As the United States continues its fight against a surging COVID-19 pandemic, the population weary of mitigation measures meant to save lives, public health officials are racing to get ahead of the virus. The backbone of the public health response has been social distancing, testing, and contact tracing. And while success with these measures in the U.S. has been spotty, one Bay Area county is innovating by drawing on a rich tradition of collaboration with Stanford University. Santa Clara County’s Public Health Department and Emergency Operations Center (EOC) are rolling out new measures to inform its public health orders, reach its residents more efficiently and effectively, and help neighborhoods most at risk of COVID transmission. These initiatives are informed by a unique partnership with Stanford’s Regulation, Evaluation, and Governance Lab (RegLab) and the Stanford Future Bay Initiative (SFBI), which began in July after extensive work early in the pandemic with the City of San Jose, and is now having a real impact in the county.

“The pandemic has exposed serious weaknesses in our public health infrastructure. We are proud to partner with the county to develop data-driven insights to inform the county’s Covid response efforts,” says Dan Ho, the William Benjamin Scott and Luna M. Scott Professor of Law, professor of political science, senior fellow at the Stanford Institute for Economic and Policy Research, associate director for the Stanford Institute for Human-Centered Artificial Intelligence (HAI) and director of RegLab.

Dan Ho, along with Derek Ouyang, (BS ’13, MSE ’15), lecturer at SFBI, Jenny Suckale, assistant professor of Geophysics, and a multidisciplinary Stanford team have been helping to unlock insights from a wide range of data sources, including COVID testing data, contact tracing data, mobility data from cell phones, building records, and consumer data. The core aim is to help the county leverage its resources and build out analytical capacity for its Covid response. Efforts to date include developing a mobility data dashboard to understand the spread of the disease within the county, reducing language barriers for contact tracing, and a newly launched door-to-door COVID-19 testing project in key high-transmission neighborhoods.

“Contact tracing is only one part of the work we do in terms of identifying what we think is driving transmission. Another critical element is analyzing large datasets that can inform our policies and allow us to better utilize our resources,” explains Greta Hansen, chief assistant county counsel and co-director of the EOC. “For instance, when we have modified certain policies, where do we see the greatest change in people’s mobility? What policy interventions seem to really change how much contact people are having with others in the community?”

The projects are about more than data, though. “The partnership helps us surface the most important problems to help the county design, implement, and evaluate the most important interventions to address community spread,” says Ho. 

A Community Partnership

While mobility data may alert public health officials to an approaching spike in COVID-19, responding to that surge is the next order of business. In one project, the team helped to develop an algorithm to match contact tracers to individuals diagnosed with COVID based on language because they found that language barriers were impeding contact tracing.

Another project led by Dr. Analilia Garcia, the community engagement and racial equity lead at the county’s EOC, is bringing community-based health workers directly to areas with high levels of COVID-19. To do this, the county has partnered with teams of multilingual promotores de salud, local health workers who have roots in key neighborhoods, to conduct door-to-door testing. The teams also coordinate high-touch support services to help residents navigate the complexities of self-isolation—from concerns related to food and shelter to employment issues.

The Stanford team is using data-driven techniques to identify the most effective assignment of promotores and a process for updating the system as case reports come in. This system offers a powerful model for how to move testing and contact tracing in a more data- and community-driven direction, which will be crucial during the months ahead. “If we can rigorously demonstrate this in East San Jose, we can build public confidence in such approaches across the Bay Area and beyond,” says Ouyang.

Data-Driven, but More

Triaging and linking existing data held in siloed departments—such as the geolocation of addresses in the IT department, the size of a building in the tax assessor’s office, and population information in the Registrar of Voters—can help form a more comprehensive view for policymakers.

“No one had linked those datasets before, so it was a bureaucratic and technical challenge,” says Ouyang. “Add on innovative new data sources from mobility to wastewater, careful guidance on how to interpret and leverage these new data with privacy and equity in mind, and new technologies for data capture and learning from community-based pilots, and I think we’ll emerge from this  with a local government that works better for our communities.”

“Among our biggest challenges is the incredible amount of information that needs to be analyzed and brought together to inform our response. And this work has been really essential,” says Hansen. “Where we have to make big policy decisions, we’re making them with the very best data possible.”