Validation of a novel smartphone-based photoplethysmographic method for ambulatory heart rhythm diagnostics

Abstract

Abstract Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): Vinnova - The Swedish Agency for Innovation Systems. Region Stockholm Innovation Fund. Introduction Devices for long-term and ambulatory recording of heart rhythm have limited availability within health care systems worldwide, potentially leading to missed diagnoses and limitations in peri-procedural heart rhythm management. As smartphones are becoming ubiquitous the availability of heart rhythm diagnostics and management of atrial fibrillation (AF) using smartphone photoplethysmography (PPG) compared to electrocardiography (ECG) is increasing. Previous validation studies of smartphone-PPG applications have all been performed under supervision in healthcare settings. In addition, no previous study has validated smartphone-PPG compared to simultaneous ECG recordings, with manual heart rhythm interpretation of the PPG recordings nor have they included patients with atrial flutter (AFL). Purpose The aim of this study was to validate a novel smartphone-PPG method for heart rhythm diagnostics in patients with AF and/or AFL when in use unsupervised in an ambulatory setting. Methods Unselected patients undergoing direct current cardioversion at a University Hospital for treatment of AF or AFL were asked to perform one-minute heart rhythm recordings post-treatment at least twice daily for 30 days in their home environment. All included were provided with an unmodified iPhone 7 smartphone running the CORAI Heart Monitor PPG application simultaneously with a single-lead ECG recording (KardiaMobile). PPG and ECG recordings were interpreted independently by two readers. Results In total 280 patients, with median age of 69.0 years (31% women) were included from November 2018 to July 2020 and registered 18 005 simultaneous PPG and ECG recordings. Of the PPG recordings 96.9% had sufficient quality for diagnosis compared to 95.1% of the ECG recordings (p < 0.001). Precardioversion ECG recordings were interpreted as AF in 82.1%, AFL in 14.3% and as having insufficient quality for diagnosis in 3.6% of the patients. After removal of recordings with insufficient quality 69.7% of ECG recordings were interpreted as sinus rhythm, 28.2% as AF and 2.1 % as AFL. Manual interpretation of the PPG recordings diagnosed AF/AFL (sensitivity) in 97.7% and sinus rhythm (specificity) in 99.4% of the recordings compared to manually interpreted ECG recordings, with an overall accuracy of 98.9%. Results excluding recordings interpreted as AFL on ECG or PPG diagnosed AF (sensitivity) in 99.0% and sinus rhythm (specificity) in 99.7% of the recordings, with an overall accuracy of 99.5%. Conclusion A novel smartphone-PPG method can be used by patients unsupervised in their home environment for accurate heart rhythm diagnostics of AF and AFL with high sensitivity and specificity.