Solving Our Diabetes Problem With A System That Predicts Who Needs Help

If you can figure out who is getting sick and help them before they get to the hospital, the health benefits (and the cost savings) could be astronomical.

Solving Our Diabetes Problem With A System That Predicts Who Needs Help

It’s no secret that the U.S. health care system operates using an acute care model: The health care industry generally solves problems after they’ve become large enough to send people to the hospital, not before. That’s one of the major reasons why we spend so much on health care. It’s a problem that’s especially visible in the ever-growing population of diabetics. People often don’t adhere to treatment plans in the early, symptomless stages of the disease; it’s only when symptoms become severe that they take real action.


Every year, Sanofi U.S.’s Data Design Diabetes Innovation Challenge asks entrants to use data and design to change the diabetes landscape in the U.S. for the better. Last year, an app that senses when users are feeling upset (and thus not managing the disease correctly) took the $100,000 grand prize. This year, the prize went to the n4a Diabetes Care Center, an initiative that uses big data to hone in on undertreated portions of the diabetic population and then helps them manage their disease.

“The n4a Diabetes Care Center inspired us by incorporating big data in a highly-personalized approach to care management. In fact, it was precisely that linkage–of data leading to insights and ultimately on-the-ground action that convinced us that this team had the most potential to broadly improve care at the patient level,” explains Aneesh Chopra, a judge in the competition and the former US Chief Technology Officer.

The initiative, created as a collaboration between the National Association of Area Agencies on Aging (n4a) and the National Minority Quality Forum, is specifically targeting the 18% of the diabetic population that ends up in the hospital every year, before they get there. The key is the National Minority Quality Forum’s massive health database of 800 million patient records, organized by zip code.

“We have taken that database and built predictive modeling tools which allow us to forecast what portion of the diabetic population is likely to have an acute event,” explains Gary Puckrein, president of the National Minority Quality Forum. “We have partnered with n4a, helping them to identify at-risk populations and at community level [allowing them to] go in and do the interventions.” Factors that increase risk include age, race, and a record of prior hospitalization.

N4a’s member groups, located in almost every community around the U.S., already offer services to the aging U.S. population–things like Meals on Wheels, chronic disease management care, and personal care. The n4a Diabetes Care Center will add another layer onto those services, but targeted directly at the diabetic population. “We know that managing diabetes for anyone is a very complicated process. Even the best situations are hard to manage for the individual,” says Steve Lindstrom of n4a.

After the National Minority Quality Forum pinpoints at-risk diabetes populations, n4a will give individual assessments that examine everything from patients’ disease knowledge to depression risk. Each patient will receive a highly personalized plan. If they’re depressed, n4a might focus on increasing their social activity. If they need access to different kinds of food, n4a will work with them, and so on. “We are leveraging existing systems and adding in big data,” says Lindstrom.


N4a Diabetes Care Center is still in the early stages, looking for potential partners and advancing its business plan. Upon being asked about a potential roll-out date for the program, Lindstrom joked, “We’re waiting for the $100,000 [from the competition].” The prize money will be enough to help with the business plan, but then the program will need cash to fund a pilot program in at least one U.S. community.

Eventually, Lindstrom and Puckrein hope to turn n4a Diabetes Care Center into a private entity. “We think that the answer to many of our health care problems lies in a different model of approaching this, with the government doing some things, nonprofits doing some things, and private companies doing some things,” says Lindstrom.

About the author

Ariel Schwartz is a Senior Editor at Co.Exist. She has contributed to SF Weekly, Popular Science, Inhabitat, Greenbiz, NBC Bay Area, GOOD Magazine and more.