Diagnostic accuracy of transcutaneous bilirubinometry and icterometry compared with capillary bilirubin measurement in neonates with jaundice in Bhutan: a cross-sectional study

During the early neonatal period, particularly the first week of life, hyperbilirubinemia remains an important cause for hospital admission and is associated with considerable morbidity. Although total serum bilirubin (TSB) is the gold standard for diagnosing hyperbilirubinemia, obtaining serum bilirubin requires blood sampling and access to laboratory facilities, which may be limited in resource-constrained healthcare settings. Hence, non-invasive methods such as transcutaneous bilirubinometry and icterometry may provide a useful alternative for screening hyperbilirubinemia. This study aims to evaluate the diagnostic accuracy of transcutaneous bilirubinometer (TcB) and icterometer in relation to capillary bilirubin measurement using the Toitu BL-200 bilirubinometer among jaundiced neonates in Bhutan. A hospital-based cross-sectional diagnostic accuracy study was conducted among neonates with hyperbilirubinemia attending the Community Health Unit (CHU) of Jigme Dorji Wangchuck National Referral Hospital (JDWNRH), Thimphu, Bhutan, between December 2021 and November 2022. Neonates presenting for postnatal care were initially screened using TcB. Those with TcB values above the age-specific phototherapy threshold based on the American Academy of Pediatrics (AAP) guideline were recruited by convenience sampling and underwent icterometer assessment and capillary bilirubin measurement using the Toitu BL-200 bilirubinometer as the reference method. The diagnostic accuracy of TcB and icterometer was evaluated using correlation analysis and Bland-Altman analysis to determine the agreement between the TcB and Capillary bilirubin measurements. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), positive likelihood ratio (+LR) and negative likelihood ratio (-LR) were calculated. Receiver operating characteristic (ROC) curve analysis at different bilirubin thresholds and the area under the ROC (AUC) was calculated to assess the discriminatory ability of the non-invasive methods for detecting hyperbilirubinemia. Among 723 infants, 683 (94.5%) neonates were term babies with a mean gestational age of 39.0 ± 1.4 weeks. The mean capillary bilirubin and TcB levels were 13.8 ± 2.1 mg/dL and 20.1 ± 2.4 mg/dL, respectively, while the median Icterometer score was 3.0 (IQR 2.5–3.5). In comparison to Icterometer, TcB demonstrated a stronger correlation with capillary bilirubin (r = 0.601vs 0.723) with both p < 0.001. TcB overestimated capillary bilirubin by a mean of 6.30 mg/dL (95% CI, 6.18–6.42), with 95% limits of agreement between 3.00 and 9.60 mg/dL. The area under the receiver operating characteristic Curve (AUCs) for TcB at capillary bilirubin thresholds of ≥ 11, ≥13, ≥ 15, and ≥ 17 mg/dL were 0.847, 0.846, 0.891, and 0.912, respectively. In comparison to AUCs for icterometer, it was 0.747, 0.765, 0.824, and 0.885. At the ≥ 17 mg/dL threshold, an Icterometer cutoff of ≥ 3.5 yielded sensitivity of 91.2% and specificity of 67.4%, while a Youden-derived TcB cutoff of ≥ 21.9 mg/dL yielded 87.7% sensitivity and 84.1% specificity. The difference in AUC between TcB and Icterometer at this threshold was not statistically significant (ΔAUC = 0.027; 95% CI, − 0.027 to 0.081; P = 0.331). The icterometer shows potential as a low-cost screening tool for neonatal hyperbilirubinemia, particularly where TcB may not be readily available. However, further validation against TSB using age- and gestational-specific thresholds is needed before routine clinical implementation.