my personal edition > stroke > news


  E-Mail this DGReview to a colleague

DGReview

New Model Incorporating Magnetic Resonance Helps Predict Stroke Outcome

A new model incorporating magnetic resonance diffusion-weighted imaging and clinical assessment may more accurately predict the outcome for patients following a stroke.

Dr Alison Baird and colleagues from the United States National Institute of Neurological Disorders And Stroke at the National Institutes of Health, Bethesda, Maryland, aimed to assess whether the extent of ischaemic brain injury on magnetic resonance diffusion-weighted imaging (MR DWI) could provide additional prognostic information to clinical factors.

"Doctors have been searching for decades for a technique that will allow early and accurate assessment of prognosis after ischaemic stroke," the investigators point out. It is established that by six months after a stroke, about 20 percent to 30 percent of patients have died, 20 to 30 percent are moderately to severely disabled, 20 to 25 percent have mild to moderate disability, and the remainder are without deficit.

"However, it is often not clear for some days after stroke onset how patients are likely to fare. Patients with initially similar clinical deficits can improve dramatically or worsen during the first 48-72 hours," they add. " Changes in clinical status might sometimes be related to pathophysiologic events, such as early reperfusion, hemorrhagic transformation or edema of the ischaemic lesion, and prognoses can markedly differ between experienced physicians."

There also is the probability that new but possibly risky stroke therapies will need to be administered within the first hours after stroke and that decisions based upon the relative risks and benefits of treatment could be aided by knowing the likely outlook of the patient.

The researchers studied 66 patients from a North American teaching hospital who had: MR DWI within 36 hours of stroke onset; the National Institutes of Health Stroke Scale (NIHSS) score measured at the time of scanning; and the Barthel Index measured no later than three months after stroke.

Using statistical assessment to derive a predictive model for good recovery; this model was applied to an independent series of 63 patients from an Australian teaching hospital.

Investigators developed a three-item scale "low, medium or high--for the early prediction of stroke recovery. Combined measurements of the NIHSS score, time in hours from stroke onset to MR DWI, and the volume of ischaemic brain tissue on MR DWI gave the best prediction of stroke recovery and the model was externally validated on the Australian sample.

This combination of clinical and MR DWI factors for patients shortly after admission to hospital provided better prediction of stroke recovery than any factor alone, the investigators said.

"Potential uses for our results include early decision-making on aggressiveness of care, and discharge planning and rehabilitation, which is of particular relevance in the current health-care climate where there are extreme pressures to shorten length of stay, limit the number of tests, and commence discharge planning soon after the patient's admission," Dr Baird and colleagues point out.

Early computed tomography signs can be difficult to detect, the investigators said. However, "if these results are confirmed we could incorporate computed tomography measurements in an attempt to make the prediction rule much more widely available for clinical use."
Lancet 2001; 357: 2095-99

E-Mail this DGReview to a colleague

To print, use this version