Validity of BiliDx as a point-of-care bilirubin measurement device to diagnose and monitor neonatal jaundice at Muhimbili National Hospital, an observational study

Background Neonatal jaundice is a condition caused by elevated levels of bilirubin in the bloodstream. Laboratory determination of serum bilirubin concentration by total serum bilirubin (TSB) test is still considered as gold standard for clinical guidance and practice. In developed countries, diagnosis of neonatal jaundice is shifting towards point-of-care medical devices. BiliDx is a device developed to allow a fast, blood-based determination of bilirubin levels at the point of care. This study aimed to determine the accuracy of the BiliDx device relative to a standard laboratory total serum bilirubin to diagnose and monitor jaundice among neonates admitted at Muhimbili National Hospital (MNH). Material and methodology This was a prospective hospital-based observational study conducted at the Neonatal Ward – MNH, Dar-es-Salaam, Tanzania from November 2022 to January 2023. A total of 180 neonates admitted at the neonatal ward with jaundice and whose parents consented were enrolled in the study. Blood samples were collected; 2 ml of venous blood into the vacutainer bottle for standard laboratory measurement of total serum bilirubin (TSB) and 25µL blood collected into a transfer pipette tube and applied to BiliDx. STATA version 15.1 was used for data analysis. Results Out of 180 neonates, 39.4% (71/180) had birth weight between 1500 − 2499.9 g, approximately 2/3rd (120/180) were preterm, 92/180 (51.1%) were males and 100/180 (55.6%) were undergoing phototherapy treatment the moment sample taken. The mean bilirubin concentration was 92 mmol/l for BiliDx and 118 mmol/l for standard laboratory TSB. The minimum and maximum values obtained with BiliDx were, 3.4 and 427.5 mmol/l respectively, compared with 10.7 and 382.1 mmol/l using standard laboratory TSB. A linear relationship and correlation coefficient of 0.8408 ( p = 0.000) between BiliDx and standard laboratory TSB was found. The regression analysis showed the presence of constant error [coefficient of BiliDx/slope = 0.91, 95% CI (0.82–0.99), p = 0.000] and random error exclusively [coefficient of constant/y-intercept = 48.52, 95%CI (37.70-59.34), p = 0.000]. The Bland–Altman plot showed an acceptable mean difference of 39.1mmol/l, limits of agreement of -48.3mmol/l to 126.4mmol/l, and 179 points (179/180 = 99.4%) lying inside the limits of agreement. Conclusion The results support the use of BiliDx for rapid and accurate testing of elevated levels of bilirubin in the bloodstream among neonates since 99.4% of the differences between BiliDx and standard laboratory TSB lie between the lines of agreement.