ABSTRACT
BACKGROUND: Clinically significant variation has been reported within and between plasma and whole blood total bilirubin methods used to identify neonates for whom clinical intervention for hyperbilirubinemia may be required. OBJECTIVE: To evaluate total bilirubin measurements between the Radiometer whole blood co-oximeter and plasma bilirubin methods from Roche Diagnostics and Ortho Clinical Diagnostics using neonatal specimens. METHODS: Total bilirubin levels were analyzed by whole blood co-oximetry (Radiometer® ABL90). Specimens were centrifuged and plasma analyzed for total bilirubin with a diazo method (Roche Cobas® C-601) and a reflectance spectrophotometric BuBc dry film method (Ortho Clinical Diagnostics VITROS® 350). Results were evaluated by regression, Bland-Altman comparisons and t-tests. RESULTS: The patient correlation study yielded the following regression equations in µmol/L: a) Radiometer=1.03 Roche - 3.5µmol/L b) Radiometer=0.98 Ortho - 5.7µmol/L c) Roche=0.97 Ortho - 2.4µmol/L. The mean bias over the range of total bilirubin levels examined was -1.0µmol/L for the Radiometer versus the Roche (p≤0.305); -4.4µmol/L for the Radiometer versus Ortho (p≤0.005) and -4.4µmol/L for the Roche versus Ortho (p≤0.002). CONCLUSIONS: Whole blood total bilirubin measurement using the Radiometer ABL90 blood gas analyzer provides accurate and precise results compared to the Roche plasma diazo method. Compared to the reflectance spectrophotometric method, results are precise and had a small but statistically significant bias of -4.4µmol/L.