Page 70 - 16Neonatal Jaundice_compressed

Factors that influence hyperbilirubinaemia and kernicterus

(low, medium and high risk based on the hour-specific bilirubin centiles, gestational age and
direct antiglobulin test (DAT) results) and the difference between their serum bilirubin levels
and the AAP phototherapy threshold levels. Data on all variables were extracted from electronic
and paper records of admissions, outpatient visits and home health visits. The cases and controls
did not differ significantly by sex, race, birthweight or duration of hospitalisation. Moreover, the
two groups had similar mean serum bilirubin levels and percentage weight loss from birth.
Bivariate analysis showed that lower gestational age, bruising on examination, serum bilirubin
concentration between 291 and 392 micromol/litre occurring during birth hospitalisation, serum
bilirubin increase of ≥ 102 micromol/litre per day, and exclusive breastfeeding (after qualifying
serum bilirubin levels) were statistically significantly associated with an increased risk of
hyperbilirubinaemia (P < 0.04), while inpatient phototherapy was found to significantly reduce
the risk. Multivariate analysis revealed that the strongest predictors of increased risk of severe
hyperbilirubinaemia were lower gestational age (adjusted OR 3.1, 95% CI 1.2 to 8.0 for 38–
39 weeks and adjusted OR 3.7, 95% CI 0.6 to 22.7 for 34–37 weeks compared with 40+
weeks as the reference), bruising on examination (adjusted OR 2.4, 95% CI 1.2 to 4.8), serum
bilirubin increase of ≥ 102 micromol/litre per day (adjusted OR 2.5, 95% CI 1.2 to 5.5) and
exclusive breastfeeding after reaching the qualifying serum bilirubin levels (adjusted OR 2.0,
95% CI 1.03 to 4.0). It was also reported that male sex, race, and the mode of feeding before
the qualifying bilirubin level did not predict severe hyperbilirubinaemia. [EL II]
12
In a retrospective cohort study conducted in a community teaching hospital in the USA, a
clinical risk factor score was developed and its predictive accuracy was compared with pre-
discharge serum bilirubin measurements plotted on the bilirubin nomogram. The study
population included babies with birthweight ≥ 2000 g (if gestational age ≥ 36 weeks) and
birthweight ≥ 2500 g (if gestational age ≥ 35 weeks) who participated in the hospital’s early
discharge programme and who had both pre- and post-discharge serum bilirubin measured.
Hyperbilirubinaemia was taken as post-discharge serum bilirubin level > 95th centile on the
nomogram. Hospital records were reviewed retrospectively to collect information on various
risk factors (baby, maternal, pregnancy and delivery factors), and their association with
hyperbilirubinaemia was explored by univariate analysis. All factors found to be associated with
the outcome at P < 0.2 level of significance were considered for the final risk factor score
based on logistic regression modelling. For univariate analysis, the baby factors found to be
associated with an increased risk of hyperbilirubinaemia (at P < 0.2 level of significance)
included gestational age < 38 weeks and ≥ 40 weeks, large for gestational age, high pre-
discharge serum bilirubin and higher birthweight; the maternal factors included maternal
diabetes, breastfeeding and combined breast- and bottle-feeding; the pregnancy, labour and
delivery factors included vacuum extraction, prolonged rupture of membranes and oxytocin
use. Three factors were found to be associated with decreased risk of hyperbilirubinaemia: small
for gestational age, parity and caesarean section. All these factors were then analysed for the
final risk factor model using step-wise logistic regression, except for pre-discharge serum
bilirubin level/risk zone, which was analysed separately. Results from the regression analysis
showed the following factors to be statistically significantly associated with hyperbilirubinaemia:
gestational age < 38 weeks (adjusted OR 2.6, 95% CI 1.5 to 4.5), oxytocin use during labour
(adjusted OR 2.0, 95% CI 1.2 to 3.4), vacuum delivery (adjusted OR 2.2, 95% CI 1.5 to 3.6),
exclusive breastfeeding (adjusted OR 2.6, 95% CI 1.5 to 4.5), combination of breast- and bottle-
feeding (adjusted OR 2.3, 95% CI 1.1 to 4.9), and birthweight (for every 0.5 kg increase above
2.5 kg: adjusted OR 1.5, 95% CI 1.2 to 1.9). The predictive accuracy of pre-discharge serum
bilirubin level/risk zone was evaluated separately from the risk factor model, and it was shown
to predict hyperbilirubinaemia more accurately than the risk factor model alone. [EL II]
13
A prospective cohort study from Israel evaluated the ability of prenatal and intrapartum
characteristics and early serum bilirubin measurements to predict hyperbilirubinaemia in
healthy term babies. The study included 1177 babies (≥ 37 weeks of gestation). Babies with
either blood group incompatibility with a positive direct DAT or G6PD deficiency were
excluded. Serum bilirubin levels were obtained within the first 8 to 24 hours of life and
repeated daily for the next 4 days. In all, 5.1% (60 of 1177) of babies developed
hyperbilirubinaemia (defined as serum bilirubin level > 171 micromol/litre at day 2,
> 239 micromol/litre at day 3, and > 291 micromol/litre at day 4–5. Using multiple logistic
regression analysis, serum bilirubin level > 85 micromol/litre on day 1 had a statistically

43