ABSTRACT
Approximately 50% of newborn develop jaundice, however only 6% of them have significant hyperbilirubinemia that need more evaluation. Actually, most of sampling [50-90%] for detection of hyperbilirubinemia is not necessary and cause some problems such as infection, tissue injury, pain, discomfort, wasting time and money. So non-invasive methods is desirable to decrease unnecessary sampling. This study performed on neonates who referred to Shiraz University of Medical Science affiliated Nemazee hospital neonatal emergency room due to jaundice. Initially 92 neonates were selected. After taking digital images and calculating mean and 95 percentile yellow wave's lengths of forehead skin and detection of total serum bilirubin, the correlation between wave's lengths and serum total bilirubin were determined. 95 percentile of skin yellow wave lengths that determined bilirubin level of 5, 10, 15 and 20 mg/dl were used for making 4 yellow strips of the icterometer. This icterometer was used for screening of 209 neonates who referred for evaluation of jaundice. Using icterometer in evaluation of 209 neonates with jaundice demonstrated sensitivity 94.7%, specificity 55.2%, positive predictive value of 85%, negative predictive value of 95% and accuracy 87%. This Icterometer can be applied as a screening tool in neonatal hyperbilirubinemia to detect less significant degree of hyperbilirubinemia that doesn't need sampling in Shiraz and even other regions of Iran with the same skin color