Megadata  is detecting nursing home outbreaks up to two weeks early. FIND OUT HOW

How Data Saves Lives In The Age Of COVID-19

Research is moving fast in the age of COVID-19, and even new methods of early detection can save lives because of the virulent nature of the virus. Learn how data is saving lives, especially in nursing homes, through careful analysis of information.


Share this infographic on your site!

How Data Saves Lives In The Age Of COVID-19

In the U.S., at least 2,300 long term care facilities have reported cases of coronavirus* — and the virus continues to spread

What We Know Now

Social Distancing Is Working

  • Comparing cities with similar population density the results of early, mandated stay at home orders are clear
    • San Francisco
      • First reported case: March 5
      • Stay at home order: March 17— 13 days later
      • Cases per 100,000: 114*
    • Boston
      • First reported case: February 1st
      • Stay at home order: March 24— 53 days later
      • Cases per 100,000: 663*

Nursing Homes: Unfortunate Front Lines

  • From March 23-30 2020, coronavirus cases among long-term care patients spiked 172%
    • Seattle, Washington: Life Care Center
      • One of the first U.S. epicenters of the outbreak
        • More than 2/3rds of residents have tested positive
        • Nearly 40 residents have died from the virus
    • As of April 24, nearly 10,000 US deaths are connected to nursing home residents or caregivers
      • In New Jersey alone, nearly 400 long-term facilities have been affected by an outbreak, resulting in more than 1,500 deaths
    • Many U.S. health experts have criticized lack of testing and data gathering — tools vital to developing new strategies to combat the virus

In 1918, the Spanish Flu wasn’t named for the origin of the pandemic  — As one of the only nations with a free press, Spanish newspapers were the chief source of public information on the disease

How Data Saves Lives

    • Worldwide, organizations are already making efforts to use emerging technologies and data science to combat the virus
      • The World Health Organization has partnered with Facebook, Microsoft, Twitter, and other large tech companies to host a coronavirus “hackathon”
      • The Rockefeller Foundation committed $20 million to the development of better data-science tools for tracking and managing the spread of the virus
    • The challenge is not a lack of data, but determining what is the most relevant data
      • What we need to know
        • Severity of the problem
        • Risk factors
        • How it spreads
        • Societal and environmental context

Why Local, Specialized Data Matters

    • If predictive models are built solely on macro data we may miss vital differences in an epidemic’s spread 
      • Data must take into account
        • Geographic location
        • Socioeconomics
        • Cultural customs and behaviors
    • In 2010, a cholera epidemic in Haiti killed 7,000 and infected half a million others 
      • It took weeks for official sources to report details — BUT on Twitter, news of the disease traveled far more quickly


Finding A Solution Start With Collecting Data

Early Warning Signs

    • UC San Francisco is studying 2,000 healthcare workers using Oura ring
      • Their immediate goal: Detect early onset symptoms of COVID-19
      • Their ultimate goal: To determine algorithm to predict many illnesses
    • The ring tracks symptoms through the finger, including respiration rate and body temperature

You Can Help From Home — Even If You’re Not A Data Scientist

    • [email protected] lets you share your unused computing power to aid research into potential cures for coronavirus and other diseases

Tracking the Spread: Smart Thermometers — Kinsa

    • Using anonymized data to track fevers across the country, using
      • 500,000 thermometers 
      • 162,000 readings per day
    • Adapted algorithms to detect fevers that are inconsistent with typical flu spread patterns — identifying likely clusters of coronavirus
    • Based on this data, Kinsa has consistently predicted the spread of COVID-19 — accurately forecasting 2-3 weeks ahead of the CDC

Focusing On Specialized Populations

    • SECURE-IBD is a database of anonymized patient data those COVID-19 and irritable bowel disease
      • Caregivers around the world can contribute to the data collection
      • Each week participants receive an updated summary of the data
    • DarwinAI has released open source tools to help researchers and clinicians diagnose and study COVID-19 through chest x-rays
      • COVID-Net: A neural network to designed to aid diagnosis
      • COVIDx: A dataset of 5941 chest x-rays from 2839 patients
    • Nursing Home Dashboard — MegaData
      • All-inclusive dashboard for nursing home business intelligence
      • Creates a facility-wide dashboard of COVID-19 symptoms for early detection of the disease — Helping to identify problems up to 3 days sooner
        • Extracts likely symptoms from nurses’ notes — like cough or shortness of breath
        • Tracks daily temperature readings and flags high fevers
      • With the severity of symptoms so closely linked to survival, early identification of COVID infections — especially in high-risk patients — may be essential to saving Lives

        • With the right data, staff can react quickly and stop the spread of infection


A New Breakthrough in Early COVID-19 Detection?

    • Temperature checks aren’t enough, could indicate other things are wrong, and above all – ARE TOO LATE
    • Pulse Oximeter: Measuring oxygen levels twice a day shows the underlying problems before patients fully outwardly present COVID-19 symptomsAnd more importantly, save lives and contain coronavirus in its tracks

The answer to COVID-19 is in data — Are you looking in the right place?