Article Text

Download PDFPDF

American Academy of Pediatrics guidelines for detecting neonatal hyperbilirubinaemia and preventing kernicterus
  1. D Manning
  1. Correspondence to:
    Dr Manning
    Wirral Hospital, Arrowe Park, Wirral, Merseyside CH49 5PE, UK;

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Are they applicable in Britain?

In 2004 the American Academy of Pediatrics (AAP) revised its guidelines for management of severe hyperbilirubinaemia in the newborn.1 The objective was to strike a balance between preventing severe hyperbilirubinaemia and its sequelae on the one hand and minimising overinvestigation and treatment of physiological jaundice on the other.


The previous guidelines, published in 1994,2 reflected a “kinder, gentler” approach to neonatal jaundice.3 Thanks to improvements in the management of rhesus isoimmunisation, bilirubin encephalopathy and kernicterus had virtually disappeared among term infants in the western world. Many paediatricians were concerned that this experience was extrapolated inappropriately to neonatal jaundice in general, and that infants with physiological and breast milk jaundice were undergoing unnecessary investigation and treatment. The 1994 guidelines were directed to reducing such intervention. They included the recommendation, however, that all infants discharged within 48 hours of birth should be followed up within three days.

Increasingly early postnatal discharge of term and near term infants, and initiatives to promote breast feeding, coincided with the more relaxed approach to neonatal jaundice. Ironically, these developments may have contributed to the re-emergence of severe jaundice and bilirubin encephalopathy in North America and Europe.4–6 Reported associations included relatively short gestation7 and comorbidity, particularly glucose-6-phosphate dehydrogenase deficiency.4 Possibly early discharge, before jaundice has reached a maximum and breast feeding is established, leaves some infants, particularly if immature or ill, vulnerable to severe jaundice, which would previously have been identified during inpatient postnatal stay. Non-white infants may be particularly vulnerable because, compared with white infants, jaundice is more difficult to evaluate clinically and glucose-6-phosphate dehydrogenase deficiency is more common.

The pendulum thus swung away from the “kinder, gentler” approach and led to the recent revision of the AAP guidelines. The experts who revised them acknowledged the dearth of evidence to underpin their recommendations. The general objectives were to promote breast feeding, to encourage assessment before discharge of the risk of developing later jaundice, to individualise follow up arrangements, and to review treatment advice.


One of the most important changes is to extend the scope from full term infants to those of 35 weeks gestation or greater. This change acknowledges that near term infants are being discharged earlier, and may be particularly vulnerable to jaundice because of relative feeding difficulty and hepatic immaturity.

Another substantial change is the inclusion of rather complex recommendations on risk assessment before discharge for severe jaundice, and targeted follow up based on this after discharge. Risk assessment uses a nomogram derived from measurements before discharge of serum bilirubin in 2840 term and near term infants in Philadelphia.8 These measurements were reported to predict later significant hyperbilirubinaemia. Two options are proposed for risk assessment: (a) universal total serum bilirubin measurement; (b) bilirubin measurement combined with assessment for clinical risk factors. The timing of the recommended follow up depends on the timing of discharge and the risk assessment before discharge. All infants should be examined by a health professional within a few days of discharge, and those discharged within 24 hours should be reviewed by 72 hours. Infants with risk factors should be reviewed earlier and more often than those without. The recommended location of follow up is not specified but, according to the 1994 guidelines, might be the paediatrician’s office, a community clinic, or the family home.

Acknowledging that visual assessment for jaundice severity can be unreliable, the report reviews research in transcutaneous bilirubinometry and recommends that this be evaluated further.

Finally, the guidelines give advice on treatment of moderate to severe jaundice and bilirubin encephalopathy. The phototherapy and exchange transfusion bilirubin thresholds for well infants are slightly lower than in the 1994 guidelines, and even lower thresholds are recommended for infants who are ill or have risk factors for severe jaundice.


To address this question, we need to know whether severe jaundice is re-emerging in the United Kingdom. The only recent published information is from Wirral, where the incidence of moderate jaundice (serum bilirubin ⩾340 μmol/l) increased from 2.4/1000 live births in 1991 to 5.5/1000 in 2001 (p<0.0001).9 This increase coincided with a progressive reduction in length of postnatal stay for mothers and infants.

