Early detection of SARS-CoV-2 and other infections in solid organ transplant recipients and household members using wearable devices

ABSTRACT

Purpose: Wearable devices that measure physiological parameters have shown utility for detecting infections such as influenza and recently COVID-19 up to 10 days before clinical symptoms appear. Combining symptom data with wearable biosensor data has proven to increase discrimination between COVID-19 and non-COVID-19 infection compared to using symptom data alone (AUC 0.80 vs. 0.71, p<0.01). Here we study the utility of wearable devices in early detection of SARS-Cov2 and related infections in pediatric solid organ transplant recipients. Early remote detection of infections may guide treatment responses to improve clinical outcomes such as rates of hospitalization. Methods: This is an ongoing prospective cohort study of pediatric solid organ transplant recipients and their non-transplanted household members. We are currently remotely recruiting all participants from multicenter sites and heart, liver and lung transplant patients from a single transplant center. We continuously monitor heart rate (HR), body temperature, oxygen saturation, blood pressure, sleep and respiratory patterns, and electro-dermal activity. We use MyPHD, a HIPAA compliant information architecture that supports EHR integration, for remote patient recruitment, secure data collection, and analyses. We apply two real-time algorithms to the data to identify changes that are associated with COVID-19. The algorithms are based on Resting Heart-Rate-Difference (RHR-Diff) and identify periods of elevated HR based on outlier interval detection, calculating standardized residuals for each HR observation compared to a baseline of clinically validated “healthy days” for each patient. Results: Continuous real-time physiological monitoring of transplant patients may provide syndromic surveillance and inform healthcare management. The primary outcome is time to infection diagnoses, with a particular emphasis on SARS-CoV2 and common post-transplant infections (Influenza, EBV, CMV, and BK virus). The secondary outcomes are to optimize our algorithms for the pediatric transplant setting and to monitor for other complications including cardiometabolic complications and eGFR decline. Conclusions: The potential impact of this study include algorithm-guided early detection of infection signatures coupled with provider clinical-decision-support and return-of-results to manage transplant patient care.