RESUMO
OBJECTIVES: Haemodilution during cardiopulmonary bypass is associated with increased perioperative blood transfusions and is thought to reduce intraoperative oxygen delivery to the brain. We sought to evaluate our method of rapid antegrade prime displacement in the context of the perioperative blood transfusion rate, intraoperative cerebral saturations and postoperative hospital stay. METHODS: Retrospective analysis of 160 propensity-matched patients undergoing elective coronary artery bypass grafting was performed comparing different perfusion strategies on perioperative blood transfusion and length of postoperative stay. Eighty patients who had rapid antegrade prime displacement and vacuum-assisted venous drainage (RAD-VAD) were compared with 80 patients who had conventional cardiopulmonary bypass with gravity drainage (CB). RAD-VAD involved displacing all or most of the prime in the circuit with the patient's own blood prior to the initiation of cardiopulmonary bypass within a 15-20 s window. Within each group, 10 patients had intraoperative cerebral saturation measurements. RESULTS: There were no differences in the baseline characteristics between the groups. Both groups had a significant fall (P < 0.05) in haematocrit during cardiopulmonary bypass from preoperative values, however, the fall in haematocrit was significantly less in the RAD-VAD group (P < 0.05). There was significantly (P < 0.05) less intraoperative and postoperative homologous blood transfusions in the RAD-VAD group (47.892 ml ± 8.14 and 76.58 ml ± 21.58) compared with the CB group (229.06 ml ± 105.03 and 199.91 ml ± 47.13). There was a significant fall in cerebral saturations within both groups (P < 0.05) but it was not significant between the groups. The postoperative stay was significantly (P < 0.05) shorter in the RAD-VAD group compared with the conventional group (7.74 days ± 0.51 vs 10.13 days ± 0.95). CONCLUSIONS: RAD-VAD is associated with a significantly lower blood transfusion rate perioperatively and shorter hospital stays compared with CB.