When Randomization Is Not Enough: Divergent Beta-Blocker Trials and the Illusion of Universal Causality

Two contemporary randomized trials of β-blocker therapy after myocardial infarction (MI) reached apparently conflicting conclusions. REBOOT, conducted in Spain and Italy, found no reduction in death or major cardiovascular events among patients with preserved ejection fraction. In contrast, the BETAMI-DANBLOCK study, conducted in Scandinavia, reported a modest but statistically significant reduction in a composite endpoint. The prevailing response has been to reconcile this divergence by elevating left ventricular ejection fraction-particularly the 40-49% range-as the decisive explanatory variable. This interpretation offers analytic simplicity but rests on fragile statistical and conceptual grounds. Drawing on Nancy Cartwright's account of causal capacities and Judea Pearl's transportability framework, this paper argues that β-blockers exhibit stable physiological effects whose clinical benefits depend on the causal environment. Randomization secures internal validity within trials but cannot homogenize genetic, systemic, or environmental contexts. Meta-analysis, by averaging across heterogeneous environments and individual treatment effects, risks obscuring rather than clarifying dependencies. The divergence between REBOOT and BETAMI-DANBLOCK does not represent a failure of evidence-based medicine but exposes the limits of reductionist interpretation. Rather than seeking universal truths, cardiovascular medicine should map the causal environments in which therapeutic capacities become manifest.