Objective The Essen Stroke Risk Score (ESRS) is widely used to predict the risk of recurrent ischemic stroke but may not have sufficient discriminatory ability for predicting early recurrence. In this study, we developed a modified ESRS for predicting recurrent ischemic stroke at one year based on Chinese population. Methods Eight hundred and three consecutive patients were included in the study. We performed a univariate analysis to investigate the effect of previously identified risk factors on recurrent stroke at one year. The candidate predictors were included in a forward stepwise logistic regression model. Prediction scores were assigned based on β coefficient and a modified ESRS system was constructed. We used receiver operating characteristic (ROC) curves to compare the predictive ability of the modified and original ESRS. Results Recurrent ischemic stroke occurred in 85 (11.00%) patients and death occurred in 68 (8.80%) patients within one year. Hypertension over 15 years, diabetes over 10 years, large artery atherosclerosis cerebral infarction by TOAST criteria and a history of transient ischemic attack (TIA) or ischemic stroke (IS) were identified as independent risk factors for recurrent stroke (all P < 0.05). The areas under the ROC curves of ESRS and modified ESRS were 0.58 (P < 0.05, 95%CI: 0.51-0.64) and 0.70 (P < 0.05, 95% CI: 0.63-0.76), respectively. The modified ESRS appears to be superior to the original ESRS for predicting recurrent ischemic stroke at one year (P < 0.05). Conclusions The modified ESRS increased the ability of the original ESRS to predict recurrent stroke at one year among the Chinese population.
Keywords: Essen Stroke Risk Score (ESRS); Ischemic stroke; predictive ability; recurrent stroke; risk.