Development And Validation Of A Nomogram For Predicting The Risk Of Post-Stroke Cognitive Impairment

Objectives:The aim of this study was to develop and internally validate a nomogram for predicting the risk of post-stroke cognitive impairment(PSCI)risk in patients with mild acute ischemic stroke(AIS).Methods:A total of 315 mild AIS patients admitted to the Department of Neurology of the First Hospital of Jilin University from April 2019 to January 2021 were selected,including 181 patients in the PSCI group and 134 patients in the non-PSCI group.We chose 15 potential predictors associated with vascular cognitive impairment(VCI),including age,gender,education level,hypertension history,history of diabetes,heart disease history,previous history of stroke,hyperlipidemia,smoking history,history of drinking,Fazekas score,diameter of maximum transverse section(DMTS),TOAST classification,OCSP classification and intracranial artery stenosis(ICAS).The least absolute shrinkage and selection operator regression(LASSO)model was used to optimize forecasting indicators selection for the PSCI risk model.Finally,it was reduced to 10 forecasting indicators that had a greater impact on PSCI.Multivariable logistic regression analysis was applied to build a predicting model incorporating the 10 forecasting indicators selected in the LASSO model.Discrimination,calibration,and clinical usefulness of the predicting model were assessed using the C-index,calibration plot,and decision curve analysis(DCA).Internal validation of the PSCI predicting model was used the Bootstrapping validation.Results:Forecasting indicators contained in the prediction nomogram model of PSCI included age,sex,education level,past stroke history and DMTS.The nomogram displayed good discrimination with a C-index of 0.708(95% confidence interval: 0.651-0.765).High C-index value of 0.682 could still be reached through the internal validation.The calibration curve also demonstrated good agreement in this study.DCA showed that when the PSCI threshold probability is more than 27%,higher net benefits can be obtained by using the PSCI nomogram model developed in the current study.Conclusions:This PSCI nomogram model incorporating the age,sex,education level,past stroke history and DMTS could be conveniently for clinicians to predict the PSCI risk of patients with mild AIS and provide timely and reasonable treatment and intervention for patients at high risk of PSCI.This PSCI nomogram model is a tool worthy of clinical promotion and application.