The Analytical Performance of Neonatal Total Bilirubin Assays Does Not Meet the Clinical Need

Severe neonatal hyperbilirubinemia is a critical condition that requires phototherapy (PT) and/or exchange transfusion to prevent irreversible neurological sequelae. Management of neonatal hyperbilirubinemia is largely based on total bilirubin (TBIL) concentration. However, TBIL measurements are known for interassay differences, possibly leading to discrepant treatment decisions depending on the analytical method (1). Past studies using adult serum may not properly reflect the analytical performance of neonatal samples as between-method differences vary between the adult and neonatal matrix (1). The aim of this study was to compare the performance of 10 commonly used bilirubin assays in neonatal samples to the reference measurement procedure (RMP) and evaluate whether bilirubin conjugates or photoisomers contribute to the between-method differences. Methods used include BuBc slide (QuidelOrtho), vanadate oxidase (Siemens), spectrophotometry (Radiometer), and 7 diazo (Abbott, 2x Beckman Coulter, 2x Siemens, 2x Roche). Residual plasma from neonates with no significant interfering hemolysis was collected before, during, or after PT. Samples from neonates with different gestational age, time after birth, and risk factors were included and were pooled according to the PT status and similar TBIL concentration range. Additionally, leftover exchange transfusion serum was collected. Altogether, a total of 11 samples in the concentration range 8.5 to 20.3 mg/dL (146–347 µmol/L) were measured in duplicate using 10 routine TBIL assays and 2 RMPs, i.e., the Doumas and Klauke methods (2, 3).