Learn more about our Long Covid research study and find out if you are eligible.
Long Covid (or Post Covid Syndrome) has become a major cause of morbidity and functional impairment for many people who were afflicted with COVID-19. To address this health challenge, we designed a project to study patterns in wearable sensor data that reflect how the body responds to COVID-19 months after the initial acute infection. Our objectives in the study were: a) To identify signature wearable data patterns that indicate Physiological abnormalities from baseline in COVID Long haulers; b) To develop Machine learning and deep learning models that can use pre-infection or infection period wearable data to predict post-infection Long COVID trajectories; c) To formulate preemptive protocols.
We perform statistical disease modeling of Post Acute COVID-19 by identifying signature patterns in clinical symptoms. This will eventually help to improve the understanding of the pathophysiologic basis behind Long COVID and also help to formulate pre-emptive treatment protocols that could significantly reduce disease burden in patients. Our initial results from analysing the data collected through MyPHD show that at a population level, irrespective of the presence of lingering symptoms, we see a sustained elevation in the mean resting heart rate even upto 90 days after recovery in people who were affected with SARS-CoV-2 infection.
These persistent deviations from baseline in wearable signals and clinical symptoms following recovery from SARS-CoV-2 infection requires more in-depth analysis. To better understand the risk factors and symptomatology of Long COVID and correlate various severity stages with abnormalities in resting heart rate, we sent out a detailed survey through our myPHD mobile application to all 5506 participants in our study. We are performing a two-way analysis. First by using symptoms to differentiate COVID long haulers from normal subjects, we analyse the wearable data. Next by using wearable data patterns to differentiate those with abnormal deviations from normal subjects, we analyse their symptoms. We have attached the full survey as an annexure. In summary, the survey gathered data on the following:
Shown below is an example figure showing the tracing of resting heart rate for subjects experiencing Chronic Post Covid Syndrome (symptoms over 12 weeks after recovery). The monthly median RHR shows a consistent step up from the previous value throughout the Post Covid phase. By finer classification of Long Covid subjects based on which symptom was most reported, features of wearable signals that may be specific to particular types of symptoms will be identified.
Contributed by Ekanath Srihari Rangan
