Why we built “Covid-19 Near Me”: A prototype for using mobile app data to encourage social distancing

Social distancing is a simple idea: we all minimize direct contact with each other in order to slow the spread of the coronavirus that causes Covid-19.

Why is this important? Because the virus spreads at an exponential rate. One person gives it to two, two give it to four, four give it to eight and before we know it, tens of thousands of people have it. If we do nothing, tens of millions of Americans will get the disease. Even a tiny percentage of people needing hospital care will overwhelm our healthcare system. If we practice “social distancing” (meaning, we avoid direct contact with people as much possible) we can slow the spread and keep more people from getting sick, resulting in less people who need to go to the hospital.

This is called “flattening the curve” and can help keep the number of cases below the capacity of our healthcare system.

This gif illustrates the concept nicely.

In South Korea, authorities have been able to control the outbreak and flatten their curve without resorting to draconian quarantine efforts (unlike China). Just compare the rates of new cases each day in South Korea to those in Italy:

As you look at the above histograms, don’t just look at the shape of the curve. Look at the magnitude. South Korea’s peak was 851 on March 4 and is currently only about 60 new cases a day. Italy’s peak was two days ago with 6,557 new cases. Italy is still recording thousands of new cases each day.

Unfortunately, new cases translate directly to more new deaths and less new recoveries. And since Italy’s hospital system is now overwhelmed, they can’t provide the care to get people healthy again.

These charts below show that the number of people in Italy who have recovered has remained flat while the number of new cases rises. This is in contrast to South Korea, where the number of new cases is falling and the number of recoveries rises. Again, pay close attention to the magnitude (number of cases) and not just the shape. Things are out of control in Italy right now.

So what has South Korea done differently than Italy? Al Jazeera lays out the story nicely. In addition to social distancing, they’ve used existing data combined with large scale testing, public health detective work, telemedicine, and hospital demand management to beat the virus.

• Data: The government used a law that allowed them to nationalize consumer data in times of emergency. In other words, the government accessed GPS data, CCTV data, social media data, and other sources in order to best identify people at risk.
• Testing: They began testing on a large scale from the beginning of the crisis.
• Public health detective work: If someone tests positive, they use the consumer data to identify those who are at the most risk and have them self-isolate.
• Telemedicine: Anyone self-isolating is given a telemedicine app to monitor them remotely.
• Demand management: Doctors can remotely decide if they need to get a monitored patient to the hospital, send a specialized team to transport them in isolation, and get them to the hospital with the most bandwidth to care for them based on their needs.

On the other hand, Italy’s large scale social distancing efforts haven’t stemmed the tide of cases. Front line healthcare workers are now left to simply manage the demand coming in (often leaving doctors with tragic decisions to make). Testing, while increasing, has not been widespread or targeted enough.

Unfortunately, the growth of the virus in the US is starting to look like Italy.

It’s still early, but we are on a dangerous path in America.

So what can we accomplish with the existing resources to encourage social distancing and support public health efforts?

A lot.

Our phones give us the ability to share where we are at any point in time. Consumers have the choice to share data with apps, and we do it everyday. Think about building a map on RunKeeper and sharing it with your friends. Or checking into a restaurant on Yelp. Or getting directions on Waze. Or writing a business recommendation on Google. Or looking at events near you on Facebook. We all do it every day.

To prove the concept, we built an app with simple functionality that will:

• let users see the number of cases near them
• let users be alerted if they’ve interacted with anyone in the last two weeks who has the virus
• help public health officials understand where hot spots are occurring
• help public health officials identify individuals who are at risk

So how does Covid-19 Near Me work? It’s simple.

• A user downloads the app and agrees to anonymously share their location.
• They select their disease status (i.e., positive or negative)
• If their disease status changes, they change it in the app.
• If a person changes their status to positive, Covid-19 Near Me will alert anyone whose phone has been at the same place at the same time as a person who has the disease.

If everyone downloads the app and selects their status, we can start building more powerful information for when large scale testing begins to roll out.

A note about privacy

We’ve built the app to use GPS data available on your device. We aren’t requiring any login or signup so you can remain completely anonymous. Other users won’t see your status and you won’t see theirs. All you’ll see is the total number of devices that have been near you that are owned by someone who has or might have the disease.

Phone GPS data can be accurate from a few feet upwards depending on the device and location. We’re not trying to locate the exact position of any individual, but instead we want to show when you were relatively close to others. It’s playing by the same rules that every other app you use that uses your location data to improve its experience.

Check out these screenshots. It may not look fancy now, but it’s simple and it works.

So what are the limitations?

Due to current testing availability, we should expect the vast majority of people to self-select that they don’t have the disease. This will tell a story that there isn’t a lot of risk. However, we know this isn’t true. People can spread the disease asymptomatically (they can give it to others even if they aren’t sick yet) and without large scale testing, most people won’t be able to be tested.

And since Covid-19 Near Me doesn’t know anything about your location before you downloaded the app, the data is limited to what you share going forward.

As health authorities roll out large scale testing, we’ll see more and more people test positive. But since Covid-19 Near Me only knows what you’ve shared with the app, people downloading the app to enter their status as positive won’t have as much of an impact as people who change their status from negative to positive. After all, if you test positive, you probably aren’t leaving your house for the next two weeks anyway!

In other words, because Covid-19 Near Me is most powerful when it can see where you were, the more people who sign up before getting tested will mean a more powerful database to help people know their risk later.

How can we overcome limitations?

Because we are playing by the existing rules of how consumer data can be used, we can only see information you share with us from the time you decide to opt-in and share it.

Right now, that means Covid-19 Near Me can’t “look back”. It can’t see where you were before you downloaded the app. It can’t access any information that you don’t give permission to share with it.

But large tech companies like Google, Facebook, Twitter, Uber, or Lyft could use this same approach to look at their data and identify people who are at risk. They could work with public health officials in the most privacy compliant ways to identify anyone who has direct exposure to people who have tested positive and use existing technology channels to contact them.

America could then do what South Korea has done. We can combine social distancing with existing data, large scale testing, public health detective work, telemedicine, and hospital demand management to beat the virus!

How can you help?

If you are a healthcare company or government agency or public health group or university or established Apple developer that wants to collaborate on saving lives, then we want to hear from you at cv19tracker@gmail.com.

We are currently stuck in a bit of a regulatory queue with Apple. They are not approving apps related to Covid-19 that aren’t submitted by relevant existing entities (i.e., healthcare companies or universities) so our submission was rejected. The app exists and we can invite people to test it, we just can’t get it live in the App Store for general consumption.

We also need help with additional software development (immediate need for React Native help), UX/UI design, hosting costs, expedited app store approval (iTunes and Play), and help getting the word out. If you can help, please reach out.

We can beat this virus if we all chip in. And we don’t need to resort to quarantines. Let’s do this!

Who are we?

We’re a group of folks who work in technology and want to give back to the greater good. We are not trying to monetize this app and have no current plans to make money off of it. We just want to help.

Tom Cummings: A business focused data enthusiast who happens to read books like Germs in his spare time. LinkedIn, Facebook, Twitter

Derek Yimoyines: Tech guy with 20 years of experience at various Software companies in the Boston Area (linkedin)

Norman Croan III: Full Stack Enthusiast, linkedin.com/in/ncroan

Special thanks to Elly Brady (LinkedIn) for the logo design, Dr. Elizabeth Martin, VMD, DACVECC an ICU veterinarian who provided scientific feedback, and Steven Sampson for help with editing this post.