The COVID-19 pandemic has underscored the significance of staying healthy, and our mission to assist individuals reside healthier lives has never been more essential. Since the start of this international health crisis, the Fitbit research staff has been diligently working to help make a distinction in the battle towards COVID-19. This consists of accelerating work on early illness detection, an effort led by our group of knowledge scientists with deep experience in machine learning and predictive modeling.
In May, we announced the launch of the Fitbit COVID-19 research aimed at building an algorithm that detects COVID-19 earlier than signs start. In simply over two months, greater than 100,000 Fitbit customers across the US and Canada have enrolled, with more than 1,000 constructive instances of the virus reported. This research presents an exciting opportunity to see how the facility of the Fitbit group will assist us better understand this new and sophisticated illness. If you’re focused on contributing to this analysis, you’ll be able to be a part of right here or within the Assessments & Discover tab within the Fitbit app.
Because knowledge to assist detect COVID-19 is of crucial international significance, we’ve submitted our early research for publication in a peer-reviewed journal. While we work to finalize the publication, we’ve made the complete manuscript publicly out there as a preprint, which permits us to share some of the preliminary findings.
Early Findings From Fitbit COVID-19 Study
So far, we’re encouraged to see physiological signs of illness detected by Fitbit units concurrently with research individuals’ reporting the onset of COVID-19 symptoms, and in some instances even earlier than.
Based on the findings of our research, we will detect almost 50 % of COVID-19 instances in the future before members reported the onset of symptoms with 70 % specificity.
This is necessary because individuals can transmit the virus before they understand they have signs or once they haven’t any signs at all. If we will let individuals know they need to get examined a day before signs begin, they will isolate and seek care sooner, serving to to scale back the spread of COVID-19.
As researchers, we are all the time working to discover a stability between sensitivity (alerting people who may be sick) and specificity (the power to determine people who find themselves healthy), as there are trade-offs to each. We will continue to work with the medical and public health communities to guage totally different models for creating this know-how to ensure the optimum stability.
Our research additionally reinforces that respiration fee, resting coronary heart fee and coronary heart fee variability (HRV) are all helpful metrics for indicating onset of illness and are greatest tracked at night time, when the physique is at rest. Our analysis exhibits that HRV, which is the beat-to-beat variation of the guts, typically decreases in people who find themselves exhibiting signs of illness, while resting coronary heart fee and respiration price are sometimes elevated. In some instances, those metrics start to sign modifications almost every week before individuals reported symptoms.
Other findings embrace:
- On average, heart price variability hits its lowest point the day after symptoms are reported
- Increases in resting coronary heart fee normalize, on average, at least 5–7 days after the start of symptoms
- Breathing fee peaks sometimes on day 2 of symptoms, however there’s a slight elevation, on average, for as much as 3 weeks after symptoms start
Understanding Symptoms and Severity of COVID-19
In addition to detecting early alerts of COVID-19, we also are gleaning some insights into widespread symptoms, as well as severity, period of illness and the signs probably linked to hospitalization. Many of these observations align with what we’re hearing from different researchers and public health officers. For instance, being older, male, or having a high BMI increases the probability of severe outcomes.
In addition, our research found that shortness of breath and vomiting are the symptoms most certainly to predict that someone with COVID-19 will have to be hospitalized, while sore throat and stomachache have been the symptoms least more likely to predict the necessity for hospitalization.
We’re additionally seeing that the most typical symptom reported by people with COVID-19 was fatigue, which was present in 72 % of members reporting having COVID-19. This was adopted by headache (65 %), physique ache (63 %), decrease in taste and odor (60 %), and cough (59 %). Of word, fever was present in simply 55 % of individuals reporting COVID-19, an indicator that temperature screening alone is probably not sufficient to know who may be contaminated.
Mild instances (those who recovered at residence on their own) present a median period of 8 days, while average instances (those who recovered at residence with assist from others) final a few week longer, with a median period of 15 days. For severe instances (sufferers who find yourself requiring hospitalization), the median period of sickness was approximately 24 days. But this period had a large unfold, with a number of instances lasting longer than two months.
What’s Next for the Fitbit COVID-19 Study
It’s clear that our our bodies start to signal impacts from the illness before extra noticeable signs appear. With these initial alerts identified, we’ll continue our work in creating an algorithm to detect illnesses like COVID-19 and concentrate on expanded research in a real-world setting.
Early detection is crucial, and we hope to deliver this sort of info to shoppers as soon as attainable. As a subsequent step, we’ll continue to work with our research partners like Scripps Research Translational Institute and Stanford Healthcare Innovation Lab to further validate the know-how and intend to interact with the appropriate regulators globally to find out the perfect path to convey this to shoppers.
This info is for instructional purposes only and isn’t meant as an alternative to medical analysis or remedy. You shouldn’t use this info to diagnose or deal with a health drawback or condition. Always examine together with your physician earlier than altering your food regimen, altering your sleep habits, taking dietary supplements, or starting a brand new fitness routine.