Abstract:
Objective To analyze the risk factors of post-stroke cognitive impairment (PSCI), construct the prediction model of nomogram, and verify its application effects.
Methods A total of 356 patients with the first acute ischemic stroke were selected, including 211 cases in the modeling group and 145 cases in the verification group. After 3 months of stroke, the cognitive impairment of patients in the modeling group was assessed by using the Montreal Cognitive Assessment Scale, the influencing factors of PSCI were analyzed, the PSCI risk prediction histogram model was constructed. The differentiation and calibration degree of the model were evaluated by ROC curve and Hosmer-Lemeshow goodness of fit test, and the clinical decision curve of the prediction model was drawn.
Results The age, education level, occupation, NIHSS score on admission, intracranial artery stenosis and baseline social support rating scale score were the independent factors of PSCI (P ＜ 0.05 to P ＜ 0.01). The PSCI nomogram prediction model was constructed based on independent risk factors, and the model consistency index was 0.826. ROC curve showed that the AUC of PSCI predicted by the model was 0.842, the sensitivity was 80.5%, and the specificity was 73.0%. The AUC of PSCI in the verification group was 0.839, the sensitivity was 86.5%, and the specificity was 68.4%. The clinical decision curve showed that the prediction model had clinical application value.
Conclusions The PSCI risk prediction model has good predictive efficacy, which is conducive to early clinical identification of PSCI high-risk groups.
下载: