Total Bilirubin methods for neonatal samples across 6 NHS trusts within North West England

The National Institute for Health and Care Excellence (NICE) guideline: management of neonatal jaundice [CG98] outlines total bilirubin (TBil) thresholds for the treatment of neonatal jaundice. Despite wide analytical variation in TBil, evidenced on external quality assessment schemes, the thresholds are based on a single Roche diazo assay. Adoption of these thresholds may result in inappropriate treatment decisions depending on the analytical methodology. This study assesses TBil method bias using pooled neonatal (<29 days) samples, prompted by incidents within Greater Manchester where differences in neonatal TBil measurements obtained due to transfer of care across sites caused clinical confusion.

Seventeen serum and 9 plasma neonatal pools were analysed in duplicate on 6 platforms: Roche Cobas 701, Roche Pro, Abbott Alinity, Beckman Coulter AU5800, Siemens Atellica and Siemens XPT , utilising 5 diazo and 2 vanadate oxidation methodologies. Measurement biases compared to Roche Cobas were assessed.

Siemens Atellica and XPT vanadate oxidation methods exhibited significant positive bias: 17.4% [95% CI: 15.0-19.8%] and 17.0% [95% CI: 15.0-19.0%], respectively. All diazo methods were within desirable bias limits (+/-8%) as defined by the European Federation of Clinical Chemistry and Laboratory Medicine analytical performance specifications. Bias for Siemens vanadate oxidation and Beckman Coulter diazo methods exhibited matrix variation.

Use of Siemens TBil vanadate oxidation method may lead to over treatment of neonatal jaundice when using NICE [CG98] thresholds. Laboratory professionals should evaluate the appropriateness of their TBil methodology in collaboration with their neonatal teams to minimise risks of overtreatment.