Page 72 - 16Neonatal Jaundice_compressed

Factors that influence hyperbilirubinaemia and kernicterus

hyperbilirubinaemia showed a three-fold higher risk of developing hyperbilirubinaemia
compared with those who had prior sibling without hyperbilirubinaemia (10.3% versus 3.6%;
OR 3.1, 95% CI 1.4 to 6.8). In the next stage of analysis, potential confounding factors (race,
sex, gestational age, maternal age, year of birth, delivery type, gravidity, breastfeeding, obstetric
anaesthesia and neonatal asphyxia) were adjusted in a logistic regression analysis and the risk of
recurrence assessed for different degrees of jaundice: mild (peak serum bilirubin levels
≤ 205 micromol/litre), moderate (205–256 micromol/litre) and severe hyperbilirubinaemia
(≥ 256 micromol/litre). The results showed a clear trend of increasing sibling risk with
increasing severity of hyperbilirubinaemia. There was a 2.7 times higher risk of mild jaundice in
newborns who had a sibling with mild jaundice (25.3% versus 11.1%; OR 2.7, 95% CI 1.8 to
4.1), and the risk was four times greater for the moderate jaundice group (8.8% versus 2.3%;
OR 4.1, 95% CI 1.5 to 10.8). Babies who had a prior sibling with severe hyperbilirubinaemia
showed a 12 times higher risk of developing jaundice compared with those who had no sibling
with severe hyperbilirubinaemia (10.5% versus 0.9%; OR 12.5, 95% CI 2.3 to 65.3). [EL II]
17
In another nested case–control study from the USA, the charts of 11 456 babies were searched
electronically to identify babies who had been readmitted for hyperbilirubinaemia (total serum
bilirubin > 291 micromol/litre). Babies who had received phototherapy before discharge were
excluded. A total of 75 babies (0.7%) constituted the test group, and these were matched with
75 randomly selected controls who had not been readmitted. The two groups were compared
and a step-wise logistic regression analysis to determine the smallest subset of predictors of the
difference between the groups. Three factors were identified: early gestation (for 35 to
36 6/7 weeks: adjusted OR 20.79, 95% CI 2.34 to 184.74; for 37 to 37 6/7 weeks: adjusted
OR 14.86, 95% CI 1.91 to 115.38), exclusive breastfeeding (adjusted OR 10.75, 95% CI 2.37 to
48.82 and finally transcutaneous bilirubin levels above the 95th percentile on the Bhutani
nomogram (adjusted OR 149.89, 95% CI 20.41 to > 999.99). [EL II]

A survey of mothers of babies with gestational age ≥ 35 weeks discharged from a well-baby
18
nursery of a health maintenance organisation in the USA was conducted to evaluate how
closely mother’s race documented in medical records correlated with self-reported race, and to
analyse the correlation between mother’s and newborn’s race in the context of risk for neonatal
hyperbilirubinaemia. Maternal and neonatal data were extracted from the organisation’s
database and maternal race was placed in one of seven categories. Further information from the
mothers about their experience of breastfeeding, neonatal care, hyperbilirubinaemia detection,
interventions and education, and racial ancestry for mother, father and newborn (allowing up to
five responses for ancestry of each) was elicited through a computerised telephone survey. Of
the 3021 mothers available for potential inclusion, only 41% could be contacted and, of them,
69% (866 of 1248) completed the survey. Of these, 145 mothers were documented as white in
the medical records, but only 64% of them self-reported as white, while, of 427 mothers
documented as black in medical records, only 70% self-reported as black. For mothers of Asian
and Middle Eastern origin, the agreement between the two sources was 35% and 50%,
respectively. About 15% of the mothers described themselves as being of multiracial (two or
more races) origin and 9% reported that the father was multiracial, but only 11% (93 of 866)
reported their baby as multiracial. When racial ancestry was further explored among the
newborns reported as being of two or more races, the primary race matched that of the parents
in 41% of cases only. In 23% of babies, the primary race was assigned to the mother’s race and
in 25% to the father’s race, with 11% assigned to the race of neither mother nor father.
Moreover, of the 70 newborns born to parents of different ethnic origins, only 64% were
reported as multiracial. [EL III]

Evidence summary
There is consistent evidence from good-quality studies to show that four factors are
independently associated with an increased risk of hyperbilirubinaemia – gestational age
< 38 weeks, jaundice within 24 hours of birth, increase in severity of clinically apparent
jaundice and intention to breastfeed exclusively. Five studies evaluated family history of
jaundice as a risk factor and four found it to be statistically significantly associated with
hyperbilirubinaemia. Bruising was reported as a statistically significant risk factor in only two
studies. Results from most studies show no statistically significant association between
cephalohaematoma, vacuum delivery, male sex or race and hyperbilirubinaemia.

45