Validation of Tele-HRV: A Novel Digital Health Solution for Cardiovascular Risk Detection in Diabetes

Abstract

BackgroundHeart rate variability (HRV) in a sensitive test of Cardiac autonomic neuropathy (CAN) and predicts high mortality in diabetes subjects. Our innovative TeleHRV solution conforms to the HRV Task Force Standards and consists of mApp controlled battery powered handy device that uploads data to the cloud and delivers eReport. It eliminates all cost and technical constraints and makes HRV available for mass application. Aim of this study was to compare the computation accuracy of standard HRV indices using TeleHRV analysis software and its open-source equivalent software Kubios HRV which is considered the gold standard. MethodsData for this project was downloaded from Physionet.org. Standard Time domain indices (mean RR, SDNN, RMSSD, NN50) and Frequency Domain Indices (LF, HF and LF/HF) were calculated with TeleHRV analysis software and Kubios HRV software and Correlation Coefficient (CC) as well as Intra-Class Correlation Coefficient (ICCC) were computed for the indices obtained from the two software. ResultsThe Maximum of ICCC =1 (which is perfect agreement) was obtained for mean RR interval while the Minimum CC= 0.9976 was obtained for LF/HF which is extremely well correlated for a minimum observation; CC and ICCC for other indices were in between for the two software. ConclusionThe statistics strongly suggests that TeleHRV's analysis software faithfully matches the HRV indices values calculated by the gold standard Kubios HRV and support the validity of TeleHRV, which provides a cost effective, handy and user-friendly technology in a resource constrained and remote locations for early detection of CAN in diabetes.