β-Blocker Discontinuation after Myocardial Infarction without Reduced Ejection Fraction in the Revascularization Era: A Design-Coherent Bayesian Review & Meta-Analysis
TL;DR
In contemporary revascularized patients with preserved or mildly reduced EF, β-blocker discontinuation was not associated with a credible increase in death or MACE, whether stopped in the stable phase or the acute-withdrawal phase; the prior harm likely reflects immortal-time bias and inclusion of non-revascularized patients.
Abstract
Background & Objective: β-Blockers are prescribed indefinitely after myocardial infarction (MI), yet in contemporary revascularized patients with preserved or mildly reduced ejection fraction (EF), their value remains uncertain, and the question of whether an established β-blocker can be electively discontinued is common. Prior syntheses pooled three distinct scenarios (acute withdrawal, immortal-time–susceptible duration comparisons, and elective late discontinuation), obscuring their differences. Methodology: We performed a Bayesian random-effects meta-analysis comparing β-blocker discontinuation versus continuation, restricted to design-coherent studies (shared post-establishment baseline) in contemporary, revascularized patients with EF >40% (or no LV dysfunction), no heart failure, and on a β-blocker for ≥1 month. Duration-based and acute-withdrawal designs were analyzed only in sensitivity analyses. We report pooled hazard ratios with 95% credible intervals, the probability of direction (pd), and the fraction within a region of practical equivalence (ROPE). Results: Seven design-coherent studies (2 randomized trials, 5 cohorts; 88,514 patients) were pooled. No estimate was credible (pd ≥97.5%): all-cause mortality 1.01 (0.83–1.19) and major adverse cardiovascular events (MACE) 1.07 (0.93–1.20) sat closest to practical equivalence, while recurrent MI (1.20) and cardiovascular death (1.13) were inconclusive. Estimates were stable when randomized acute-withdrawal trials were added but inflated with duration-based studies (cardiovascular death 1.64). Conclusion: In contemporary revascularized patients with preserved or mildly reduced EF, β-blocker discontinuation was not associated with a credible increase in death or MACE, whether stopped in the stable phase or the acute-withdrawal phase; the prior harm likely reflects immortal-time bias and inclusion of non-revascularized patients. Discontinuation is a reasonable, individualized option; the recurrent-MI signal warrants confirmation. Key Words: Beta Blocker, Myocardial Infarction, Preserved Ejection Fraction.