Development and validation of predictive model for cognitive impairment after stroke / 西安交通大学学报(医学版)

ABSTRACT

【Objective】 To construct and validate a risk prediction model for cognitive impairment after stroke based on demographic, clinical, and neuroimaging characteristics. 【Methods】 Through the medical record system, we collected all data of the patients. We finished cognitive function testing three months after the indexed stroke. The Mini-Mental State Examination Scale score≤26 was defined as cognitive dysfunction. Optimal subset regression analysis was used to screen variables, Logistic regression analysis was used to construct a predictive model for cognitive impairment, and C-index, calibration chart and clinical decision curve analyses were used to evaluate the discrimination, consistency, and clinical availability of the model. And nomograms were used to express the performance of the model. 【Results】 Seven variables were selected cognitive function before stroke, age, years of education, National Institutes of Health Stroke Scale score at admission, history of ischemic heart disease, the number of old lacunar infarct lesions, and medial temporal lobe atrophy scale. The prediction model had a C-index of 0.845 (95% CI 0.805-0.885). The clinical decision curve showed that the model had a positive net benefit when the threshold probability was 9.0%-90.0%. 【Conclusion】 The predictive model of cognitive impairment in stroke patients has good predictive efficiency and provides an effective assessment tool for screening high-risk cases of cognitive impairment in patients with stroke of various subtypes.