Transcutaneous bilirubinometry: a comparison of old and new methods.

Two devices for hyperbilirubinemia screening were compared: the Minolta jaundice meter and the 30-year-old Ingram icterometer. Serum bilirubin values were obtained from a population of 106 healthy newborns with jaundice. A linear correlation existed between serum bilirubin values and the readings on the jaundice meter (y = 0.91x + 9.25, r = .74, standard error of the estimate (sy X x) = 2.02). The device classified hyperbilirubinemia with a sensitivity of 94% and specificity of 77.5%. A linear correlation existed between serum bilirubin values and the readings on the icterometer (y = 0.17x + 1.74, r = .63, sy X x = 0.38). The device classified hyperbilirubinemia with a sensitivity of 82% and specificity of 74%. The population studied was preselected for the presence of jaundice and each data point obtained was from a different patient. As a result, the correlation coefficient obtained from the jaundice meter is lower than others reported, but is more representative of the value that should be obtained under routine nursery situations. Both devices perform well as screening devices; at optimal cutpoints, no statistically significant difference in their efficacy can be demonstrated. Despite the availability of more complex and expensive devices, the icterometer continues to serve as a cost-effective screening device for hyperbilirubinemia.