Accuracy and precision of transcutaneous bilirubinometry in postdischarge Asian neonates

Abstract

A prospective cross-sectional study was conducted in a tertiary care center to determine the accuracy of transcutaneous bilirubin measurements (TcB) measured by the Konica Minolta JM-103™ meter compared to total serum bilirubin (TSB) in Asian infants aged 5-14 days. There were 405 late-preterm and term infants involved, and 455 paired samples were obtained by venepuncture and analyzed for bilirubin levels. TcB measurements were performed using the average of three measurements (TcB3) and a single measurement (TcB1) method. The overall correlation between TSB and the TcB was 0.80 (p ≤ 0.001) for TcB3 and 0.76 (p ≤ 0.001) for TcB1, respectively. The mean (SD) difference between TcB3 and TSB was -17.6 (29.5) μmol/L (median, -17.0; interquartile range (IQR), -39.1 to 1.7) and between TcB1 and TSB was -20.7 (32.3) μmol/L (median, -20.4; IQR, -42.5 to 1.7). The mean difference (SD) between the TcB3 and TSB in the low-risk (TSB < 170 μmol/L), intermediate-risk (TSB 170-254 μmol/L), and high-risk (TSB ≥ 255 μmol/L) groups was -2.8 (27.2), -13.4 (27.0), and -33.4 (29.1) μmol/L, respectively. To detect a TSB level of ≥255 μmol/L, using the TcB cutoff level of 204 μmol/L provides a sensitivity of 96 % with a specificity of 58 %. The TcB meter using a specific cutoff level can be reliably used as a screening tool for jaundice detection in older, postdischarge neonates, including the Asian population. Lower cutoff values can be set to capture all infants who merit closer surveillance, potential investigation, and treatment with higher accompanying screening costs.