657 - Agreement of Ear and Forehead Transcutaneous Bilirubin Measurements with Serum Values in Neonates

Publication Number: 1634.657

Rafael Mendelsohn, Tel Aviv Medical Center, tel aviv, Tel Aviv, Israel; Laurence Mangel, Tel Aviv Medical Center, Tel Aviv, HaMerkaz, Israel; Tal Ber, tel aviv medical center, Tel Aviv-Jaffa, Tel Aviv, Israel; sigal Goldberg Sharon, Tel Aviv Medical Center, Tel aviv, Tel Aviv, Israel; Dror Mandel, Dana Dwek Children's Hospital, Tel Aviv Medical Center, Tel Aviv, Tel Aviv, Israel; Jacky Herzlich, Dana Dweck Children's Hospital, Tel Aviv, Tel Aviv, Israel; Sharon Morag Cooper, Tel Aviv medical center, Tel Aviv, Tel Aviv, Israel; RONELLA MAROM, Tel aviv Medical Center, Tel Aviv, Tel Aviv, Israel

Background: Neonatal hyperbilirubinemia is common and may lead to complications at high bilirubin levels. Transcutaneous bilirubin (TcB) provides a noninvasive screening alternative to serum bilirubin (TSB), though accuracy may vary by device and bilirubin range.

Objective: To assess the agreement between TcB measurements obtained with the BiliCare (ear) and Dräger JM-105 (forehead) devices in term and near-term neonates, and to compare each with paired TSB values.

Design/Methods: In this, ERB approved, prospective study, term and near-term neonates (≥37 weeks) were assessed during routine screening or jaundice evaluation. Bilirubin was measured noninvasively using the BiliCare (ear) and Dräger JM-105 (forehead) devices, with TSB obtained when clinically indicated. Duplicate TcB readings were averaged in 77.5% of low-bilirubin and 46.8% of jaundice work-up cases. Bland–Altman analysis assessed agreement between methods.

Results: In the low-bilirubin subgroup (n = 102, TcB < 11 mg/dL), forehead and ear TcB values were strongly correlated (ρ = 0.70, p < 0.001; 95% CI 0.58–0.79) with moderate variability (R² = 0.56). The mean bias was –0.4 mg/dL (95% LoA –3.7 to +2.9 mg/dL), and only 44.1% of paired readings differed by ≤ ±1 mg/dL.
In the jaundice work-up subgroup (n = 94), TcB correlated strongly with TSB: forehead r = 0.88 (p < 0.001; R² = 0.77) and ear r = 0.77 (p < 0.001; R² = 0.59). Mean biases were +0.29 mg/dL (ear) and –0.4 mg/dL (forehead) with wide LoA (–3.0 to +3.6 mg/dL; –3.22 to +2.42 mg/dL). A proportional bias was observed only for the forehead device (B = 0.23, p < 0.001), indicating overestimation at higher bilirubin levels. Approximately half of TcB–TSB pairs fell within ±1 mg/dL.

Conclusion(s): Both TcB devices showed strong correlations with serum bilirubin and small mean biases but wide limits of agreement and moderate precision. The forehead device tended to overestimate at higher levels, while the ear device underestimated in the low range. TcB remains valuable for screening and monitoring, yet serum confirmation is essential when values approach treatment thresholds.

Figure 1: Correlation between forehead and ear TcB values in the low-bilirubin subgroup
bili1.pdf

Figure 2: Agreement between forehead and ear TcB measurements in the low-bilirubin subgroup
bili2.pdf

Figure 3: Agreement between transcutaneous (TcB) and serum bilirubin (TSB) measurements.
bili3.pdf