Validation Of Health-related Software Application

Abstract

In today’s evolving technological society, telehealth, telemedicine, and e-health are powered by the confluence of mobile devices, ubiquitous internet access, and Artificial Intelligence. PURPOSE: To determine the validity of a health-related software application for measurement of heart rate and blood pressure. METHODS: Physiological variables collected twice in participants (Age = 20.8±3.0 yrs; Height = 168.7±8.6 cm; Body Mass = 68.3±14.3 kg; % Body Fat = 21.4±8.3) were heart rate (HR), systolic blood pressure (SBP), and diastolic blood pressure (DBP) via digital monitor (DM) and software application (SA) at rest, seated in a chair with legs uncrossed. Pearson Correlation Coefficient was used to determine the relationship between the measurements. Paired sample t-test was used to determine if statistical mean differences in HR, SBP, and DBP existed, and mean absolute error (MAE) were calculated to get an error estimate on an individual basis. RESULTS: Table 1, for HR, the SA was highly correlated with DM, there were no significant mean differences, and the MAE was 4 bpm. For SBP, correlations were moderate to low, the SA significantly underestimated SBP, and the MAE ranged from 10.6-17.9 mmHg. For DBP, results were similar to the SBP with the exception of the Male Only DBP which was not significant different between devices, however the MAE was 8.8 mmHg.Table 1. Heart rate and blood pressure values for male and female, female only, and male only.CONCLUSION: Resting HR values were consistent between both devices amongst all participants. Systolic BP, however, SA significantly and systematically underestimated DM measurements amongst all participants. Regarding DBP, SA significantly and systematically underestimated DM measurements amongst all participants, except for male only.