OBJECTIVE Anemia is common in severely ill patients, and blood sampling plays a relevant causative role. Consequently, blood transfusions are frequent an related to several complications. Trying to reduce the transfusion-related risk, minimizing blood loss is mandatory. Thus, this work aimed to evaluate a closed blood sampling system as a strategy to spare unnecessary blood losses and transfusions. METHODS This
was a prospective, randomized, controlled, multicenter, 6 months, clinical trial. The patients were assigned to either VAMP (Venous Arterial Blood Management Protection) group, using a closed blood sampling system, or control group. The groups' transfusion rate, as well as hemoglobin (Hb) and Hematocrit (Ht) changes were compared for 14 days. RESULTS Were included 127 patients, 65 assigned to the control group, and 62 to VAMP. During the intensive care unit stay, both groups experienced both hemoglobin and hematocrit drops. However, when the final Ht and Hb were compared between the groups, a difference was identified with higher values in the VAMP group (p=0.03; p=0.006, respectively). No statistical difference was found for both groups transfusion rates, although the VAMP group had an absolute 12% blood transfusion reduction. CONCLUSION The use of a closed blood sampling system was able to minimize blood count values changes, however failed to reduce transfusions rate.