[Evaluation of transcutaneous bilirubin measurement and agreement with bilirubinemia]

Evaluation of the neonates for jaundice and kernicterus is indispensable when early hospital discharge has become standard practice. Transcutaneous bilirubin (TcB) measurement is an advantageous option because of its non-invasive nature and the possibility of its use near the patient. The point of care device BiliCheck has been compared in numerous instances to serum bilirubin. However, its clinical utility remains a subject of discussion. We have compared total blood bilirubin (TBB) concentrations to TcB values using the BiliCheck in newborns at 48 +/- 12 hours of life, at the time of discharge when they have lost weight. One hundred and ninety-six term neonates were initially included into the study. Transcutaneous bilirubin could be compared to whole blood bilirubin for 178 of them. Methods were compared by linear regression analysis and by the non-parametric Bland and Altman method. The correlation between BiliCheck and whole blood bilirubin was adequate (r(2): 0.7768). However, the Bland-Altman analysis revealed a 95% CI of -50.4 to 47.5 micromol/L. Transcutaneous bilirubin was also compared to a measure on plasma in a sub-group of 53 infants, the correlation was 0.7749 with a 95% CI of -35.8 to 46.5 micromol/L. Comparing total blood bilirubin with plasma bilirubin in 35 patients, we observed a similar results with a correlation of 0.7583 and a 95% CI of -34.6 to 40.7 micromol/L. Finally, the extent of weight loss observed in our group of patients had little influence and did not affect the agreement between the 2 approaches. We conclude that the BiliCheck may be used to monitor bilirubin in term neonates at 48 hours of life even with a weight loss. Clinicians have however to be conscious of the limit of the precision of the measures both for the BiliCheck and the laboratory methods.