Attempt to improve transcutaneous bilirubinometry: a double-blind study of Medick BiliMed versus Respironics BiliCheck

Objectives: To compare the accuracy of a new transcutaneous bilirubinometer BiliMed (Medick SA,, Paris, France) with BiliCheck (Respironics, Marietta, GA, USA), a widely available instrument, and with total serum bilirubin measurement. Design: A prospective double-blind study comparing the two devices was carried out. 686 healthy newborns needing measurement of their bilirubin were enrolled over a 4-month period. Serurn and transcutaneous bilirubin measurements were taken with both devices within 15 minutes. The order of use of the instruments was randomised. Setting: Well-baby nursery ward in a university hospital, tertiary referral centre. Results: The linear regression analysis showed a better correlation between BiliCheck and serum bilirubin (r = 0.75) than between BiliMed and serum bilirubin (r = 0.45). BiliCheck variability (+/- 2 SD of the mean bias from serum bilrubin) was within -87.2 to 63.3 mu mol/l, while BiliMed variability was within -97.5 to 121.4 mu mol/l. The receiver operating characteristic analysis (for serum bilirubin levels >205.2 mu mol/l or >239.4 mu mol/l) showed significantly higher areas under the curve for BiliCheck than those for BiliMed (p<0.001). Conclusions: Despite the potential practical advantages of BiliMed, its reduced diagnostic accuracy in comparison with BiliCheck does not justify its use in clinical practice.