Background Currently, maternal Rh-D antibody levels are primarily used to triage which alloimmunized women warrant enhanced surveillance with middle cerebral artery Doppler. Traditionally, maternal Rh-D antibody levels ≤15 IU/ml have indicated, at worst, mild anaemia and provided reassurance. This threshold has not been widely studied.

Methods A prospective cohort study of all intrauterine fetal transfusions (IUT) for Rh-D alloimmunization performed at our tertiary fetal medicine unit from 1996–2011. Fetal haemoglobin (Hb) levels at the time of IUT were adjusted for gestational age (multiples of median [MoM]) and correlated with the maternal serum Rh-D antibody level taken on the day of IUT, or ≤2 weeks prior to the transfusion.

Results 260 IUTs were performed, of which 195 were for Rh-D alloimmunzation in 82 pregnancies. No significant correlation was demonstrated between fetal Hb and serum antibody levels (Spearman r = 0.08; p = 0.35). Rates of mild (0.65–0.84 MoM), moderate (0.55–0.64 MoM) and severe (<0.55 MoM) fetal anaemia were 32%, 22% and 31% respectively. The sensitivity, specificity, PPV and NPV of a maternal antibody threshold of >15 IU/ml for detecting any fetal anaemia (<0.84 MoM) were 88%, 14%, 85% and 18%. The equivalent results for a threshold of >15 IU/ml detecting moderate-severe anaemia (<0.65 MoM) were 88%, 12%, 52% and 47%. Using a lower antibody threshold of >8 IU/ml, the sensitivity, specificity, PPV and NPV of maternal serum antibody levels detecting any fetal anaemia were 96%, 5%, 85% and 17% respectively.

Conclusion The widely used Rh-D threshold of >15 IU/ml may miss a substantial proportion of cases of fetal anaemia.