P0718 - Correlation between Total Serum Bilirubin and Transcutaneous bilirubin levels in extremely preterm infants less than 30 weeks gestation

TITLE: Correlation between Total Serum Bilirubin and Transcutaneous bilirubin levels in extremely preterm infants less than 30 weeks gestation

Background: Measurement of Transcutaneous bilirubin (TCB) level is a safe, noninvasive, cost effective and widely used method for screening term and late preterm infants with hyperbilirubinemia. Limited data exists regarding the validity of TCB use in extremely preterm (EP) infants.

Objective: The primary objective was to evaluate the correlation and agreement between total serum bilirubin (TSB) and TCB levels in EP infants.
Design/

Methods: This was a prospective multicenter study conducted at tertiary and regional NICUs in California. EP infants between 23.0-29.6 weeks of gestation admitted to the NICU without congenital anomalies were included. TCB measurements were performed within 30-60 minutes of TSB measurements during the first 3 weeks of life. Trained NICU staff measured TCB levels by using Drager Jaundice meter (JM- 103) or the BiliChek transcutaneous bilirubinometer. Neonates with TSB readings in the phototherapy range received phototherapy as per the standard protocol. We examined the difference between TCB and TSB values, gestational age, maternal race, ethnicity and neonatal morbidities. Descriptive statistics were generated for entire group using IBM SPSS statistical software version 26. Correlation between TSB and TCB levels was assessed with Pearson’s correlation analyses. Bland–Altman analysis was used to show the differences against the mean of the two methods.

Results: 299 paired TCB and TSB measurements (58 infants) were analyzed. Median gestational age was 27.2 weeks (IQR 25.5-28.3) and postnatal age ranged from 1 to 18 days of life. TSB values ranged from 0.41 to 14.7 where as TCB values ranged from 0 to 15.5. Mean difference between TCB and TSB was 0.37 (SD +/-2.56). Pearson bivariate correlation testing revealed moderate level of correlation between TCB and TSB with a coefficient of 0.625 (p<0.001). Proportionate bias was noted in infants with higher bilirubin values. Bland Altman analysis of log transformed data (Figure 1) showed a good agreement at 95 percent limits with mean of 0.0074 (SD 0.267). Regression equation predicted TSB= 2.68+0.42(TCB) with r squared of 0.39 suggesting weak to moderate correlation (Figure 2).
Conclusion(s): TCB levels correlated moderately with TSB levels in this ethnically diverse population of EP babies in California NICUs. TCB may be a useful screening tool for monitoring jaundice in EP newborns. Larger studies are needed to validate these findings.