Restrictive transfusion strategy in cardiac surgery remains noninferior


Clinical question: Does using a restrictive transfusion strategy with patients undergoing cardiac surgery affect long-term outcomes?

Dr. Cassie Shaw

Background: Using a restrictive transfusion strategy in patients undergoing cardiac surgery is known to use fewer units of allogeneic red cells, compared with a liberal strategy, while still having noninferior short-term clinical outcomes. At this time, little is known about such a strategy’s long-term effects.

Study design: Randomized, open-label, noninferiority trial.

Setting: 74 hospitals in 19 countries.

Synopsis: 5,243 adults undergoing nontransplant cardiac surgeries and having at least a moderate predicted risk for death were randomly divided into a liberal or restrictive transfusion strategy. Restrictive-­strategy participants received a transfusion when hemoglobin was less than 7.5 g/dL, compared with either a hemoglobin of 8.5 g/dL on the floor or 9.5 g/dL in the ICU for the liberal-strategy group. During the hospitalization, the restrictive group received fewer U of red cells and had a lower mean predischarge hemoglobin. At 6 months, the groups were compared for the primary outcomes of death, MI, stroke, or renal failure requiring dialysis, finding an occurrence of such in 402/2,317 in the restrictive-strategy group and 402/2,347 in the liberal-strategy group (P = .006 for noninferiority). Limitations include the study being a noninferiority trial and the very specific patient population selected.

Bottom line: In patients undergoing cardiac surgery, a restrictive transfusion strategy is noninferior to a liberal strategy with respect to death from any cause, MI, stroke, and new renal failure requiring dialysis at 6 months postop.

Citation: Mazer CD et al. Six-month outcomes after restrictive or liberal transfusion for cardiac surgery. N Eng J Med. 2018 Sep 27;379(13):1224-33.

Dr. Shaw is an assistant professor in the division of hospital medicine,University of New Mexico.

Next Article:

   Comments ()