A multicenter randomized controlled trial to assess the feasibility of testing modified ultrafiltration as a blood conservation technology in cardiac surgery

ObjectivesModified ultrafiltration is a technique after cardiopulmonary bypass whereby blood withdrawn from the aortic cannula is passed across a semipermeable membrane to hemoconcentrate. Unblinded trials have suggested that modified ultrafiltration is efficacious for blood conservation. The objective of this trial was to assess the feasibility of a model testing modified ultrafiltration in which all members of the surgical team were blinded to the intervention.MethodsPatients (<65 kg) undergoing procedures involving cardiopulmonary bypass were randomized to undergo either modified ultrafiltration (n = 29) or sham (circulation without an interposed filter, n = 36) for 15 minutes. The circuit was shielded from all members of the team except the perfusionist. A questionnaire was administered to determine the blinding success.ResultsModified ultrafiltration resulted in a removal of 1000 ± 251 mL of fluid and a reduction in the pump balance (1025 ± 807 vs 1804 ± 838; P < .001) with an increase in hemoglobin immediately after intervention (increase of 7.7 ± 8.8 g/L in modified ultrafiltration vs 3.8 ± 5.1 g/L in sham; P = .04). Introduction or increase in dose of vasopressors was more frequent in the modified ultrafiltration group (52% vs 28%; P = .048). Differences in red cell transfusion rates between groups did not reach statistical significance (P = .59). Blinding was successful for the anesthetist (blinding index 0.13 [95% confidence interval, 0.11–0.38] and the intensivist (blinding index, 0.09 [95% confidence interval, 0.14–0.31]) but not for the surgeon (blinding index, 0.24 [95% confidence interval, 0.05–0.42]). The compliance rate for the transfusion protocol was greater than 90%.ConclusionsModified ultrafiltration was effective for hemoconcentration after cardiopulmonary bypass in patients of low body weight, but it is associated with an increased need for vasopressor support. The anesthetist and intensivist were successfully blinded to the intervention.