Drop-out from cardiovascular magnetic resonance in a randomized controlled trial of ST-elevation myocardial infarction does not cause selection bias on endpoints.

Abstract

The extent of selection bias due to drop-out in clinical trials of ST-elevation myocardial infarction (STEMI) using cardiovascular magnetic resonance (CMR) as surrogate endpoints is unknown. We sought to interrogate the characteristics and prognosis of patients who dropped out before acute CMR assessment compared to CMR-participants in a previously published double-blinded, placebo-controlled all-comer trial with CMR outcome as the primary endpoint. Baseline characteristics and composite endpoint of all-cause mortality, heart failure and re-infarction after 30days and 5years of follow-up were assessed and compared between CMR-drop-outs and CMR-participants using the trial screening log and the Eastern Danish Heart Registry. The drop-out rate from acute CMR was 28% (n = 92). These patients had a significantly worse clinical risk profile upon admission as evaluated by the TIMI-risk score (3.7 (± 2.1) vs 4.0 (± 2.6), p = 0.043) and by left ventricular ejection fraction (43 (± 9) vs. 47 (± 10), p = 0.029). CMR drop-outs had a higher incidence of known hypertension (39% vs. 35%, p = 0.043), known diabetes (14% vs. 7%, p = 0.025), known cardiac disease (11% vs. 3%, p = 0.013) and known renal function disease (5% vs. 0%, p = 0.007). However, the 30-day and 5-years composite endpoint rate was not significantly higher among the CMR drop-out ((HR 1.43 (95%-CI 0.5; 3.97) (p = 0.5)) and (HR 1.31 (95%-CI 0.84; 2.05) (p = 0.24)). CMR-drop-outs had a higher incidence of cardiovascular risk factors at baseline, a worse clinical risk profile upon admission. However, no significant difference was observed in the clinical endpoints between the groups.