New Tool Allows Early Prediction Of Patient's Stroke Outcome

Scientists have developed a new tool that may help physicians predict, during the first several hours a stroke patient is in the hospital, the degree of recovery the patient will eventually experience. The tool uses three factors for the accurate prediction of stroke outcome: measurement of brain injury using magnetic resonance imaging (MRI); the patient's score on the NIH stroke scale; and the time in hours from the onset of symptoms until the MRI brain scan is performed.

Alison E. Baird, M.D., Ph.D., of the National Institute of Neurological Disorders and Stroke (NINDS), and colleagues will publish their findings in the June 30, 2001 issue of "Lancet" (1). "We hope this new tool will not only help physicians manage their patients more efficiently, but also will alleviate the distress and anxiety about prognosis that patients and their families suffer in the first days after stroke," said Dr. Baird.

The researchers set out to see if a new type of brain imaging technology called magnetic resonance diffusion- weighted imaging (MR-DWI), in addition to standard clinical assessments, could yield prognostic information about a stroke. The MR-DWI can measure the volume of the lesions that appear during the first few hours after an ischemic stroke, which is caused by a clot obstructing blood flow to the brain. The study results showed that this measurement correlates with the severity of the stroke as well as the patient's outcome. Patients with small lesion volumes (less than 14.1 milliliters) were five times more likely to recover from their strokes than patients with larger lesion volumes.

Another important prognostic tool that has been used widely for many years is the National Institutes of Health Stroke Scale (NIHSS), used to measure the severity of neurological dysfunction at the time of a stroke. In the study reported in "Lancet", the Stroke Scale was measured within 1 hour of the MRI scan. A score greater than 25 indicates very severe neurological impairment; a score between 15 and 25 indicates severe impairment, while a score between 5 and 15 mild to moderately severe impairment, and a score less than 5 mild impairment. The mean score of the patients in this study was 11.

The third measurement in the scale is the time from the onset of the patient's symptoms to the time of the brain scan. If a patient suffered the stroke while asleep, the time was backdated to the last time the patient was known to have no stroke symptoms. Surprisingly, the patients who waited the longest before receiving their scans were more likely to recover. The investigators speculate that this time relationship may reflect an "instability" factor in the earliest hours of stroke. There may be more certainty of a good outcome when the patient is assessed beyond the first 6 hours, by which time the most critical changes in blood flow in the brain have occurred.

The NINDS study involved looking at the data from a total of 129 stroke patients -- 66 at the Beth Israel Deaconess Medical Center in Boston and 63 at the Royal Melbourne Hospital in Australia -- and then developing a three-item scale for early prediction of stroke recovery. (Good stroke recovery in this study was defined as a score of greater than 90 on the Barthel Index, indicating a patient who has nearly full functional independence.)

Using the three-item scale the researchers assigned points (with a maximum score of 7) based on the brain lesion volume, the patient's score on the NIHSS, and the time from symptoms to scanning. The clinicians rated a patient's likelihood of recovery using categories of low (total score 0-2), medium (total score 3-4), or high (total score 5-7).

The new scale proved to be a very accurate predictor of stroke recovery with high sensitivity and specificity. In the Boston group, only 6 percent of the patients who had a low score (0-2) recovered. Forty-seven percent of the patients with a medium range score recovered, and 93 percent of those with a high score recovered. In the Australian group, 8 percent of patients with a low score recovered, 57 percent of patients with a medium score recovered, and 78 percent of patients with a high score recovered.

The researchers concluded that the combination of clinical and imaging data allowed more reliable early prediction of stroke recovery than any single factor alone.

Future studies will focus on evaluating the potential of CT scanning in the prediction of recovery from stroke. CT scanning is the standard imaging tool for stroke patients and, unlike MRI, is available in nearly every hospital. "This would allow us to make a prognostic stroke scale more widely available for clinical use," said Dr. Baird. Dr. Baird is a member of the NIH Stroke Team at Suburban Hospital in Bethesda, Maryland. The NIH Stroke Team is headed by Steven Warach, M.D., Ph.D., of NINDS, who is senior author on the "Lancet" paper.

NINDS is a component of the National Institutes of Health in Bethesda, Maryland, and is the leading supporter of biomedical research on the brain and nervous system. The Institute is celebrating its 50th anniversary this year.

This release will be posted on EurekAlert! at http://www.eurekalert.org  and on the NINDS website at http://www.ninds.nih.gov/news_and_events/index.htm.

(1) Baird, A.E., Dambrosia, J., Janket, S., Eichbaum, Q., Chaves, C., Silver, B., Barber, P.A., Parsons, M., Darby, D., Davis, S., Caplan, L.R., Edelman, R.R., Warach, S. "A Three-Item Scale for the Early Prediction of Stroke Recovery." "Lancet", June 30, 2001, Vol. 357, No. 9274, pp. 2095-2099.