Because of the dearth of British data, a prospective study, supported by the British Paediatric Surveillance Unit, to determine the incidence of severe neonatal jaundice (unconjugated serum bilirubin ⩾510 μmol/l in the first month of life) in Britain and Ireland is in progress. Secondary objectives are to document associated clinical and demographic variables and short term outcomes. Surveillance started in June 2003 and finished in May 2005. Early returns indicate that severe jaundice and bilirubin encephalopathy, although uncommon, are occurring in Britain and Ireland. Associations are similar to those reported in North America: relative immaturity, early discharge, breast feeding, ethnic minority origin, and comorbidity such as haemolysis and sepsis. Raising awareness of these associations therefore is just as relevant here as in North America.

The aim of the guidelines to support breast feeding is laudable, but the specific recommendations—advising mothers to feed their infants 8–12 times a day and advising against routine supplementation with water and dextrose—are accompanied by little detail on appropriate support. In Britain, this support is provided by the midwife who, by continuing care after hospital discharge, bridges the gap between secondary and primary care of mother and infant. The AAP recommendations on supporting lactation have little to add to current British practice.

Given the lack of evidence to support specific bilirubin threshold levels for phototherapy and exchange transfusion, the recommendations on treatment of moderate and severe jaundice seem reasonable and do not differ greatly from British practice. The principle of recommending lower thresholds for ill or vulnerable infants is also reasonable, as is the promotion of further research in transcutaneous bilirubinometry.

The most controversial recommendations, and those that are arguably least applicable here, are those pertaining to individualised risk assessment and targeted follow up. Even if early bilirubin measurements were sufficiently predictive of risk, arguably a nomogram would have to be devised for each population of infants to which it applied, reflecting the population incidence and severity of hyperbilirubinaemia, the uptake of breast feeding, and the incidence of haemolytic disease due to blood group incompatibility and glucose-6-phosphate dehydrogenase deficiency.

I am concerned, moreover, that risk assessment and follow up based largely on bilirubin concentrations may be insufficient safely to identify infants at risk for severe hyperbilirubinaemia. Although the guidelines rightly draw attention to other relevant clinical and demographic risk factors, these are too common to be sufficiently predictive of risk. Individualised frequency and timing of follow up arrangements may be too complicated to be implemented effectively. Indeed, this may be the American experience; of 61 infants known to the Pilot Kernicterus Registry who were readmitted from home with severe jaundice, follow up according to the 1994 guidelines had not been arranged in 44 cases.4 Appointments had been arranged for a further three infants, but had been missed. This system failure may have indicated complacency about severe jaundice, but probably also reflected the difficulties of implementing individualised follow up programmes.

Currently in the United Kingdom, all mothers and infants receive regular postnatal home visits from their midwives.10 Although practice may vary by district, usually the first visit takes place on the day after hospital discharge. Although there is no evidence that this practice is superior to targeted surveillance,10 the British system may be more effective than the American in the identification and timely management of severe jaundice. This can progress rapidly in haemolytic disease and in breast feeding failure complicated by dehydration, and so may not always be predicted by bilirubin concentration before discharge nor intercepted by interval follow up. Also, the midwife’s visits serve more purpose than visual assessment for jaundice. They offer emotional and practical support to the mother, and permit assessment of the infant’s state of hydration and general wellbeing. There is, indeed, some evidence to suggest that home visiting is effective in promoting successful breast feeding.10 It may also offer better access to health professionals, particularly for poor and isolated families, than systems that require the infant to be taken to a clinic or office and which put the onus on the family to seek help if the infant is unwell.

Although British postnatal home visiting has been subjected to little evaluation, a recent American study compared the cost effectiveness of universal follow up with review based on bilirubin concentrations before discharge, and estimated that the former was more expensive.11 The authors acknowledged, however, that universal follow up may have benefits other than surveillance for jaundice, so its overall cost effectiveness is likely to be greater than their estimates. Pending further evidence, I see no reason for British practice to change to the American model of targeted surveillance for severe jaundice.


The revised AAP guidelines are valuable in combating complacency about neonatal jaundice, in highlighting clinical and demographic factors that demand vigilance, and in informing treatment. All of these are relevant to British neonatal care. The case has not been made, however, to abandon the British system of midwifery based universal surveillance of recently discharged infants in favour of risk assessment and targeted surveillance.

Are they applicable in Britain?


View Abstract


  • Competing interests: none declared