Autotransfusion in trauma: a comparison of two systems

Autotransfusion is a potentially valuable tool in the resuscitation of hypovolemic trauma patients; its acceptance in this setting has been limited by fears of the induction of coagulopathic and septic complications. It has been inferred that the addition of a cell washing step would obviate these concerns but at the cost of speed. To assess the validity of these concerns, we have retrospectively compared two autotransfusion devices: one without (the modified Bentley device) and one with (the Baylor Rapid Autologous Transfusion system) a cell washing step, over a 48-month period. In the Bentley group (n = 13), the mean estimated blood loss was 8,423 ml and the mean amount of blood autotransfused was 1,826 ml. Overall, the device returned 0.54 units of whole blood for every unit of banked blood used. Sixty-two per cent of these severely injured individuals died. Among survivors, there was a 20 per cent incidence of significant complications. In the BRAT group (n = 13), the mean estimated blood loss was 11,177 ml and the mean amount of blood autotransfused was 3,681 ml. Overall, the device returned 0.82 units of washed, packed red blood cells for every unit of banked blood used. Overall mortality was 26 per cent, and 30 per cent of survivors had complications. While we have been unable to demonstrate an advantage of the cell washing step, there is no evidence that this step in this unit limited the rate or volume of autologous blood replacement.