Previously submitted to: JMIR AI (no longer under consideration since Feb 24, 2025)
Date Submitted: Nov 4, 2022
Warning: This is an author submission that is not peer-reviewed or edited. Preprints - unless they show as "accepted" - should not be relied on to guide clinical practice or health-related behavior and should not be reported in news media as established information.
Machine learning predicting atrial fibrillation as an adverse event in the Warfarin Versus Aspirin in Reduced Cardiac Ejection Fraction (WARCEF) trail
ABSTRACT
Background:
Atrial fibrillation (AF) and heart failure (HF) commonly coexist due to shared pathophysiological mechanisms. Being able to recognise and identify patients with HF at risk of developing AF would allow clinicians the opportunity to implement appropriate monitoring strategy and timely treatment, reducing the impact of AF on patient’s health.
Objective:
We aim to utilise machine learning models to identify risk factors useful in predicting AF in patients with heart failure
Methods:
Utilising patient level data from Warfarin and Aspirin in Patients with Heart Failure and Sinus Rhythm (WARCEF) study, four machine learning (ML) models combined with logistic regression and cluster analysis were applied to identify factors which predict development of AF in patients with HF. Out of the (n = 2219) patients included for analysis, (n = 215, 9.7%) presented AF as an adjudicated adverse event during the 6 years (mean [±SD], 3.5±1.8) follow up period.
Results:
Logistic regression applied to patients of a white racial ethnicity only shows that white divorced patients have a 1.75-fold higher risk of AF than white patients reporting other marital statuses (95% CI 1.19–2.57, p-value = 0.002). By contrast, similar analysis for the non-white racial ethnicity only patients suggests that non-white patients who live alone have a 2.58-fold higher risk of AF than those not living alone (95% CI 1.45–4.59, p-value < 0.001). ML analysis also identified “marital status” and “line alone” as relevant predictors of AF. Apart from previously well-recognised factors (e.g., age, heart failure), the ML algorithms and cluster analysis identified 2 clearly distinct clusters, namely white and non-white ethnicities. This should serve as reminder of the impact of social factors on health and the need to explore the impact of social factors on the under-represented non-white ethnicity group of patients and the potential impact social interventions, or the lack thereof, have on them.
Conclusions:
The use of ML can prove useful in identifying novel risk factors. Our analysis has highlighted that “social factors”, such as living alone, may disproportionately increase the risk of AF in the under-represented non-white patient groups with HF. The study also highlights the need for more studies focusing on stratification of multiracial cohorts to better uncover the heterogeneity of AF mechanisms across different racial groups. Clinical Trial: N/A
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.