Declining stroke and vascular event recurrence rates in secondary prevention trials over the past 50 years and consequences for current trial design

Abstract

Background: It is widely supposed, but not well-demonstrated, that cumulative advances in standard care have reduced recurrent stroke and cardiovascular events in secondary prevention trials.

Methods and results: Systematic search identified all randomized, controlled trials of medical secondary stroke prevention therapies published from 1960 to 2009. Randomized, controlled trials narrowly focused on single stroke mechanisms, including atrial fibrillation, cervical carotid stenosis, and intracranial stenosis, were excluded. From control arms of individual trials, we extracted data for baseline characteristics and annual event rates for recurrent stroke, fatal stroke, and major vascular events and analyzed trends over time. Fifty-nine randomized controlled trials were identified, enrolling 66 157 patients in control arms. Over the 5 decade periods, annual event rates declined, per decade, for recurrent stroke by 0.996% (P=0.001), fatal stroke by 0.282% (P=0.003), and major vascular events by 1.331% (P=0.001). Multiple regression analyses identified increasing antithrombotic use and lower blood pressures as major contributors to the decline in recurrent stroke. For recurrent stroke, annual rates fell from 8.71% in trials launched in the 1960s to 6.10% in the 1970s, 5.41% in the 1980s, 4.04% in the 1990s, and 4.98% in the 2000s. The sample size required for a trial to have adequate power to detect a 20% reduction in recurrent stroke increased 2.2-fold during this period.

Conclusions: Recurrent stroke and vascular event rates have declined substantially over the last 5 decades, with improved blood pressure control and more frequent use of antiplatelet therapy as the leading causes. Considerably larger sample sizes are now needed to demonstrate incremental improvements in medical secondary prevention.

Figures

Figure 1

Trends over time of the event rates of recurrent stroke, fatal stroke, and major vascular events The size of each circle on the graph indicates each trial's weight which was derived by the inverse of variance of each trial's event rate. SE (β), standard error of beta-coefficient; CV, cardiovascular