A predictive score for atrial fibrillation in poststroke patients

Area of Science:

• Cardiology
• Neurology
• Medical Informatics

Background:

• Atrial fibrillation (AF) is a significant risk factor for cerebral ischemia.
• Detecting paroxysmal AF can be challenging, necessitating efficient risk stratification tools, especially in resource-limited settings.

Purpose of the Study:

• To develop and validate a predictive scoring tool for incident atrial fibrillation in patients undergoing poststroke follow-up.

Main Methods:

• Retrospective longitudinal study of 872 patients with cerebral ischemia from 2014-2021.
• Utilized electronic medical records, including demographic, clinical, laboratory, ECG, echocardiogram, and neuroimaging data.
• Employed stepwise logistic regression to identify predictors and created an integer-based risk score.

Main Results:

• The developed score incorporates left atrial diameter (≥ 42 mm), age (≥ 70 years), septal aneurysm, and NIH Stroke Scale score (≥ 6).
• A score of ≥ 2 indicated a fivefold increased risk of incident AF detection.
• The score demonstrated good predictive accuracy with an Area Under the Curve (AUC) of 0.77.

Conclusions:

• An accurate risk score tool for incident AF has been developed.
• This tool can assist in identifying high-risk individuals for targeted AF investigation poststroke.
• Further validation in multicenter samples is recommended for broader clinical application.