How San Diego is Using Big Data to Improve Public Health

In San Diego, a conflict is brewing over a proposal to build a dense residential development in a rural part of the county. The developer has permission to build 110 homes, but has requested to build 1,706 homes.

The conflict pits the county’s goals for healthier urban growth—concentrating development to cut commute times, reduce greenhouse gas emissions and lower wildfire risks—against a proposal for LEED-certified buildings and a water recycling plant to service the new development.

“These developments continue to be proposed and approved with only minimal ability to evaluate their impact on health and quality of life overall,” says Kevin Patrick, director of the Center for Wireless Health at the University of California, San Diego.

Patrick intends to change that. He’s working on a project that, if successful, would give city officials the ability to forecast the consequences of a given development “like the weather,” he says.

Patrick is working with a team of data scientists, physicians, and government officials to conduct an experiment called “Big Data and a Culture of Health,” and it could be the future of public policy.

Forecasting Health

San Diego County’s public health department has been collecting “big data” on health-related issues for years. The information is available to residents via a user-friendly website.

Through a combination of research from marketing companies, mobile health apps, and their own demographic data, the county has information on everything from where in San Diego people eat the most vegetables, who watches “smart” TV versus old-fashioned cable, and where diabetes and Alzheimer’s are most prevalent.

Officials have already used these data to make positive changes, through the county’s LiveWell public health initiative. The Chula Vista elementary school district saw it had some of the highest childhood obesity rates in San Diego county, and started a program of increased physical activity and healthier meals for kids in every grade. The initiative worked: the number of overweight students has decreased five percent since 2010 .

But the “Big Data and a Culture of Health” project, led by Dr. Steven Woolf of Virginia Commonwealth University in collaboration with Patrick and San Diego county, is designed to go beyond straight-forward correlations.

This time, researchers are overlaying all the datasets together, and using new computing and data visualizing techniques to tease out correlations they weren’t looking for.

Any new public policy will have health effects, Patrick says, even a new bus route or jobs program. But it can be hard to understand, let alone predict, the effects a new transportation policy or major housing development will have. That’s what the new project is designed to help do.

“This is going to give us an opportunity to see things we hadn’t thought of,” says Dr. Leslie Ray, San Diego’s senior epidemiologist. “Once you have a disease, you’re already in my database and I can figure out a lot about you. But if I know what is correlated with the disease before you get it, that’s the phase I’m looking for.”

The data model will help her department anticipate resident’s health issues before they emerge, Ray says. “We can predict before early disease begins.”

Personal Data, Citywide Benefits

One of the project’s guiding principles of the project is that researchers will focus their efforts on on what county officials and local leaders think is important.

For example, Alzheimer’s is the third leading cause of death in the region—much higher than the rest of the country, where it’s the sixth leading cause . Dianne Jacobs, chair of the county Board of Supervisors, has made it a priority to figure out why. So investigating what factors are correlated with Alzheimer’s will be one of the project’s tasks.

As the data become more refined, citizens will have to decide how much privacy they are willing to trade for big or small health improvements. All the health and demographic information the project uses now is anonymous, and the health records aren’t linked to exactly where people live, just their region.

But in the future the doctors hope to work with health data precisely linked to where people live, and it’s easy to imagine how anonymous records could be put together to identify an individual.

Woolf knows the potential for individual profiling concerns people, but it could have significant benefits, he says.

“Everybody in the public obviously is very worried about data privacy,” he says, but “if they understand some of the benefits they might appreciate why we’re thinking about things like this.”

For example, he says, when a patient with asthma visits their doctor, the physician could pull up their electronic medical record and see not only the patient’s labs and x-rays, but also pull up how close they are to the freeway, and what resources are available in their neighborhood.

“Those kinds of capacities exist, and are part of the future,” Woolf says. Advertisers are already using them to understand your shopping habits, he points out, so why shouldn’t your doctor use them to help improve your health?

San Diego County: A Petri Dish

San Diego County has several unique attributes that Woolf and Patrick say made it the perfect location for this experiment. It’s roughly equivalent to the U.S. overall in terms of demographics, socioeconomic status, and age, he says, and it’s contained.

“We have the ocean to the west, the desert to the east, Mexico to the south and Camp Pendleton to the north, so we're kind of this buffered community,” Patrick says.

But even more importantly, according to Woolf, is the fact that the county government has already identified health goals for the whole population, invested in a community-wide public health initiative, and gathered and made available the data that makes the whole project possible.

“That kind of community investment doesn't exist everywhere,” says Woolf. Although he hopes someday, it will.