A-B-O incompatibility and haemolytic disease of the newborn.

Since the work of Boorman, Dodd and Trinick (1949), incompatibility in the ABO system has been recognized with increasing frequency as a cause of haemolytic disease. Robinson, Phillips and Prystowsky (1951), Shumway, Miller and Young (1955), Davidsohn (1956) and others have written on this condition. And yet haemolytic disease due to ABO incompatibility is much less widely known than that due to Rhesus incompatibility. The failure widely to recognize ABO disease may be due to several reasons. Chiefly it is due to the fact that it is usually not a serious condition. In its severe degree it is uncommon. In its common form it is so slight as scarcely to be recognized. So the disease is held to be rare. Belief that a disease is rare leads to a lessened vigilance in its detection, and it is remarkable how once a disease is found to be not uncommon how quickly it becomes commonplace. Then again, that elegant test, the direct antiglobulin test of Coombs, which so readily detects haemolytic disease in blood group incompatibilities other than ABO, is usually negative by the more usual techniques. By the special technique of Rosenfield (1955), however, it may be shown to be positive. With heterospecific pregnancies with A or B incompatibility, antibodies are by definition present in the blood of the mother, although where disease is present they are believed to be of a different nature from the 'natural' antibody. In Rhesus disease the presence of antibody with Rhesus incompatibility indicates disease. It is not necessarily so in ABO disease. Moreover, no immune antibody may be detectable ante-natally in ABO disease, although it may appear after delivery. Unlike Rhesus disease, ABO disease cannot be anticipated by ante-natal testing with any certainty. Finally, it may be that the term 'physiological jaundice' makes for a too easy road to diagnostic apathy, and one that is trod too often. It is stated by Davidsohn (1956) that ABO haemolytic disease appears in 1 in 3,000 random pregnancies and by Gunson (1956) that it accounts for 15% of all cases of haemolytic disease of the newborn. Certainly it is generally believed to be less common than disease due to Rhesus incompatibility. However, Rosenfield (1955) found laboratory evidence of abnormality attributable to ABO incompatibility in 38 cord blood specimens of 1,480 random samples. The present small survey purports to show that clinically recognizable ABO haemolytic disease is a common condition and that it may quite often need close watching and active treatment.