How Wearables Track And Prevent The Spread of COVID-19
Key information and trends to identify potential determinants
I want to start to this write-up by saying people diagnose diseases and infections better than machines ever can. Even with the advancement of Machine Learning(ML), Artificial Intelligence (AI) and Quantum Computing, people can help us identify this disease more than anything. However, we are all practicing the act of social distancing to separate ourselves from the pandemic, wearable technology may help us predict outcomes of the virus. To find symptoms for people that may not be able to see the details themselves.
Wearables are used by a high percentage of the public. According to Statista.com, nearly 60M Americans currently have operating at the moment and the market continues to grow.
I first caught wind of the idea from the Scripps Research Translational Institute who are launching a project called DETECT, that gathers data from smartwatches and activity trackers to analyze how activity, heart rate, sleep patterns and other health data can be used to track COVID-19 information through their MyDataHelp app.
Previously this information has been used by Apple Health, Fitbit, Garmin, WHOOP and other boutique brands to correlate trends of common seasonal respiratory infected flu cases. The question is, can these devices help track the outbreak of the COVID-19 case?
Currently symptoms for identifying COVID-19 are a fever, tiredness/fatigue, dry cough and in some rare cases, difficulty breathing.
Wearable and smartwatch technology have an array of trackers that can detect heart rates and even specific motions. The key functions of wearable tracking that can help find COVID-19 symptoms are:
Respiration Tracking: Abdomen Strain Levels, Respiratory Rate (RR), Tidal Volume (TV)
Heart Rate Tracking: Resting Heart Rate (RHR), Heart Rate Variability (HRV),
Temperature Tracking: Body Temperature, Skin Temperature, Room Temperature, Heat Loss (UV Tracking), Oxygen Level Variations
Sleep Tracking: Rapid Eye Movement (REM), Deep Sleep, Light Sleep, Awake
Distance Tracking: Steps, Floors
Location Tracking: Global Positioning System (GPS), Barometer Levels
Variables at Play
We have all of these trackers, but what do they mean? Here are some trends that can identify the virus:
Increase in Respiratory Measurements: When using a tracker that is connected to 8-10th region of the abdomen (right below the left side of the pectoral muscle), one can assess proper resting and exercise heart rates. The ideal resting breathing rate is 6-8 breaths/min for an average adult and the average teen breathes 15-20/min. Abnormal increases in flow can become a potential symptom.
Increase in Resting Heart Rate: HRV decreases abruptly as people get older. 50% of 20-25 year olds usually have an average HRV in the 55-105 range, while 60-65 year olds tend to be between 25-45.
For our standard heart rates, an average resting heart rate can be anywhere from 54 beats/min (BPM) to 80, depending on health.
If a sudden decrease of HRV is detected and an increase an RHR spikes. These are signs of potential symptoms.
Increase in Temperature & Variances of Oxygen Levels: For more pricey and advanced tracking models, your wearable can act as a thermometer at the skin level and detect with some accuracy the temperature of your body. Giving your body a heat check, compared to the given room temperature can predict accuracy of a fever occurring.
If you happen to be active outside, it can combine that information with UV levels of the sun so that your skin temperature will not be conflicted with your exposure to the sun.
Through Fitbit in particular they have an SpO2 sensor that takes in a reading of the oxygen levels in your blood using a red LED in the optical heart rate monitor. Examining the difference in the highs and lows frequencies of your blood oxygen.
An increase in the temperature of your skin, a high or lower variance score with your oxygen levels, relative to your given environment, can detect symptoms related to COVID-19 as well.
Sleep Score Inequalities: One can reasonably say that a build up of mucus can correlate to a lack of sleep, but so can less REM and Deep Sleep statistics. As a body must be still and relaxed in order to have those precious conditions from sleep.
The intake of drugs, alcohol or tobacco is another variant that can mess with those numbers, but for a healthy and normally restful sleeper, this information can be categorized as a commonality to upper respiratory issues.
Less Distance/Location Tracking Numbers: Not the best measurement due to our country-wide social distancing measures of staying in one spot. But we can assume that participants who are normally very active would still find workout routines given quarantine guidelines. Along with different areas a device is being carried to locationally, this variable can be considered minor.
What To Do With This Information
The best resources to use with the information are to create data science modeling tools so that our health care, federal and public service partners can access the information and help prevent the spread.
I suggest using the epidemiological SIR (Susceptible, Infected, Recovered or alternatively Removed) model, Contact tracing model or traditional mapping models to give a visualization representation of the data.
Could This Work?
Healthcare providers have low confidence levels that the devices give an accurate representation of vital signs, but these tools in my opinion are best used for predictive purposes, rather than determining immediate outcomes. There are a lot of things that can go wrong with the devices themselves like an incorrect bluetooth signal, inappropriate placement of the wearable on body and people manually entering in false information.
Given this rich data, we can determine a percentage of Americans who currently have symptoms, who are recovered and who may not have had any respiratory illness signs at all.
Identifying and predicting outcomes though will be able to give the American public more peace of mind while at home.
____________________________________________________________________________
If you found this newsletter valuable please subscribe to this page!
If you’re already a follower of my updates, I would appreciate it if you share this with people who may find my posts valuable. Feel free to reply to this email with comments, questions, or feedback.