Evaluation of the feasibility and accuracy of remote mobile app-based self-reported atrial fibrillation risk factor assessment in patients with atrial fibrillation: TeleCheck-AF results

Abstract Background Previously, we introduced the TeleCheck-AF approach, which is an on-demand mobile health (mHealth) infrastructure incorporating mobile app-based heart rate and rhythm monitoring as well as mobile app-based self-reported atrial fibrillation (AF) risk factor assessment to allow comprehensive remote AF management through teleconsultation. Herein, we evaluated the feasibility and accuracy of remote mobile app-based self-reported AF risk factor assessment in AF patients. Methods In our University Medical Center, 545 patients were managed within the TeleCheck-AF project by an on-demand heart rate and rhythm mHealth infrastructure through teleconsultation. Patients were asked to fill in a short mobile app-based 10-item questionnaire related to AF risk factors. A reminder to complete the questionnaire automatically popped-up after the following four heart rate and rhythm recordings. Furthermore, patient's medical history was retrieved from the electronic health records (EHRs). Results Out of 545 patients, 542 (99.4%) patients (217 female, age 67 (59–72) years) completed the mobile app-based 10-item questionnaire and were included in this analysis. The number of patients with diabetes mellitus was similar in the EHRs and mobile app-based questionnaire (both 11.3%, p=1.000). There was no significant difference in the number of patients who had a medical history of transient ischemic attack (TIA)/cerebrovascular accident (CVA) and artery disease (coronary artery disease and peripheral artery disease) in the EHRs and mobile app-based questionnaire (11.4% vs 12.2%, p=0.608 and 14.8% vs 13.3%, p=0.366, respectively). Heart failure was more frequently reported in the mobile app-based questionnaire compared to the EHRs (33.4% vs 14.0%, p<0.001). A total of 260 (48.0%) patients had a diagnosis of hypertension verified in EHRs and only 239 (44.1%) patients reported hypertension in the mobile app-based questionnaire (p=0.044). There was no significant difference in number of patients with CHA2DS2-VASc-score ≥2 between the EHRs and mobile app-based questionnaire (64.2% vs 66.1%, p=0.275). The accuracy of mobile app-based assessment of diabetes mellitus was 85.4%, of TIA/CVA 78.9%, of artery disease 60.9%, of heart failure 78.8%, and of hypertension 89.3%. Conclusion Patient self-reported AF risk factors by a remote mobile app-based assessment is feasible and may be useful for future digital trials and comprehensive remote AF management through teleconsultation. Funding Acknowledgement Type of funding sources: None.