A new technique in exchange transfusion.

In experienced hands exchange transfusion carries a very low operative risk. Nevertheless, infants who are initially very anaemic present a grave problem as this treatment is their only hope of survival. There are also other infants who appear well at the start of the transfusion, but become ill or die during the procedure. Walker and Neligan (1955) discussed these problems and listed the probable factors involved as heart failure, shock and possibly biochemical changes. Campbell (1955) considered that deterioration during exchange transfusion was in many cases probably due to electrolyte disturbance, mainly hyperkalaemia. He noted, however, that in two infants having similarly high levels of serum potassium, one was ill and one quite well. Farquhar and Smith (1958) stated that 'changes in serum electrolytes were shared by infants who were disturbed and by those who showed no abnormality'. In a series of 18 consecutive cases of haemolytic disease studied in this hospital during 1958-59, serum electrolyte estimations and simultaneous electrocardiograph tracings were done at every 100 ml. during the exchange. There was no correlation between deterioration of the infant's condition observed clinically and the serum level of either calcium or potassium, and not a single electrocardiogram showed any evidence of electrolyte disturbance. Even when donor blood with a potassium level of 19-2 mEq/litre was used in one case there was no clinical distress, and the infant's serum potassium and electrocardiogram remained normal. It has been my impression that with increasing experience and our present method of using packed donor cells, disturbances due to hyperkalaemia and hypocalcaemia play a decreasing role. I agree with O'Neill and Gordon (1959) that cardiac embarrassment is a more frequent cause of clinical distress. This may occur not only in infants who are already very anaemic and ill, but also during the exchange in an otherwise well baby. This paper attempts to assess the value of venous pressure measurements in the umbilical catheter during exchange transfusion and control of the speed of transfusion in relation to cardiac failure.