Smartwatches in Clinical Pre-diagnosis: Enhancing Tilt Test Analysis for Prolonged COVID-19 Symptoms

Ana Leticia Gomes dos Santos1, Samuel Minucci Camargo1, William Tsutomu Watanabe2, Stella Tassinari Maximo3, Kelly Correa Baioco Da Silva4, Christian Goncalves Sassaki1, Jose L Puglisi5, Daniel Gustavo Goroso1
1University of Mogi das Cruzes, 2University of Mogi das Cruzes, 3Sao Leopoldo Mandic, 4Universidade Mogi das Cruzes, 5California North State University


Abstract

With technological advancements, smartwatches have emerged as multifunctional devices that enable non-invasive real-time collection of biomedical data, continuously monitoring heart rate with optical sensors and determining postural changes with an integrated accelerometer. However, conventional apps face limitations in analyzing heart rate variability (HRV) due to reduced sampling rates. In this context, we developed and validated a specific app for smartwatches, focusing on clinical pre-diagnosis for the "Tilt Test," emphasizing patients with Long-COVID-19 and post-COVID-19. The App focuses on HR measurement and accelerometer analysis to determine body transitions, serving as a basis for HRV analysis in different postural situations, aiming to evaluate the autonomic response of the sympathet-ic/parasympathetic system. We selected 44 participants based on specific criteria, divided into control and study groups, applying a 50-minute tilt-test protocol with three phases: 15 min. horizontal, 15 min. inclined, and a 20 min. recovery. Throughout the procedure, we collected data simultaneously using ECG as the reference standard and smartwatches positioned on both the wrist and ankle. To ensure app accuracy, we conducted calibration tests, in-cluding linear regression and the Bland-Altman test to analyze result con-cordance. Additionally, we used the Mann-Whitney test to observe potential differences between study groups. Data from the smartwatch were compared with ECG data, showing a high correlation (r=0.98). Concordance corrections were also conducted. Statistical analysis of data collected by both ECG and smartwatch revealed significant differences between study and control groups (p<0.05). The App has the potential to enable "at-home tilt-tests," enhancing accessibility to clinical pre-diagnosis through smartwatch use and potentially aiding in daily life activities. The data collection method is particularly use-ful for detecting Postural Orthostatic Tachycardia Syndrome (POTS) in pa-tients with LongCOVID-19 or post-COVID-19 symptoms.