Validation Of The Pooled Cohort Risk Equation For Predicting Recurrent Stroke And Prognosis In A Chinese Population Of Non-atrial Fibrillation Ischemic Stroke And TIA Patients

Objective Pooled Cohort Risk equation(PCRE),estimates the 10-year risk for the first hard atherosclerotic cardiovascular disease(ASCVD)including stroke,for ASCVD-free adults.We explored its value in predicting stroke recurrence and unfavorable functional outcome in a Chinese ischemic stroke(IS)and transient ischemic attack(TIA)population in China National Stroke Registry(CNSR).Method and Material Patients with IS and TIA were prospectively registered in CNSR.We excluded patients with history of atrial flutter and atrial fibrillation and missing data,leaving eligible patients for validation.The primary endpoint was recurrent stroke,and the secondary endpoints were poor clinical outcome and all-cause death.As for continuous variables,we used t-test;and as for classified variables,we used x2 test.Calculate the endpoint rates after stratify each prediction model.The area under the receiver operating curve(AUC)and 95%CI was used to evaluate the discrimination ability.All tests were 2-tailed,and P<0.05 was considered statistically significant.All analyses were conducted using SAS 9.1.Results 1.Validation of the Pooled Cohort Risk Equation for predicting 3-and 12-month recurrent stroke in non-atrial fibrillation ischemic stroke and TIA patients.Areas under the curve for 3-and 12-month recurrent stroke estimated by PCRE-WH were 0.541 and 0.537,respectively.Areas under the curve for 3-and12-month recurrent stroke estimated by PCRE-AA were 0.539 and 0.537,respectively.2.Validation of the Pooled Cohort Risk Equation for predicting 3-and12-month poor clinical outcome and all-cause death in non-atrial fibrillation ischemic stroke and TIA patients.Areas under the curve for 3-and 12-month poor clinical outcome estimated by PCRE-WH were 0.560 and 0.579,respectively.Areas under the curve for 3-and 12-month poor clinical outcome estimated by PCRE-AA were 0.536 and 0.551,respectively.Areas under the curve for 3-and 12-month poor clinical outcome estimated by ESRS were 0.578 and 0.630,respectively.Areas under the curve for 3-and 12-month poor clinical outcome estimated by SPI-II were 0.626 and 0.649,respectively.Areas under the curve for 3-and 12-month all-cause death estimated by PCRE-WH were 0.546 and 0.559,respectively.Areas under the curve for 3-and 12-month all-cause death estimated by PCRE-AA were 0.530 and 0.538,respectively.Areas under the curve for 3-and 12-month all-cause death estimated by ESRS were 0.596 and 0.600,respectively.Areas under the curve for 3-and 12-month all-cause death estimated by SPI-II were 0.610 and 0.622,respectively.Conclusion The predictive capability of the Pooled Cohort Risk Equation to predict 3-and 12-month recurrent stroke in non-atrial fibrillation ischemic stroke and TIA patients is relatively poor.The predictive capability of the Pooled Cohort Risk Equation to predict 3-and 12-month poor clinical outcome and all-cause death in non-atrial fibrillation ischemic stroke and TIA patients is relatively poor.