Infants born at the well baby nursery of the University “Federico II” of Naples with gestational age over 35 weeks were enrolled between January and December 2009. The investigation protocol was approved by the local ethical committee and parental consent was obtained.
Infants with Rh or ABO isoimmunisation, major congenital malformations, haemoglobinopathies or evidence of liver disease were excluded from the study.
Infants with post-natal age from 4 to 75 hours were measured TSB by an experienced resident (SL) as ordered by the attending clinician unaware of the study purpose. Samples were obtained from heelsticks in capillary tubes, centrifuged at 3000 rpm for five minutes and then read by a direct spectrophotometer (GB13A, Bertocchi Elettromedicali, Cremona, Italy). Spectrophotometry was used as a reference as it is the technique used in daily routine and has better agreement with high pressure liquid chromatography (HPLC) than any other technique .
Within 20 minutes before TSB determination, TcB was measured on the infant’s forehead, protected from direct sunlight and avoiding areas with hair, bruises, nevi or other skin anomalies.
We used three well known transcutaneous bilirubinometers according to the manifacturer’s instructions. Bilicheck (SpectRX, Norcross, GA) was calibrated using a Bilical before each measurement, calculated averaging five readings. The device scans the whole spectrum of visible light and automatically subtracts the light reflected by confounding factors like haemoglobin or melanin. BiliMed (Medick, S.A. France) operates by ten light emitting diodes that form a single beam when hitting the newborn skin. A silicon diode captures the reflected light and the internal software calculates the highest intensity for every wavelength range. BiliMed was used as recommended at 2 cm from the neonate’s forehead. JM-103 (Draeger Medical Systems, Inc., Telford, PA) uses 2 wavelengths and a dual optical path system; each measurement was calculated averaging five readings.
All measurements were performed in the ambient morning light of the nursery, in random order according to a computer generated randomization table.
Data were analyzed using MedCalc software v.11 (MedCalc Software, Mariakerke, Belgium). Linear regression analysis, Pearson correlation coefficients, Bland-Altman plots and receiver operating curve (ROC) were performed.