A. Meberg et Kb. Johansen, Screening for neonatal hyperbilirubinaemia and ABO alloimmunization at thetime of testing for phenylketonuria and congenital hypothyreosis, ACT PAEDIAT, 87(12), 1998, pp. 1269-1274
In a population-based study including 2463 infants, serum bilirubin measure
ments were added to the neonatal screening programme for phenylketonuria an
d congenital hypothyreosis. This screening programme detected 11/17 (65%) o
f infants with serum bilirubin levels >350 mu mol l(-1), of whom 7 (3 per 1
000) were readmitted from home (6 treated with phototherapy). A total of 13
9 infants (5.6%) received phototherapy. Maternal blood type O occurred sign
ificantly more often in term infants treated (30/54; 55.6%) compared with p
reterm infants treated (32/85; 37.6%) and with blood type O occurrence in t
he total population of mothers (906/2426; 37.3%) (p < 0.05). The blood type
constellations mother O/infant A or B showed a sensitivity of 64%, specifi
city 65%, positive predictive value 12% and a negative predictive value of
96% for the requirement of phototherapy for the whole material. Exchange tr
ansfusion was not required in any of the infants. No infant developed bilir
ubin encephalopathy (kernicterus). Adding bilirubin to a neonatal screening
programme detects some cases with unexpectedly high bilirubin levels in ne
ed of intervention. Routine ABO blood typing of pregnant women, ABO cord bl
ood typing and Coombs' test in infants of mothers with blood type O cannot
be recommended because of low positive predictive value for the requirement
of intervention (phototherapy) by these tests.