Retrospective analysis of the results of an external quality assessment of the determination of bilirubin in newborn serum using a unified calibrator.

Overview

Objective: Retrospective analysis of the impact of a unified calibrator on the interlaboratory variability of the results of the total bilirubin determinations in newborn serum.

Design: retrospective cross-sectional
Material and Methods: Retrospective reassessment of the results of participants analyses´ in the external quality assessment (EHK) in the period 2023 to 2025 using the unified calibrator and graphical and statistical evaluation of the results.

Results: Reassessment of the three-year results of the EHK system provided by SEKK s.r.o. and their statistical analysis demonstrated a significant positive impact of implementing a single calibrator on reducing interlaboratory variability and the degree of harmonization of analytical platforms. The data analysis focused on the four dominant platforms of Abbott, Beckmann, Roche and Siemens, on which about 95% of all neonatal bilirubin determinations are performed in clinical labo -⁠ ratories. The original three-year pooled datasets exhibited significant variability in results, with relative bias values ranging up to ±22%. The most pronounced differences in trueness were identified between the Roche (Cobas) and Siemens (ADVIA and Atellica) platforms, with median relative biases of -5.45% (interquartile range, IQR 5.4) and +12.6% (IQR 10.35), respectively. However, the application of a single calibrator led to a significant statistical correction by reducing the variance for all four platforms. The average interquartile range (IQR) decreased on average to 4.1, indicating an increase in precision, and the median relative bias stabilized within a narrow interval around zero, thus stabilizing both the trueness and the overall accuracy of the analyses. Retrospective analysis of biases demonstrated long-term systematic differences between the diagnostics of dominant manufacturers. The main source of undesirable variability in the determination of bilirubin in newborn serum is not the chemical principle of the method itself, but inconsistent calibration of IVD products.

Conclusion: The results of our paper show that the implementation of a unified calibration system is of key importance for the standardization of measurements of total bilirubin in newborn serum. However, there is still a need to define a unified analytical procedure, metrological traceability, redefinition of adequate permissible limits of D_max, special calibrators and control materials suitable for the purpose of measurement.

Keywords:

bilirubin – standardization – hyperbilirubinemia – reference method