46Smartphone apps for the detection of AF

Abstract

Background: Improving availability of screening methods of AF, could reduce the number of thromboemobilc strokes. Commercially available smartphone applications that utilise the smartphones intrinsic function without the need for a peripheral device, are available for consumer use. However, there is limited information on their reliability. Method(s): Two smartphone applications were identified in app stores that claim to detect AF using the smartphones own intrinsic functionality; BeatScanner (utilises the built in accelerometer and gyroscope sensors in a smartphone device) and Preventicus (utilises smartphone camera). Patients ware recruited from inpatient wards and outpatient clinics. Each had a 12 lead ECG and filled out a questionnaire on their user experience. Exclusion criteria included patients with frequent premature ectopic beats, paced rhythms and sinus arrhythmia. Result(s): 140 patients were recruited; 70 patients in AF rhythm and 70 patients in sinus rhythm. Mean age was 73+/-12 years( 60% Male) in AF group and 59+/-20 years (66% Male) in sinus group. The Preventicus app had a sensitivity of 94.2% (95% CI: 85.2- 98.1) and specificity of 95.7% (87.2-98.9). The BeatScanner app had a sensitivity of 88.6% (0.784-0.946) and specificity of 67.1% (54.6- 77.6). From the cohort, 59.2% were smartphone users. 94% of smartphones users reported that they would be interested to self screen with smartphones apps and 55% would pay for an app to do so. 96% of smartphone users found the apps easy to use. 63% of smartphone users preferred the Preventicus app to the BeatScanner app. Conclusion(s): Smartphones apps have a good sensitivity for detection of AF, particularly the Preventicus app which demonstrated excellent specificity and sensitivity. Technology that uses the smartphones own intrinsic functions are easy to use, and should be explored further for screening of AF in the general population.