RESUMO
OBJECTIVE: Transcutaneous bilirubinometers are useful in screening and monitoring neonatal hyperbilirubinemia. These instruments reduce the number of blood samples needed, but their validity has been questioned. The aim of this study was to evaluate the accuracy and reliability of a classic transcutaneous bilirubinometer. METHOD: Descriptive observational cross-sectional study of a sam-ple of 117 simultaneous measurements of transcutaneous and serum bilirubin in 82 jaundiced newborn infants. We analyzed the agreement between repeat measurements and measurements in different sites. We developed predictive models of serum bilirubin and analyzed their validity. RESULTS: All points of measurement displayed excellent agreement between repeat measures (intraclass correlation co-efficient higher than 0.92). Forehead measurements had a linear correlation coefficient of 0.75; in the sternum II-near correlation coefficient varied according to the absence or presence of phototherapy (0.86 and 0.54 respectively). The error intervals in which predictions of se-rum bilirubin were found were #+ 2 mg/dl for low values and #+ 4 mg/dl for high values (#> 15 mg/dl). CONCLUSIONS: Reliability of the transcutaneous bilirubinometer evaluated was high, but the substantial degree of error in se-rum bilirubin make its accuracy questionable, especially in patients with high levels.
