Article Text
PDF
Abstract
Objective Early prediction of atrial fibrillation (AF) development would improve patient outcomes. We propose a simple and cheap ECG based score to predict AF development.
Methods A cohort of 16 316 patients was analysed. ECG measures provided by the computer-assisted ECG software were used to identify patients. A first group included patients in sinus rhythm who showed an ECG with AF at any time later (n=505). A second group included patients with all their ECGs in sinus rhythm (n=15 811). By using a training set (75% of the cohort) the initial sinus rhythm ECGs of both groups were analysed and a predictive risk score based on a multivariate logistic model was constructed.
Results A multivariate regression model was constructed with 32 variables showing a predictive value characterised by an area under the curve (AUC) of 0.776 (95% CI: 0.738 to 0.814). The subsequent risk score included the following variables: age, duration of P-wave in aVF, V4 and V5; duration of T-wave in V3, mean QT interval adjusted for heart rate, transverse P-wave clockwise rotation, transverse P-wave terminal angle and transverse QRS complex terminal vector magnitude. Risk score values ranged from 0 (no risk) to 5 (high risk). The predictive validity of the score reached an AUC of 0.764 (95% CI: 0.722 to 0.806) with a global specificity of 61% and a sensitivity of 55%.
Conclusions The automatic assessment of ECG biomarkers from ECGs in sinus rhythm is able to predict the risk for AF providing a low-cost screening strategy for early detection of this pathology.
- atrial fibrillation
- biomarkers
Data availability statement
No data are available.
Statistics from Altmetric.com
Data availability statement
No data are available.
Footnotes
AS-G and AC are joint first authors.
GJO and JJ-B are joint senior authors.
GJO and JJ-B contributed equally.
Contributors JJ-B conceived this study. GJO and AS carried out the numerical analysis, JMC collected, selected and provided XML files. JJ-B, AC, JMC and FA evaluated and interpreted numerical results and all ECGs. AS carried out the numerical analysis, performed all the statistical analysis and the literature search. All co-authors produced the initial draft of the manuscript and reviewed the final manuscript version. GJO and JJ-B are guarantors of this paper.
Funding Authors received a research grant from the Carlos III Institute of Health under the health strategy action 2020–2022 with reference PI20/00792.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.