Aims: To determine the accuracy of Bilicheck in assessing serum bilirubin. To evaluate the effectiveness of Bilicheck as a screening device in a population of jaundiced term and near-term babies. Methods: Prospective evaluation of paired Bilicheck measurements and serum bilirubin in 300 babies. Results: There was a strong correlation between Bilicheck and serum bilirubin. The mean difference between Bilicheck and serum bilirubin was —10.7 μmol/l, but the 95% limits of agreement were wide at —80 μmol/l to +60 μmol/l. Bilicheck identified significant jaundice (serum bilirubin above 250 μmol/l) with a sensitivity (95% confidence interval) of 91% (88% to 94%) and specificity of 66% (60% to 71%). The area under the receiver operator characteristic curve was 0.85. There were five false negatives, all of whom had a serum bilirubin below 300 μmol/l. If Bilicheck had been used as a screening device, there would have been a 55% decrease in blood samples taken in these babies.
Conclusion: Bilicheck cannot be used to measure serum bilirubin in term and near-term babies with jaundice, but is an effective screening method, which can be used to safely reduce the number of blood tests in these babies.