Objective:

To develop and validate a prediction model using EMR data through natural language processing (NLP) and machine learning (ML) to assess the risk of stroke in emergency department patients with acute dizziness, addressing the limitations of current diagnostic methods.

Key Findings:

• Stroke is frequently misdiagnosed in patients presenting with dizziness, with existing diagnostic methods having notable limitations, including a 35% miss rate for vestibular strokes.
• Advanced age, male gender, diabetes, atrial fibrillation, previous cerebrovascular disease, recurrent vertigo, and high blood pressure are known stroke predictors.

Interpretation:

The study highlights the urgent need for improved diagnostic tools in emergency settings to accurately assess stroke risk in patients with dizziness, leveraging EMR data and advanced analytical techniques to enhance patient outcomes.

Limitations:

• The study is retrospective and may be subject to biases inherent in historical data.
• The reliance on EMR data may limit the generalizability of findings to other settings or populations.
• Potential biases may arise from the methods used for feature extraction.

Conclusion:

A clinical decision support system is essential for enhancing stroke risk assessment in emergency department patients with dizziness, potentially reducing misdiagnosis and improving patient outcomes.