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.

Systemic effects of intraoperative autotransfusion.

The hematological effects of autotransfusion have not been well characterized. A canine model was used in which a controlled 2000-ml blood loss was effected into the open peritoneum and replaced either by autotransfusion, using one of two devices, or by homologous transfusion. In each group hematological data were collected before, during, and after the operation. None of the parameters measured showed significant differences between the two autotransfused groups. However, intraoperative values for prothrombin time and activated partial thromboplastin time were significantly prolonged in the autotransfused groups compared with the homologous transfusion group. Fibrinogen levels and platelet counts fell significantly from baseline levels in the intraoperative period, but rose above even baseline levels postoperatively. White blood cell counts remained unchanged intraoperatively, but rose significantly and remained elevated at 24 hours and 1 week. No significant differences exist between the effects produced by the two autotransfusion systems. It is clear that prothrombin time and activated partial thromboplastin time are prolonged intraoperatively in the autotransfusion groups, whereas these parameters are unaffected by homologous transfusion.

Clotting competence of intracavitary blood in trauma victims.

In order to assess more rationally the requirement for anticoagulation during intraoperative autotransfusion, the clotting competence of blood collected from the body cavities of 31 trauma victims entering our emergency department with indications for intraoperative transfusion was assessed. Blood was collected at thoracotomy or laparotomy prior to the institution of any anticoagulant measures and was assessed for clotting competence, the presence of fibrinogen, the presence of soluble fibrin monomere, and the appearance of fibrin degradation products. The prothrombin time, partial thromboplastin time, and thrombin time of this blood were markedly elevated; fibrinogen was absent; soluble fibrin monomer was absent; and fibrin degradation products were markedly elevated. Blood collected from body cavities is then incoagulable, and we suggest that in the autotransfusion of such a product the need for anticoagulation may be reduced.