Background Cardiac magnetic resonance (CMR) may radiologically identify or confirm underlying pathophysiologies in myocardial infarction with non-obstructive coronary arteries (MINOCA), however, there are scant prospective data evaluating the impact on routine clinical care.

Methods In a multicentre international cohort study of MINOCA, clinical diagnosis, diagnostic certainty and intended clinical management were prospectively determined before and again after CMR. The primary outcome was a composite of change in clinical diagnosis and/or management. Secondary outcomes were individual components of the primary outcome, change in diagnostic certainty and number-needed-to-test for deprescription of dual antiplatelet therapy (DAPT). Predictors of the primary outcome were evaluated by multivariable logistic regression analysis.

Results In 320 patients, CMR was associated with change in diagnosis and/or management in 63% (95% CI 57% to 68%, p0.001) and significantly increased diagnostic certainty (8÷10 post-CMR (5–9) vs 6/10 pre-CMR (4–7), p0.0001). Relevant predictors of the primary outcome on multivariable analysis were early CMR (≤14 days), absence of atheroma on coronary angiography and significant pre-CMR diagnostic uncertainty (≤5/10); CMR changed diagnosis and/or management in 80% of individuals with all three predictors versus 40% in those with none. In individuals where treating physicians initially chose to prescribe DAPT despite no obstructive culprit lesion, number-needed-to-test by CMR for DAPT deprescription was 3.