2247Diagnostic accuracy of a novel mobile phone application in monitoring atrial fibrillation

Abstract

Background: Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia in adults, associated with significant morbidity, increased mortality, and rising healthcare costs. Recently, a Cardiio Rhythm Mobile Application (CRMA) was demonstrated to be an accurate screening tool for AF in a general population. We sought to evaluate the diagnostic accuracy of a CRMA for monitoring in patients with known AF. Method(s): Patients with AF, scheduled for elective cardioversion (CV), were enrolled in the study. CRMA finger pulse recordings, utilizing an iPhone camera, were obtained three times before (pre-CV) and after (post-CV) the cardioversion in each patient. Detection of an irregular rhythm on two out of three CRMA recordings was needed to label the rhythm as AF (Figure 1). The findings were validated against 12-lead electrocardiograms (ECGs) or, in rare instances, rhythm strips, obtained as part of standard CV. ECGs were reviewed by two independent cardiologists blinded to the CRMA results. Result(s): 113 patients, 87 (77%) men, mean age of 67.5+/-10.5 years, were enrolled in the study. Pre-CV CRMA readings were analyzed in 112 patients (no ECG for validation in one case). Post-CV CRMA readings were analyzed for 107 out of 109 patients that underwent CV (2 left before the recording was obtained). Overall, the CR application correctly labeled 91 of 95 atrial fibrillation recordings (95.8%) as AF and 96 of 103 non-AF recordings (93.2%) as non-AF. The sensitivity was 95.8% (95% CI = 89.6% - 98.8%) and specificity was 93.2% (95% CI = 86.5% - 97.2%). The positive predictive value was 92.9% (95% CI = 85.8% - 97.1%) and the negative predictive value was 96.0% (95% CI = 90.1% - 98.9%). Conclusion(s): The Cardiio Rhythm Mobile Application demonstrates promising potential in accurate detection and discrimination of AF from normal sinus rhythm in patients with a history of AF. Further studies are needed to assess the efficacy of CRMA in early diagnosis of AF recurrence and its effect on clinical outcomes. (Figure Presented).