New Score to Predict Early Stroke Death

January 11, 2019

A new clinical score has been developed to identify patients with a high risk of early mortality after an ischemic stroke.

Previous scores for stroke mortality have been developed, but these have generally looked at late deaths, the researchers say.

"There isn't much available for risk evaluation for early mortality after a stroke," Thomas Gattringer, MD, Medical University of Graz, Austria, who was involved in developing the score, told Medscape Medical News. "Our score is focused on outcomes in the first few days after an ischemic stroke for patients treated in a dedicated stroke unit."

The aim is to help triage patients in stroke units that are becoming more common and where stroke unit beds are limited, he said. "If we know a patient is at high risk of early death we can recognize the need for more intensive monitoring, and this also helps with communication with the family," Gattringer said.

"A strength of this score is that most of the variables used are very easily obtained on a clinical and neurological examination and from the medical history, and the score is easy to calculate," he added. "Other scores have up to 240 points and need a calculator to work out, but ours only has 12 points and can be easily formulated at the bedside in a few seconds."

Gattringer and colleagues describe how they developed the score in a paper published online in Stroke on December 14.

They used data on 77,653 ischemic stroke patients from the Austrian national stroke unit registry, who were treated between 2006 and 2017. They analyzed a comprehensive list of variables on these patients and compared characteristics in patients who died within the first 7 days after the stroke to those who survived. Multivariate analysis was then performed to ascertain which factors were increasingly associated with early stroke death.

The mortality rate at the stroke units was 2% and median time of death was 3 days.

The researchers found key variables associated with early mortality: age, stroke severity measured by the National Institutes of Health Stroke Scale (NIHSS), pre-stroke functional disability (modified Rankin Scale > 0), pre-existing heart disease, diabetes mellitus, posterior circulation stroke syndrome, and non-lacunar stroke cause.

Points were allocated to each of these variables depending how strongly they were linked to early mortality (Table 1).

Table 1. Risk Prediction Score for Early Ischemic Stroke Mortality (PREMISE)

Risk Factor



60-69 years



> 70 years


Pre-existing disability (mRS 1-5)










> 24


Diabetes mellitus



Heart disease



Posterior circulation stroke syndrome



Non-lacunar stroke



Maximum score points


Results showed that patients with a score ≥ 10 had a 35% risk of dying within the first few days at the stroke unit.

The score had an area under the curve of 0.879. Gattringer explained: "This means the score could explain almost 88% of early stroke deaths. That is a very high number. In general, anything over 0.75 is good. Considering only 2% of the patients in our registry died, that is a very high discriminatory power."

They also validated the score using two different methods. The first validation, known as bootstrapping, is one in which artificial samples are generated from the original data and re-analyzed multiple times. The second method had a temporal design in which the score was applied to patients from the same registry at a later time period. The area under the curve for the validation sample was 0.884.

Gattringer pointed out that the most important, most weighted variable was the NIHSS stroke score. "Patents with highest stroke severity on this scale scored 5 points — that's almost half the points for a maximum score," he noted.

But he cautioned that this may change in the future in the era of thrombectomy because the NIHSS score can rapidly decline after this procedure and because researchers used the at-admission NIHSS score for their calculations. "But this only applies to the relatively low percentage of patients who undergo thrombectomy, and in future there might be a refinement where the NIHSS score after thrombectomy is used for these patients," he suggested.

Another strength of the new clinical score is that it is based on real-word clinical practice. "We included every patient admitted to the Austrian Stroke Unit network in the given time period," Gattringer noted.

However, at present the score is only applicable to patients treated in a stroke unit. Gattringer elaborated: "We have not yet tested the score in patients treated in other settings, but the factors we identified for the score are appropriate for general ischemic stroke patients and it would probably translate to other settings as well."

The researchers' next goal is to prospectively test the score in the Austrian Stroke Unit network. "This will give more information about the clinical practicability and how it is influencing treatment decisions," Gattringer said.

Research funding for this project was provided by the Austrian Neurological Society. Gattringer has disclosed no relevant financial relationships.

Stroke. Published online December 14, 2018. Abstract

For more Medscape Neurology news, join us on  Facebook  and  Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.