Automated method for exchange transfusion.

Automated method for exchange transfusion Sir, Philpott and Banerjee (Archives, 1972) are to be congratulated on their method of exchange transfusion using a two-channel Holter pump. Many have been interested in this technical problem, and the Holter pump provides an acceptable solution. The pumping rate of the machine is accurate to ±3%. More importantly, the discrepancy between the flow rates of the two channels operating together, is, in our hands, in the region of 1%. In their photograph the authors show the donor blood close to the pump. It is common practice, however, for the donor blood to be 'suspended' (authors' word) high above the working area. It seemed possible that the hydrostatic pressure so achieved might introduce an important discrepancy between the flow rates of the two channels. To test this hypothesis, an in vitro system was set up. A Holter pump pumped water through two identical systems into graduated cylinders at 26 'C. The difference in flow rates with afferent hydrostatic pressure variation was measured. The channel track was varied and the increased pressure applied sequentially to both systems (Table). It is apparent that a pressure differential of 52 to 56 cm water increases the channel discrepancy up to six-fold. Though the authors equalize their volumes at 100 cm3 intervals, it appears worth while to approximate the afferent pressures on the pump channels.