After cardiac surgery, using a restrictive transfusion threshold – forgoing transfusion until hemoglobin level drops to 7.5 g/dL – does not decrease morbidity or costs of care, compared with using a liberal transfusion threshold of 9 g/dL, according to a report published online March 12 in the New England Journal of Medicine.

Several blood management guidelines and health policy statements recommend the restrictive approach in the hope that it will reduce the increasing demand on blood services and the high costs of storing, handling, and administering red-cell units, and also because transfusions following cardiac surgery have been linked to infection, low cardiac output, acute kidney injury, and increased mortality. Clinicians remain uncertain about a safe threshold for transfusions in this setting, which is evidenced by the striking variation in transfusion rates among cardiac centers in the United States (8%-93%) and the United Kingdom (25%-75%), said Dr. Gavin J. Murphy of the British Heart Foundation and department of cardiovascular sciences, University of Leicester (England), and his associates.

They performed the Transfusion Indication Threshold Reduction (TITRe2) study to test the hypothesis that the restrictive approach is superior to the liberal approach regarding both postoperative morbidity and health care costs. Adults undergoing nonemergency cardiac surgery at 17 specialty centers in the United Kingdom were randomly assigned to a restricted (1,000 patients) or a liberal (1,003 patients) transfusion threshold. The median patient age was 70 years, and 68% were men. Most of the procedures were CABG or valve surgeries.

Contrary to expectations, the primary outcome – a composite of serious infection or an ischemic event such as stroke, MI, gut infarction, or acute kidney injury within 3 months – occurred in 35.1% of patients in the restrictive-threshold group and 33.0% in the liberal-threshold group. Secondary outcomes, including length of ICU stay and rates of clinically significant pulmonary complications, also were similar between the two study groups. Rates of other serious postoperative complications were similar, at 35.7% and 34.2%, as was general health status as assessed via the EuroQol Group 5-Dimension Self-Report Questionnaire, further contradicting the study hypothesis.

Mean health care costs were similar between the two study groups: the equivalent of $17,762 U.S. dollars with restrictive-threshhold transfusions and $18,059 with liberal-threshold transfusions, Dr. Murphy and his associates noted (N. Engl. J. Med. 2015 March 12 [doi:10.1056/NEJMoa1403612]).

Unexpectedly, 3-month mortality was significantly higher with restrictive- than with liberal-threshold transfusions (4.2% vs 2.6%). This association persisted in sensitivity analyses and “is a cause for concern,” but it may be due to chance alone, the investigators added.