Management of jaundice in newborn nurseries – measuring, predicting and avoiding the sequelae

Abstract

This issue of Acta Paediatrica contains three studies of neonatal jaundice (1–3), each in its own way adding to our ability to manage neonatal jaundice safely. Although neonatal jaundice is so common as to perhaps be considered banal, few issues in neonatal care appear quite as vexing as the management of the jaundiced infant. Thus, we are faced with imprecision in measurement, imprecision in treatment and uncertainty in prediction. No wonder that we appear unable to avoid the tragedy of kernicterus (4). Measurement of bilirubin has been performed for decades mostly on blood samples (whole blood, plasma or serum). Although there is a plethora of methods available for such analysis, none appears perfect judging by the many complaints about the lack of precision in bilirubin measurements (5). A large manufacturer of analytical equipment recently announced that they had decided to shift their bilirubin standard (down) by 9%, recognizing that the numbers which had previously been reported from their machines had differed by that percentage. Not surprisingly, attempting to estimate the serum bilirubin value by measuring bilirubin in the skin is affected with even greater imprecision. In spite of this, transcutaneous measurement of bilirubin (TcB) has become a common method to screen neonates for jaundice. If TcB values are used, not to make treatment decisions, but simply to decide whether or not a blood test is needed, many unnecessary blood tests can be avoided. One of the pitfalls of TcB has been the contribution of non-bilirubin pigments (such as melanin) in the infants’ skin to the value read by the machine. Wainer et al. (1) have aimed to improve on previous studies by assessing the infants’ skin colour according to two defined reference colours, rather than by racial assignations or other visual estimates of skin colour. Using the Minolta ⁄AirShields JM-103 jaundice meter, which was designed to minimize the impact of melanin on measurements, the authors found the best precision and lowest bias in infants judged to be of ‘medium’ skin tone, i.e. between the two reference colours. The generalizability of their results is limited by the fact that only 1.9% of the studied infants had ‘dark’ skin colour. In addition, it seems that whereas the assignation into the three colour categories was performed in 79% of the cases by the study coordinator, and thus less subject to inter-observer variability, the lighting conditions under which skin colour assignment was made are not described or defined in the study. There may be two challenges here, one being that the colour assignment may not work quite as well as when performed as it probably would be under normal working conditions, by a larger group of different people. The other problem is that an assessment of colour is likely to vary with the strength and colour of ambient light. Nevertheless, studies like this are important because they highlight the strengths as well as limitations of methods in routine use, and thus are likely to make us more attentive to their potential pitfalls. In recent years, the length of stay in maternity hospitals and birthing units has become shorter. Thus, for many Invited Commentary for Wainer et al. Impact of skin tone on the performance of a transcutaneous jaundice meter, Bakkeheim et al. Maternal IgG anti-A and anti-B titres predict outcome in ABO-incompatibility in the neonate, Bental et al. Bhutani-based nomograms for the prediction of significant hyperbilirubinaemia using transcutaneous measurements of bilirubin. Acta Paediatrica ISSN 0803–5253