Bayesian Post Hoc Analysis of MINT Trial Favors Liberal Transfusion Strategy After Myocardial Infarction

A new Bayesian post hoc analysis of the Myocardial Ischemia and Transfusion (MINT) trial suggests that a liberal transfusion strategy may be superior to a restrictive one for patients with myocardial infarction (MI) and anemia. The findings were published Monday in JAMA Network Open.

The original MINT trial compared a restrictive strategy, in which patients were transfused when hemoglobin fell below 7 to 8 g/dL, with a liberal strategy that maintained hemoglobin above 10 g/dL. The primary outcome (composite of death or recurrent MI within 30 days) favored the liberal strategy but was not statistically significant (risk ratio 1.15; P = .07). The researchers noted that this borderline result has been difficult to interpret and reanalyzed the trial data using a Bayesian framework to better quantify the probability of benefit associated with a liberal transfusion strategy.

Using three prior assumptions (one neutral, one favoring the liberal strategy and one favoring the restrictive strategy), the researchers found that the probability that the liberal strategy lowered the risk of 30-day death or MI ranged from 89.1% to 98.8%. The probability of preventing at least one death or MI per 100 patients treated ranged from 62.7% to 90.4%. The liberal strategy carried a higher probability of heart failure, though the absolute excess was small, and the chance that this excess exceeded one event per 100 patients was 29.3% or lower.

Assuming a neutral prior, the authors estimated that adopting a liberal strategy would prevent roughly five deaths or MIs while causing about one additional heart failure event for every 200 patients treated. The authors concluded that the reduced risk of death or MI with a liberal strategy outweighed the increased risk of heart failure. They also noted that the findings are consistent with AABB's international clinical practice guidelines, which recommended that a liberal strategy may be reasonable for patients with MI and anemia, according to patient values and clinician risk assessment.