Meta-regression of trials using carotid intima–media thickness (CIMT) as a surrogate end point leads to unreliable results owing to heterogeneity in methods, interventions, and outcomes of pooled ...