Neonatal jaundice is a common condition, most often caused by normal physiologic mechanisms and not usually of significant concern. The decision to obtain a serum bilirubin level in a newborn usually is based on the child's appearance and the clinical judgment of the physician. More than 30 years ago, clinical jaundice was correlated with elevated serum bilirubin levels. Since then, other studies that have tried to correlate specific dermal zones with elevated serum bilirubin concentration have been inconclusive. Moyer and colleagues evaluated the accuracy of clinical observation in the management of neonatal icterus by assessing whether experienced observers agreed in describing the extent of neonatal jaundice. A second goal of the study was to evaluate the reliability of visual assessment as an indicator for elevated serum bilirubin levels in neonates.

Neonates in the well-infant nursery of an urban public hospital who were of at least 36 weeks' gestational age and who weighed at least 2,000 g (4 lb, 6 oz) were eligible for the observational study. Serum bilirubin levels were obtained initially if the infant appeared icteric or if the mother was Rh-negative or had a positive Coombs test. At the time the bilirubin level was obtained, two observers (e.g., pediatrician, pediatric nurse practitioner or pediatric resident) independently examined the infant, assessing prespecified areas of the body that are thought to reflect cephalocaudal progression of jaundice. They also assessed sites that experienced pediatricians reported as being useful to identify clinical jaundice (e.g., conjunctiva, tip of the nose, palate). For each site, the jaundice was categorized as absent, slight or obvious, and the infant's skin tone was noted to be light or dark. Each observer also was asked to predict the infant's serum bilirubin level. A serum bilirubin level was obtained within one hour of the clinical assessment.

A total of 122 infants were included in the study, and mean age at assessment was two days after birth. Interobserver agreement on infant skin tone was good, but agreement on the presence of jaundice at various anatomic levels and sites was poor. The correlation between observers' estimations of the serum bilirubin level also was poor. The correlation between observer estimation of serum bilirubin level and the measured serum level was only slightly better. The presence of jaundice that extended below the middle of the chest (nipple line) had the highest sensitivity and specificity for predicting an elevated serum bilirubin level. Infants with no jaundice below the nipple line consistently had a bilirubin concentration of less than 12.0 mg per dL (205 μmol per L). The converse, however, was not true. The presence of jaundice below the nipple line did not reliably predict serum bilirubin level.

The authors conclude that clinical examination of newborns alone is not a reliable way to predict the presence or absence of jaundice. The only exception is when there is no jaundice below the nipple line. In these cases, serum bilirubin levels are not significantly increased. The authors emphasize that serious illness from neonatal jaundice is rare; therefore, they suggest that indications for bilirubin testing be based on risk factors for severe hyperbilirubinemia rather than on clinical observation alone.