This afternoon Facebook announced its latest coronavirus-focused efforts, which includes the rollout of three new maps to its Data for Good program aimed at tracking the potential spread of the disease. The Menlo Park company will share the maps, which are based off aggregated Facebook data, with research and public health organizations.

“You really need accurate information on population movement to understand how an infectious disease might spread,” Laura McGorman, Data for Good, told MobiHealthNews. “In this particular case we need to understand how populations are or are not moving around in response to large scale preventative measures – things like social distancing and stay at home orders.” 

The first map is a co-location map, which is aimed at figuring out the probability that different Facebook users will come into contact with each other in person.

“These are maps that estimate the probability that different populations will come in contact with each other,” McGorman said. “So, if you want to understand the extent in which the coronavirus might spread from New York City to the suburbs of New Jersey, understanding the rate of co-location or populations mixing between those two groups is a critical insight.” 

The second map is called a “Movement Range Map.” This map will show if people are staying near home on a county-by-county level, and it also shows if people are still moving a lot. 

“If you take a look at that, you may be better able to understand if communities that still have high rates of essential workers, or you might be able to take a look at the extent to which people are still having to travel long distances to go to the pharmacy or grocery store. So, we’ve created insights and a movement range map to answer those questions,” McGorman said. 

The last map, called the “Social Connectedness Map” is focused on insights around Facebook connections (i.e. Facebook friends) across geographic lines. The company said this could be key to understanding where the disease could spread, and also where folks are looking for support.

“What many researchers have found is that social capital is a really important way to understand if communities are resilient after a natural disaster, as well as something like a pandemic,” McGorman said. “So, what we are really hoping is that, that data set in particular is informative for thinking about from an economic perspective how communities recover from this pandemic.” 

The maps are made up of various Facebook data points. McGorman stressed that all the data is aggregated, and users' data will remain private throughout these efforts. 

Today, Facebook also announced the news that it would be working with Carnegie Mellon University’s Delphi Research Center. As part of the efforts, Facebook users can opt into a survey, run by the research center, about possible symptoms and ways they are responding to the pandemic. An individual’s results will not be shared with Facebook, according to a statement released by the company. 

WHY IT MATTERS 

The World Health Organization reports 1,133,758 cases of the coronavirus worldwide. Organizations all over the world are working on ways to track and halt the spread of the disease. Facebook pitches this effort as a way for research institutions to see how different initiatives are working in curb the spread of the disease, and also to track new cases. 

“We’re partnering with universities in every region of the world who are all answering similar but, I think, different questions depending on the particular nature of the pandemic,” McGorman said. “Researchers in Italy I think are looking very closely at the extent to which stay-at-home orders are being taken up by the population, and then what does that mean as they come over the curve? I think everyone is curious about when is enough, enough, and when is it OK to relax social distancing? I think that is at the top of a lot of our researchers' minds.” 

THE LARGER TREND 

Facebook’s Data for Good has been around for over three years, but it started out looking at natural disaster. 

“We started our work in 2017 with a handful of partners and a small set of maps to try to answer the question of whether or not we could share privacy-preserving data in near real time with nonprofits and researchers who were responding to natural disasters,” McGorman said. 

The first efforts began with disaster maps and looked at how many people are still displaced from their homes. However, Gorman said the social media giant quickly identified disease tracking as a new need. 

“What we found over a year and a half of working with natural disasters was many of the real-time insights that we could generate and share in a privacy-preserving way with nonprofits had a disproportionate value during public health emergencies,” she said. 

This isn’t Facebook’s first effort in the coronavirus epidemic. In early February, the company announced it was working with a network of third-party fact-checkers to review information and remove any false information about the virus. 

With stay at home orders dominating the news cycle, many are looking for ways to help tackle the coronavirus pandemic from home. A group out of MIT is looking to remedy this with a way for various stakeholders to get involved.

Over the weekend, MIT COVID-19 Challenge: Beat the Pandemic hosted the first in a series of virtual hackathons and events aimed at designing new tools to address the virus. 

The challenge was inspired by co-organizer Alfonso Martinez's desire to get involved in fighting the coronavirus, but not knowing exactly how. 

“How can we make this time that we are spending at home useful for the crisis? How do we have an impact?” Martinez wondered to MobiHealthnews.

When news of the coronavirus started to spread, he and a friend were inspired to develop a digital solution. But this led to a whole host of questions that would require experts to answer. 

“We didn’t know what specific issues for this crisis that we should be tackling, and we didn’t have partnerships in place to scale whatever solutions we would put together to get them to market and have a tangible impact,” he said. 

Creating the challenge 

This got Martinez thinking. Now an MBA student at MIT, he had run hackathons before, and it was a way to get different stakeholders involved.  He reached out to Dr. Freddy Nguyen, the regional director of MIT Hacking medicine, and cofounders of the event, Stephanie MacConnell and Paul Cheek.  

“What we thought of was building a series of virtual events that can enable individuals that were looking to help in this crisis, all from very different backgrounds, come together and work towards the specific issues that were most pressing at this point in time,” Martinez said. “Not only to work on the specific issues that are relevant today but also work with the partners that can put those solutions [in place] and develop into process.” 

The challenge is meant to follow the disease as it progresses and respond to the medical community’s current needs over time. Traditionally, a mix of voices, including those healthcare professionals, is represented in health-tech challenges, but medical professionals are being called to the front lines of treating the disease. So, when looking at the focus of the hackathon over the weekend, organizers tapped healthcare organizations for input. 

“We [knew we] probably would not get the propensity of the healthcare workers and providers – the front-line workers who know about all the problems they are facing,” Nguyen said. “We decided to partner with a series of clinical partners from across the country … to really be our eyes and ears to find out what problems they are facing on the front lines.”

The Challenge included a number of partnerships with healthcare-focused organizations including MassGen Brigham, Johns Hopkins Medical, and MobiHealthNews’ parent company HIMSS. 

During the weekend hackathon there were 10 tracks that focused on everything from social isolation to treating patients and figuring out immunity to at home patient triaging.

“We are sorting those problems by helping synthesize and compartmentalize those problems. That way we can hand them off to the participants and really be acting around problems that are pressing right now,” Nguyen said. “The areas we ended up focusing on came from all of our clinical partners and really carried a wide range.”

New ideas created  

Each track had four winning teams that scored $500, AWS credits and a mentorship from one of the partner organizations. 

“One of the unique things about this challenge was we had people from every single time zone, we had 96 countries represented in the applicant pool, 49 states. So, the ability for folks to pull from the experiences back home or that they have seen in other cultures and other geographies was super unique,” MacConnell told MobiHealthNews. 

Many of the teams focused on tools that were effective in other situations and reapplied them to this challenge. 

“We saw a lot of teams coalescing around, 'How can I use this type of product or service that has been proven effective in other industries for this specific crisis?'” Martinez said. “One great example … was the team called Distance Domestic violence.”

He explained that this team looked at optimizing the process of assessing someone at risk for experiencing domestic violence, and then providing these individuals with resources when they may have to be quarantined with their abuser. 

The digital design is also allowing new voices to enter hackathon space. 

“There were a lot of participants where this was their first hackathon. Most hackathons or challenges are in person, so it depends on where you are located and this was virtual,” MacConnell said. “Some of the most successful teams were teams that hadn’t met each other before. They cluster around a certain idea set they thought would be really interesting to pursue and found really unique solutions based on their totally different backgrounds.”