The accuracy of transcutaneous bilirubin measurements in preterm infants

Abstract

Context: Jaundice is common in preterm infants and requires frequent blood sampling for total serum bilirubin (TSB) measurements. Transcutaneous bilirubin (TcB) measurement is a noninvasive method to determine TSB, although its accuracy is less conclusive in preterm than in term infants. Aims: The aim was to determine the correlation and agreement between TcB and TSB in jaundiced Saudi preterm infants. Settings and Design: Design - Prospective study. Setting - Neonatal intensive care unit at a tertiary university hospital. Subjects and Methods: Jaundiced preterm infants 24-36 weeks gestational age (GA). JM-103 device and direct spectrometry were used to measure TcB and TSB, respectively. TcB measurements were obtained at the forehead ± 15 min of blood sampling for TSB for every clinically jaundiced preterm infant. Statistical analysis used: The correlations (r) of TcB-TSB values were performed by Pearson linear regression. The bias and 95% limits of agreement were analyzed using Bland-Altman method. Results: A total of 152 paired TcB-TSB readings for 91 jaundiced preterm infants were obtained. Mean birth weight (BW) ± standard deviation was 1690 ± 500 g, mean GA was 32.2 ± 3.1 weeks, mean postnatal age was 62 ± 40 h, (r) were 0.78 and 0.88 and 0.87 for all the preterm infants and ≤ 32 weeks and very low BW, respectively. The bias was 16.6 μmol/L, 95% limit of agreement was (-46.3-79.9 μmol/L). Conclusions: JM-103 device is effective in estimation of TSB. The correlation is inversely proportional to BW < 1500 g and GA ≤ 32 weeks. It tends to mildly overestimates bilirubin values, which make it safe. Its use can reduce blood sampling.