Carotid surgery.

Carotid endarterectomy may well be the most studied operation ever. Unlike many if not most surgical interventions, there are several randomised trials, and now meta-analyses, to inform clinical practice. On balance, accepting the small but definite risk of stroke as a result, surgery-is a reasonable option for many patients with severe and recently symptomatic carotid stenosis who have such a high risk of stroke without surgery. Although operating on asymptomatic stenosis is safer, the unoperated risk of stroke is--on average--so low that surgery is generally not worthwhile. The challenge now is to make surgery even safer, and perhaps angioplasty will be safer still and as durable although so far there is no good randomised evidence. At the same time we must identify, if we can, those few patients with severe symptomatic stenosis who will have a stroke and focus surgery just on them, and not offer it to all the patients with severe symptomatic stenosis who might have a stroke. This will require the development and validation of mathematical models of risk. But, however well focused surgery becomes and so however small we can make "the numbers-needed-to-treat", it will have a negligible impact on stroke incidence because so few strokes are preceded by transient ischaemic attacks in patients with severe carotid stenosis.