Pandemic by the Numbers

When it comes to combatting a pandemic, data is key in knowing the enemy. The numbers of infections, deaths and recoveries have been important signposts for governments and public health officials when creating safety policies and distributing resources in the fight against COVID-19. In early March, Raphael Presberg M.S. ’18, co-founder of Nalia, a start-up specializing in data analysis, saw an opportunity to apply skills from his day job toward a citizens’ initiative to improve the availability of crucial public health information in his home country of France.

Incomplete information inhibits pandemic response

As the coronavirus raged in Italy earlier this year, it wasn’t surprising that it spread to neighboring countries — but where exactly were new cases appearing, and how quickly? In those early days, Presberg notes, reporting structures in France were not as robust as those in some other countries, which were able to plot the spread of the virus on a map. While an overall number of cases was reported, this information felt incomplete in a country that encompasses 18 administrative regions and 101 departments (territories similar to U.S. states). A major obstacle to transparency, he says, was the lack of a centralized data collection system to consolidate information from a wide variety of health agencies, such as hospitals, emergency medical services and nursing homes. Another concern was the potential violation of privacy laws if health data was released publicly.

Despite these challenges, it became clear to Presberg and many others that reliable data was essential to understanding the full extent of the COVID-19 pandemic and mounting an effective response. When scrolling through LinkedIn, he came across a post from Lior Perez, IT development manager for the meteorological service Météo France. Inspired by a health practitioner friend who expressed concern about the dearth of information on the pandemic, Perez issued a plea to data scientists to come together to create an open source data project to track COVID-19 in France.

‘A clear map for the people’

Presberg became interested in data through his studies in computer science and engineering in France, but he was able to delve deeper at Stevens, earning a master’s degree in data science thanks to a scholarship through the Jérôme Lohez 9/11 Scholarship Foundation. Founded by Dening Wu-Lohez M.S. ’97 to honor the memory of husband Jérôme Lohez M.S. ’95, a French citizen who was killed on 9/11, a goal of the scholarship foundation is to promote educational and cultural exchange and understanding among France, the U.S. and China. With the skills he developed at Castle Point, Presberg collaborated with Alexis Tuil (his co-founder at Nalia) and Perez to collect and organize data related to COVID-19.

Inspired by a similar data accessibility project in South Korea, as well as by the data map compiled by Johns Hopkins University in the U.S., the team combed regional medical press releases and newspaper articles. Calling the initiative Data Against COVID-19, they entered relevant information manually into a table on GitHub, an open source code and project sharing platform. “Basically, our goal was to provide a clear and aggregated view of the evolution of the epidemic in France by department,” explains Presberg. “And since it was evolving very fast, we asked for the data community to help us.”

Presberg’s team merged with another working group, which included Jérôme Desboeufs of Etalab, a government department that coordinates dissemination of public information. Outreach on social media led to additional growth. “We were just three guys,” Presberg says, “and in two days, we went from three people to 600.” With so many new members, Presberg, Tuil and Perez worked hard to scale up their operation and facilitate connections among collaborators using Slack, an online business communication platform.

The community’s first project was to synthesize collected data to create an interactive COVID-19 map of France. Updated daily, the tool shows the number of confirmed cases, hospitalizations, people in intensive care and those recovered. “We worked every day, every hour of every single day on this for two or three weeks,” Presberg recalls.

The results were impressive. So impressive, in fact, that the French government took notice. By the end of March, Santé Publique France (French Public Health) began collaborating with Data Against COVID-19. “They needed help,” Presberg says. “We had a few discussions with them, and then I think it was pretty clear — we were the only ones to have a clear map for the people.”

Big data, major impact

Today, the dashboard developed by Data Against COVID-19 appears on the website of the French government. Now managed through Etalab, it is regarded as an official source of COVID-19 data for the country and its territories. The map was even used as an illustrative tool on live TV by then-Prime Minister Édouard Philippe during a March 28 press conference.

Following the success of the dashboard, the self-managed Data Against COVID-19 community (which now counts more than 1,000 members) continues to field requests from doctors, hospitals and epidemiologists for other data projects related to the pandemic.

While Presberg acknowledges that one day (when a vaccine is developed) there may no longer be a need for COVID-19 data projects, he believes that the citizens’  initiative he’s organized will continue to have an impact. “We have created a community to talk, to discuss data, about how we can improve the government with data,” he says. “We have proven to the French government, and to every government, that they need to believe in digital transformation and the power of data.” 

This story is one of seven alumni features that make up “In Challenging Times, Look for the Helpers,” the cover story of the fall 2020 issue of The Stevens Indicator.