Estimating a two-year recurrence risk of ischemic stroke based on Essen Stroke Risk Score and retinal characteristics

Abstract

Objectives: Prediction of recurrence risk after initial ischemic stroke helps improve patient outcomes. We estimated the validity of combining the Essen Stroke Risk Score (ESRS) with retinal characteristics to predict the two-year recurrent risk of ischemic stroke. Methods: A prospective cohort study was conducted at Shenzhen Traditional Chinese Medicine Hospital. Patients with initial ischemic stroke were recruited. ESRS was measured, and fundus photographs were taken by trained physicians. All patients were followed up for 2 years to determine the outcome of recurrent stroke. Logistic models were built using ESRS alone and combined with ESRS and retinal characteristics. The prognostic value of recurrence risk was evaluated by calculating the area under the receiver operating characteristic curve. Results: The total recurrence rate of ischemic events within 2 years was 20%. The risk of recurrence was significantly higher in patients with an ESRS score of >2. Regarding retinal characteristics, patients with recurrent stroke had a lower mean asymmetry index of venules (0.76 vs. 0.77, P=0.00) and bifurcation coefficient of venules (1.29 vs. 1.28, P=0.03) but a higher mean occlusion of arterioles (0.11 vs. 0.13, P=0.01). The area under the curve was 0.5985 based on the logistic regression model with ESRS alone to 0.7294 for the model with both ESRS and retinal characteristics. This study was registered in the Chinese Clinical Trial Registry (registration no. ChiCTR1800019648). Conclusion: A model combining ESRS with retinal characteristics can accurately predict the risk of recurrent stroke.