Transfusion strategy in hematological intensive care unit: study protocol for a randomized controlled trial

BACKGROUND Packed red blood cell (PRBC) transfusion is required in hematology patients treated with chemotherapy for acute leukemia, autologous (auto) or allogeneic (allo) hematopoietic stem cell transplantation (HSCT). In certain situations like septic shock, hip surgery, coronary disease or gastrointestinal hemorrhage, a restrictive transfusion strategy is associated with a reduction of infection and death. A transfusion strategy using a single PRBC unit has been retrospectively investigated and showed a safe reduction of PRBC consumption and costs. We therefore designed a study to prospectively demonstrate that the transfusion of a single PRBC unit is safe and not inferior to standard care. METHODS The 1versus2 trial is a randomized trial which will determine if a single-unit transfusion policy is not inferior to a double-unit transfusion policy. The primary endpoint is the incidence of severe complication (grade ≥ 3) defined as stroke, transient ischemic attack, acute coronary syndrome, heart failure, elevated troponin level, intensive care unit transfer, death, new pulmonary infiltrates, and transfusion-related infections during hospital stays. The secondary endpoint is the number of PRBC units transfused per patient per hospital stay. Two hundred and thirty patients will be randomized to receive a single unit or double unit every time the hemoglobin level is less than 8 g/dL. All patients admitted for induction remission chemotherapy, auto-HSCT or allo-HSCT in hematology intensive care units will be eligible for inclusion. Sample size calculation has determined that a patient population of 230 will be required to prove that the 1-unit PRBC strategy is non-inferior to the 2-unit PRBC strategy. Hemoglobin threshold for transfusion is below 8 g/dL. Estimated percentage of complication-free hospital stays is 93 %. In a non-inferiority hypothesis, the number of patients to include is 230 with a power of 90 % and an alpha risk of 5 %. TRIAL REGISTRATION 14-128; Clinicaltrials.gov NCT02461264 (registered on 3 June 2015).