Back

A Cross Sectional Study of Atrial Fibrillation Prevalence and Associated Cardiac Arrhythmias in Stroke Patients: Insights from Commercial Data Archives

Soo, K.-M.; Liew, T.-S.; Heath, P.

2024-04-22 epidemiology
10.1101/2024.04.21.24306150 medRxiv
Show abstract

The link between atrial fibrillation (AF) and stroke has gained attention due to AFs significant impact on stroke risk. Detecting AF, especially paroxysmal AF, presents challenges, highlighting the necessity for continuous Holter monitoring. In our investigation, we utilized commercial data archives to examine AF prevalence and associated cardiac arrhythmias among stroke patients and controls. We also analyzed various factors, such as age, gender, symptoms, test duration, and duration post-stroke, to identify predictors of arrhythmias and enhance paroxysmal detection rates. We included 98 stroke patients and 98 controls for comparing AF prevalence within 72 hours post-stroke and analyzed 779 stroke patients within 4 days to 10 weeks post-stroke. Our findings revealed a higher AF detection rate in stroke patients compared to controls, with AF being the most prevalent arrhythmia observed. Older age and symptom presence were associated with significant findings, particularly AF. Temporal analysis showed a negative correlation between AF and complete heart block occurrence over time, while ventricular tachycardia exhibited a positive correlation. Paroxysmal fibrillation detection rates were highest within the initial month, with the optimal duration of Holter monitoring being 10 days, underscoring the importance of extended monitoring for detecting paroxysmal atrial fibrillation. Detecting paroxysmal atrial fibrillation could influence stroke management. Furthermore, our study suggests the potential utility of commercial data in yielding comparable outcomes to research data.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.