MINT Bayesian methods paper by Marnie Bertolet published JAMA Network Open

A secondary analysis of the Myocardial Ischemia and Transfusion (MINT) trial using Bayesian methods was published today in JAMA Network Open.  The paper was led by Dr. Marnie Bertolet, a faculty member at the Epidemiology Data Center, and included colleagues Drs. Brandon Herbert and Maria Mori Brooks as co-authors.  Key points for the article are included below:

Question
What is the posterior probability that a liberal vs restrictive transfusion strategy is associated with reduced risk of 30-day death or myocardial infarction (MI)?

Findings
This post hoc analysis of 3504 participants in the MINT randomized clinical trial found that, for the liberal vs restrictive strategy, the posterior probability of at least 1 less death or MI per 100 treated was 62.7% to 90.4%. The probability of at least 1 more heart failure event per 100 treated was 13.3% to 29.3%.

Meaning
This study suggests that, consistent with guideline recommendations, a liberal transfusion strategy may be reasonable for some patients, as its net benefit, after treating 200 patients, is 5 fewer deaths or MIs and 1 more heart failure event. 

https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2852120