A pilot cohort study of a microfluidic-based point-of-care bilirubin measurement system

Serum bilirubin is a key biomarker used to assess liver function. Total bilirubin concentration is correlated with liver physiology and pathology. It is widely used in tracking liver health and treatment effectiveness, such as in neonatal hyperbilirubinemia. Bilirubin concentration can be measured using optical and chemical techniques in clinical settings. However, due to their complexity, cost, and space requirements, existing laboratory techniques are not suitable for point-of-care (PoC) applications. 

Several studies have demonstrated the viability of PoC devices for bilirubin measurement in blood samples. These devices utilise the distinct optical absorption signature of bilirubin, operating by measuring light absorption at 2 wavelengths, typically 465 nm and 570 nm. Another study reported using wavelengths of 450 nm and 580 nm, with an additional 660 nm for the measurement of the sample holder. Our earlier work reported a miniaturised and scalable PoC device for bilirubin measurement in whole blood using the two-wavelength technique at 455 nm and 530 nm. Unlike other PoC devices, ours includes a fully integrated miniatured optical sensing module and microfluidic test cartridge. The device is economical for scalable adoption and highly portable due to its millimetre-sized sensing element. Our previous studies focused on measuring bilirubin concentration using whole blood samples from animal models spiked with exogenous bilirubin. 

This work evaluates the efficacy of our PoC device in measuring bilirubin concentration in human patient samples with varying endogenous bilirubin levels. Test cartridges with embedded microfluidic channels integrated with the optical sensor ensured a consistent optical path whilst using small sample volumes. Although we validated the setup for direct measurement of bilirubin in whole blood samples in our earlier studies, this study focuses on using blood serum to minimise the potential impact of haemoglobin variability and its interference in the measurement. This approach was taken given the small scale of this pilot cohort study. The diagnostic accuracy of the PoC device was compared to the standard clinical laboratory method known as the diazo method. The results were analyzed to determine the statistical association between measurements by the standard laboratory assay and the PoC device. The accuracy of our PoC device in detecting bilirubin levels through receiver operating characteristic (ROC) analyses against the reference values is assessed.