ObjectiveTo examine the hemodynamic effects and the oxygen transport pattern of autotransfusion of unprocessed blood on hemodynamics and oxygen transportation.DesignProspective, observational study.SettingResearch laboratory of a university medical center.SubjectsSix healthy, domestic pigs (20 to 33 kg).InterventionsA left thoracotomy was performed and a 5-mm incision was created in the descending aorta, resulting in a controlled hemorrhage of 30 mL/kg (approximate 40% of blood volume) into the thoracic cavity over a 45-min time period. During that period, mean arterial pressure (MAP) was maintained slightly more than 50 mm Hg, using intravenous lactated Ringers' solution. The blood sample was collected from the open thorax with compresses soaked in citric acid solution and then extracted by manual squeezing, filtered through several layers of gauzes, and stored in glass bottles. Repeat measurements were performed after hemorrhage, after retransfusion of the harvested blood, and thereafter every 15 mins up to 60 mins. The animals were supported for 2 more hrs and were observed for the following 48 hrs.Measurements and Main ResultsAll animals survived and were in good condition 48 hrs after the experimental hemorrhage. The circulatory and oxygen transport response at the end of hemorrhage and concomitant maintenance of blood pressure at more than 50 mm Hg resulted in: significant reductions of cardiac index, MAP, and oxygen transport (overdot DO2) (46%, 50%, and 64% reductions, respectively, p less than .01), in an increase of heart rate (HR) (plus 21%, not significant), pulmonary vascular resistance index (plus 112%, p less than .05), and oxygen extraction (plus 105%, p less than .01), as well as in a nonsignificant decrease of systemic vascular resistance index (minus 8%).After autotransfusion, the basic hemodynamic variables, MAP and HR were corrected, remaining near baseline (not significant) afterward. Cardiac index and overdot DO sub 2 increased after autotransfusion, but remained below the baseline until the end of the study protocol (p less than .05). A significant increase was noticed for pulmonary arterial pressure and pulmonary vascular resistance index immediately after autotransfusion (p less than .01). These values were corrected in part after 15 to 30 mins, but remained higher throughout the observation period compared with baseline (29.5% and 89.8%, respectively, p less than .05).The recently introduced relationship between cardiac index and oxygen extraction has been proposed to avoid problems of mathematical coupling between oxygen consumption and overdot DO sub 2 measurements. This relationship followed a similar course in all experiments throughout each phase. A shift downward and to the right represented the endpoint of the hemorrhagic phase. After autotransfusion, a shift toward baseline was noticed.Prothrombin time and partial thromboplastin time remained unchanged after autotransfusion. Free hemoglobin concentrations increased immediately after autotransfusion (plus 33%, p less than .05), but returned to baseline values 48 hrs later. Histologic examination showed no changes in the examined organs.ConclusionsReinfusion of large amounts of unprocessed blood (up to 40% of blood volume), collected with compresses from a noncontaminated surgical field is a cheap method, which may be of potential benefit in trauma patients, when more sophisticated autotransfusion devices are lacking. In the present study, this method resulted in transient but significant hemodynamic changes in the pulmonary circulation. Impairment of oxygen transport was noticed after the end of hemorrhage, but this impairment cannot be attributed to the autotransfusion technique alone, but also to factors such as hemorrhagic shock, surgical trauma, etc.(Crit Care Med 1996; 24:855-861)