A modified Essen stroke risk score for predicting recurrent cardiovascular events: development and validation

Abstract

Background: The Essen stroke risk score is widely applied to predict the risk of recurrent ischemic stroke. We developed a modified Essen stroke risk score and validated it using a large prospective Effective Vascular Event REduction after STroke (EVEREST) registry including 3588 patients with ischemic stroke in Japan. Patients with cardioembolic stroke were excluded, and follow-up was one-year.

Methods: The modified Essen stroke risk score was calculated from scores for waist circumference, stroke subtype by etiology, and gender in addition to age, hypertension, diabetes mellitus, previous myocardial infarction, other cardiovascular diseases except myocardial infarction and atrial fibrillation, peripheral artery disease, smoking, and previous stroke or transient ischemic attack. A multiple logistic regression model identified the predictors (each assigned one or two points) and provided c-statistics for the modified Essen stroke risk score. We considered two outcomes, recurrent ischemic stroke and cardiovascular events (defined as the combined outcomes of fatal or nonfatal stroke, myocardial infarction, nonfatal unstable angina, and cardiac death).

Results: Recurrent ischemic stroke occurred in 121 patients (3·7%) and cardiovascular events occurred in 133 (4·0%) within a year. The c-statistic (used for discrimination) was 0·632 for recurrent stroke and 0·640 for cardiovascular events. Patients scoring 6 or greater were classified as high risk, otherwise were classified as low risk. Kaplan-Meier analysis revealed that the modified risk score was more predictive than the Essen stroke risk score in both men and women.

Conclusions: The modified Essen stroke risk score increased the ability of the Essen stroke risk score to predict recurrent cardiovascular events. Patients with a high modified Essen stroke risk score should be candidates for intensified secondary prevention strategies.