ABSTRACT
Containing the COVID-19 pandemic requires rapidly identifying infected individuals. Subtle changes in physiological parameters (such as heart rate, respiratory rate, and skin temperature), discernible by wearable devices, could act as early digital biomarkers of infections. Our primary objective was to assess the performance of statistical and algorithmic models using data from wearable devices to detect deviations compatible with a SARS-CoV-2 infection. We searched MEDLINE, Embase, Web of Science, the Cochrane Central Register of Controlled Trials (known as CENTRAL), International Clinical Trials Registry Platform, and ClinicalTrials.gov on July 27, 2021 for publications, preprints, and study protocols describing the use of wearable devices to identify a SARS-CoV-2 infection. Of 3196 records identified and screened, 12 articles and 12 study protocols were analysed. Most included articles had a moderate risk of bias, as per the National Institute of Health Quality Assessment Tool for Observational and Cross-Sectional Studies. The accuracy of algorithmic models to detect SARS-CoV-2 infection varied greatly (area under the curve 0.52-0.92). An algorithm's ability to detect presymptomatic infection varied greatly (from 20% to 88% of cases), from 14 days to 1 day before symptom onset. Increased heart rate was most frequently associated with SARS-CoV-2 infection, along with increased skin temperature and respiratory rate. All 12 protocols described prospective studies that had yet to be completed or to publish their results, including two randomised controlled trials. The evidence surrounding wearable devices in the early detection of SARS-CoV-2 infection is still in an early stage, with a limited overall number of studies identified. However, these studies show promise for the early detection of SARS-CoV-2 infection. Large prospective, and preferably controlled, studies recruiting and retaining larger and more diverse populations are needed to provide further evidence.
ABSTRACT
INTRODUCTION: Prior research has demonstrated how immune functioning and physiological signals fluctuate across the menstrual cycle, with eumenorrheic womenmore likely to become ill during the luteal phase. Examining such changes during the current pandemic, we explored how the relationship betweenmenstrual cycle phase and physiological signals impacts a wearable medical device's ability to detect COVID-19. METHODS: The largest institutional review board-approved wearable device study monitoring SARS-CoV-2's effects on biophysiology to date, COVID-RED aims to develop a machine learning algorithm predicting an infection up to 3 days prior to symptom onset. Wearing the device nightly, participants (N=17,824) sync it with a mobile application and log SARS-CoV-2 diagnostic tests, symptoms, and menses in the app's Daily Diary. The algorithm ingests physiological and selfreported features to provide each user with a real-time update about their likelihood of infection. RESULTS: Daily infection likelihood and predictions of ovulation using proprietary algorithms were generated during a 9-month period for 1,574 eumenorrheic women (n=3,281 menstrual cycles) not currently on hormonal birth control. The negative/positive ratio of predicted COVID-19 cases during the 5-day period preceding ovulation was 2.94 compared to 4.83 in the 5 days post-ovulation (chi-square (1, N=33,920)5343.34, P<.0001). Participants reported 22 SARS-CoV-2 positive test results, with five times as many confirmed infections occurring in the postovulatory period (n510) compared to the preceding 10-day window (n=2). CONCLUSION: Demonstrating that machine-learning algorithms ingesting wearable data should consider menstrual cycle impact, our findings suggest that women may be more susceptible to SARS-CoV-2 during their luteal phase, with further studies needed to disentangle underlying mechanisms.