Page 96 - 16Neonatal Jaundice_compressed

Early prediction of serious hyperbilirubinaemia

positive in 34.1% of babies (31 of 91). A positive DAT and the cord blood bilirubin threshold
≥ 68 micromol/litre both showed a sensitivity of 92.3% in predicting subsequent severe
hyperbilirubinaemia. Specificities for both positive DAT and cord blood bilirubin tests were
75.6% and 100%, respectively. [EL III]
38
A Norwegian study examined the ability of universal DAT testing to predict the need for
53
phototherapy, using the Hillingdon Hospital bilirubin chart to inform treatment. The study
population included 2463 babies born in a general hospital. Exclusion criteria included high-
risk deliveries and severe neonatal illness but no more details were given. Information on
gestational age, birthweight, gender and ethnicity was not provided. Phototherapy was started in
term babies at serum bilirubin > 350 micromol/litre at > 72 hours and > 250 micromol/litre
at > 120 hours. Blinding of outcome assessors was not specified. DAT was positive in 4.1% of
babies (100 of 2463). The DAT test showed a sensitivity of 14.4% and specificity of 96.6% in
predicting the need for phototherapy. [EL III]

A Taiwanese study evaluated selective DAT testing and cord blood bilirubin as predictors of
40
hyperbilirubinaemia. The study population included 88 babies with birthweight > 2500 g born
to group O, Rhesus-positive mothers; 53 babies were ABO incompatible. Information on
ethnicity, gestational age and gender was not provided. Serum bilirubin levels were measured
daily for 1 week. Hyperbilirubinaemia was defined as serum bilirubin > 255 micromol/litre
within 96 hours of birth and/or early jaundice with serum bilirubin > 171 micromol/litre within
24 hours of birth. Blinding of outcome assessors was not specified. DAT was positive in 26%
(14 of 53). The DAT test and the cord blood bilirubin threshold level > 68 micromol/litre
showed sensitivity of 45% and 41% in predicting subsequent hyperbilirubinaemia. The
specificity for the DAT and cord blood bilirubin tests was 96% and 100%, respectively. [EL III]
A Turkish study examined selective DAT to predict serum bilirubin levels at 6, 30, 54, 78 and
39
102 hours. All babies > 38 weeks of gestation with blood groups A or B born to mothers with
blood group O, without a simultaneous Rhesus blood factor incompatibility, were included
(n = 150). The mean birthweight was 3212 ± 415 g and 51% were male. Ethnicity was not
specified. No exclusion criteria were specified but data from 14 babies were excluded from the
final analysis for clinical reasons (transferred to intensive care or no informed consent given).
Severe hyperbilirubinaemia was defined as serum bilirubin > 85 micromol/litre with an
increase of 8.5 micromol/litre per hour in the first 24 hours, levels > 205 micromol/litre on
day 2, > 255 micromol/litre on day 3, or > 289 micromol/litre on days 4 and 5. Blinding of
outcome assessors was not specified. DAT was positive in 4.4% (6 of 136) of babies. A positive
DAT showed a sensitivity of 20.1% and a specificity of 100% in predicting subsequent severe
hyperbilirubinaemia in babies with ABO incompatibility. [EL III]
Evidence summary

Each study compared DAT with varying threshold levels of bilirubin. In the EL II study the DAT
test showed a sensitivity of 8.5% and specificity of 97.6% in detecting haemolysis. Similar levels
of sensitivity and specificity in predicting subsequent hyperbilirubinaemia were found in three
of the other four EL III studies. Sensitivity ranged from 14.4% to 44.8% and specificity from
95.8% to 100%. The fourth EL III study showed a sensitivity of 92.3% and specificity of 75.6%.

GDG translation from evidence
Routine DAT (Coombs’) testing on umbilical cord blood does not accurately predict subsequent
hyperbilirubinaemia in healthy newborns.

The GDG appreciates that the current widespread use of antenatal anti-D prophylaxis in Rhesus-
negative women has influenced the interpretation of an early DAT in their newborns. Passive
antibody transfer commonly results in a weakly positive DAT in the absence of haemolysis.
54
However, a strongly positive DAT, particularly in a baby of a woman who did not receive anti-
D during pregnancy, should still be considered an important marker of haemolysis and forms
part of the formal assessment of a baby with significantly elevated bilirubin levels (see also
Chapter 6 on formal assessment).

67