Harvard COVID-19 Researchers to Use Facebook Data for Disease Forecasting

Harvard COVID-19 researchers have partnered with Facebook to implement user location data in order to better inform disease forecasting models, according to an April 6 Facebook announcement.

Facebook’s Data for Good program, which offers a selection of de-identified user data to nonprofits, released a series of “Disease Prevention Maps” to researchers.

Both the maps and the data used to make the maps were shared with the COVID-19 Mobility Network, a collective of infectious disease epidemiologists at over a dozen universities, to analyze aspects of the disease.

The Harvard School of Public Health and affiliates at other Harvard graduate schools are listed as network partners on the program’s website.

The Disease Prevention Maps seek to provide new tools for visualizing probabilistic contact tracing — a common epidemiological practice for tracking the person-to-person spread of disease — and overall population movements. Some of the maps are also available online for public viewing.

Other released indicators include a “social connectedness index” to identify “friendships across states and countries” to forecast how the disease might spread next or to whom the infected might reach out for medical help.

“Mobility data from Facebook’s Data for Good program provides a near real-time view of important correlates of disease transmission,” Daniel J. Klein of the Institute for Disease Modeling said in the Facebook press release. “This data, in combination with other sources, allows us to make better models to inform public health decisions.”

School of Public Health Professor of Epidemiology Caroline O. Buckee emphasized the importance of timeliness when creating these models.

“Measuring the impact of social distancing policies is absolutely critical at this stage,” Buckee, who also co-leads the COVID-19 Mobility Network, said in the Facebook announcement. “Aggregated data of this kind provides insights that protect individual privacy but are actionable for policymakers and researchers building predictive models.”

Researchers across the country are using de-identified health and medical data to analyze COVID-19 characteristics.

On Thursday, the Broad Institute announced a consortium alongside Partners HealthCare, Biogen, and Boston hospitals to create a “COVID-19 biobank” of patient data to study coronavirus symptomology.

Eric S. Lander, the president and founding director of the Broad Institute, spoke about the benefits and importance of data collection in combating COVID-19 in a press release published on the institute’s website.

“Patients who have volunteered to donate data to accelerate the shared understanding of the disease play a crucial role in the global effort to overcome COVID-19,” he said. “Through a shared biobank, researchers will be able to identify new patterns and drastically expand our knowledge of a disease.”

—Staff writer Simon J. Levien can be reached at simon.levien@thecrimson.com. Follow him on Twitter @simonjlevien.

—Staff writer Austin W. Li can be reached at austin.li@thecrimson.com. Follow him on Twitter @austinwli.