Predicting poststroke cognitive impairment after acute ischemic stroke based on admission characteristics

Area of Science:

• Neurology
• Clinical Prediction Modeling
• Stroke Research

Background:

• Post-stroke cognitive impairment (PSCI) is a common complication following acute ischemic stroke (AIS).
• Early identification of patients at risk for PSCI is crucial for timely intervention and improved outcomes.
• Existing prediction tools for PSCI may lack robustness or generalizability.

Purpose of the Study:

• To develop and validate a robust clinical prediction model for PSCI within six months post-AIS.
• To identify key predictors associated with the development of PSCI in AIS patients.
• To assess the discriminative and calibration performance of the developed model.

Main Methods:

• A cohort of 386 AIS patients was divided into training (270) and validation (116) sets.
• Multifactor logistic regression identified independent predictors of PSCI.
• The prediction model was evaluated using Area Under the Receiver Operating Characteristic Curve (AUC-ROC) and calibration plots.

Main Results:

• The final model incorporated variables including smoking, alcohol consumption, gender, education, NIHSS score, stroke progression, systolic blood pressure, diabetes, atrial fibrillation, coronary heart disease, LDL cholesterol, β2-microglobulin, and Lp-PLA2.
• The model achieved an AUC-ROC of 0.862 in the training set and 0.806 in the validation set.
• Calibration analyses confirmed the model's good predictive performance.

Conclusions:

• The developed clinical prediction model demonstrates high discriminative ability for identifying PSCI risk in AIS patients.
• The model offers valuable guidance for clinical decision-making in stroke care.
• Further validation using multicenter data is recommended to enhance the model's robustness and clinical applicability.