[Stroke--possibilities for prevention]

The practice of stroke medicine varies enormously. This is not just because of variable resources, access to services and preferences among patients, clinicians, medical centres, governments and societies, but also often because of poor evidence about the effectiveness and cost of many components of stroke care. Although the past 20 years have seen a huge increase in randomised controlled trials, the results have not always had much influence on clinical practice, perhaps because many trials used confusing outcome measures or were too small (limited statistical power) or because the findings could not easily be generalised. Our purpose is to review the best evidence for stroke treatment and prevention.