A model shows the brain vessels involved in stroke. (CBC)

A new web-based tool could help doctors better predict those more likely to suffer a second stroke.

Studies suggest that people who have a second stroke soon after a first stroke are more likely to die or have a severe disability.

"We currently don't have a well-developed tool for predicting short-term risk of early recurrent stroke, so this tool could help improve stroke care and outcome," study author Dr. Hakan Ay of Massachusetts General Hospital in Boston said in a release.

"For example, people at high risk of a second stroke can be immediately admitted to specialized stroke centres and given preventative treatment."

The new tool, described in Wednesday's online issue of Neurology, calculates a person's risk of having another stroke within three months by looking at risk factors such as:

• History of mini-stroke known as transient ischemic attack or TIA.
• Age.
• Type of first stroke.
• Brain scan information.

In the study, researchers looked at information from 1,458 people who had an ischemic stroke, which occurs when a clot blocks a blood vessel in the brain and cuts off circulation.

Long-term predictors of stroke — such as smoking, diabetes and high blood pressure — did not predict short-term strokes, while the type of stroke did, the team found.

After a three-month followup involving 806 of the participants, 60 strokes had occurred. The risk of a second stroke was 2.6 per cent at 14 days and six per cent at 90 days, the researchers found.

The higher a patient's score on the tool, known as the Recurrence Risk Estimator at 90 days, the more likely it was they would experience a second stroke.

Risk factors

The 90-day risk was about 40 times higher in people with four or more stroke risk factors than in people without any risk factor. More than 96 per cent of patients who had a second stroke showed signs of at least one risk factor.

Before doctors can start using the tool, its accuracy needs to be confirmed, the study's authors and a journal editorial noted.

As with other prediction scores, the tool will ultimately be judged on its clinical impact and how it improves patient care, Dr. Tobias Kurth of Université Pierre et Marie Curie in Paris and Dr. Christian Stapf of the department of neurology at Hôpital Lariboisière in Paris said in their editorial.

"A refined forecast may guide stroke physicians to the good weather zones, allow early identification of high-risk patients, lead to adjusted treatment plans … and help improve stroke recurrence in specific patient subgroups at risk," they wrote.

The availability of the online tool may pave the way for it to be validated and retested, the pair noted.

But predicting another type of stroke is more relevant, said Dr. David Spence, director of the Stroke Prevention & Atherosclerosis Research Centre at Robarts Research Institute in London, Ont.

Patients with TIA are less likely to be admitted than those who've had strokes, Spense noted. Stroke patients already tend to be investigated intensively and start stroke prevention therapies aimed at the underyling cause.

Another type of score, known as ABCD2, helps doctors to decide which patients can be sent home from the emergency department with an appointment in the stroke prevention clinic, and which ones should be investigated and treated right away by the neurology team. The ABCD2 score also has its flaws, Spence said.