The Knight News Challenge accelerates media innovation by funding breakthrough ideas in news and information.

Challenge brief

How can we harness data and information for the health of communities?

105 contributions
626 entries
626 final entries
39 final entries
39 final entries

The Knight News Challenge accelerates media innovation by funding breakthrough ideas in news and information.

Congratulations to all our contributors!


Seven projects that harness the power of data and information for the health of communities will receive more than $2 million as winners of the Knight News Challenge: Health.  Knight Foundation made the announcement at the Clinton Health Matters conference in La Quinta, Calif. Find out more at

Knight News Challenge: Health from Knight Foundation on Vimeo.

In our work, we’ve seen time and again that the most successful projects are those that take abstract concepts (“open government,” “community information,” “engagement,”) and make them tangible, relevant and useful to real live humans in our communities. That involves understanding how people are behaving, what they’re interested in, what their constraints are. It involves designing our work to meet people where they are, rather than where we wish they’d be.

Health—as concept, as systems, as a goal—is something every one of us touches and thinks about, over and over, for our entire lives. Few things reach that sort of ubiquity and consequence. That’s compounded now, in the age of mobile and data and consumer technology, by a revolution in what we know about our health and the health of the communities around us. From wristbands that track our sleep to electronic medical records to giant datasets released by the Department of Health and Human Services, we have access to more information that we’ve ever had before.

What can we do with all that knowledge to make our communities better informed about health? We’re doing the Knight News Challenge on this theme because we’re interested in the answer.

This contest is an opportunity to accelerate promising ideas and trends. Our definitions of “health data” and “news” are broad, and range from projects in traditional newsrooms to consumer-facing technology to crunching big datasets. We’re hoping to find and accelerate projects that use data and public information in innovative ways to create strong information flows about health in our communities.

For questions and more info, contact Christopher Sopher @cksopher or at cksopher[at] .


Join the conversation:

Photo of Ronald

We wanted to measure the prevalence of chronic illnesses (diabetes, heart disease, etc.) at the county level in order to identify spatial inequalities. National surveys (NHIS, BRFSS) do not drill down to the county level, especially in rural counties in MS. We partnered with IMS Health and purchased (via funding from the U.S. Office of Rural Health Policy) Rx-filled at the county level. The hypothesis was that if someone was taking diabetes meds, they probably had diabetes. We rolled the county numbers up to state totals and got correlations ranging from .44 to .78 depending on the illness and the year. More importantly we were able to drill down and show at the county-level variations in chronic disease R-x fill rates - a proxy measure for chronic illness prevalence rates in the population. See our article in Population Health Metrics at
With additional funding we want to extend this research in two ways. First, we want to use the "on-the-ground" monthly data to help correct for survey "drift" in national surveys. For example, BRFSS is administered at the state level, but no calls are made to area codes outside that state. Surveyors are missing the entire population with a nonlocal cell only. Dual frame (Land line and cell) are not only expensive, they still miss the "immigrant" telephone number. This means that, in the extreme case of Starkville MS (home to MSU), telephone surveyors will miss up to 90% of young adults (manuscript currently under review at Survey Practice). We would partner with federal and state agencies to help measure and correct for this drift using the IMS Health data.
The second major initiative would be to model the life course of drug treatment. There are a dozen steps between an individual thinking that they may have an illness that can/should be treated and fully realizing a drug treatment regimen.
If 70% of health care expenditures are due to chronic and preventable diseases and injury (Fries et al. 1993), we can do a better job of understanding chronic illness. R-x filled is a ground truthed data set that can be used to supplement surveys and expenditure data.

Photo of Ramona

Getting appropriate data on prescription drug use is critical to address patient compliance and other issues in healthcare today. However, the hypothesis that 'people taking meds for a disease means they have the disease' doesn't work for pain meds, which can be used for disorders ranging from hip replacements to cancer to recreation/addiction. And concerning the last type of use: is there a way to compensate for prescriptions filled in multiple counties?

Photo of Chris

Hi Ronald,
The submission phase for the challenge will open Sept. 3-17. Please be sure to submit your idea for funding during that time at here:

Photo of Jonathan

I would also say that it's important to note limitations to IMS data. particularly related to more novel products, inpatient distribution, and specialty channels that are increasingly used and may not be captured with the same acuity as traditional Rx's

Photo of david

I like the idea of the challenge and see the need for it but the challenge, of course, begs the question why aren't we on the right path. We aren't on the right path because of the nature of American culture and if I am correct cultural change is needed to fundamentally change health. We provide too much resources to experts and government officials and too little resources to the people. The government and experts believe people can't take care of their health without them and therefore, like, a self-fulfilling prophesy their orientation to health has become true. Only by empowering people with the resources they need for their health may they become healthy- but experts and government in America don't want to cede power to the people when it comes to health-after all, health care wouldn't be so profitable anyomore- and government certainly doesn't want that!

Photo of David

@David, I generally agree! I hope you'll check my entry, which is a people-powered system for surfacing resources and information and getting everyone on the same page. I think it's best first applied amongst those of us serving the community (the adage "eat your own dogfood" applies here), and can then be extended to the broader community (which would involve incorporating greater and more varied means of accessibility and other changes to meet the needs of the underserved in our community, something I have at least some experience with working with I'd love your input!

Photo of david

David- I like the possibilities of Woven and you could also include vision board material as motivation can be an important part of maintaining or improving health. Poems, books, webinars,symposia,tests, supplements, diets, online videos etc could be included. There are different ways to engage a community in health and different degrees that they can be engaged. At the community level you may get some group think and this may limit the depth of engagement and limit what each individual may take away from the Woven process. Nonetheless, I believe Woven may prove helpful just as LetscureTogether does. GetHEalthyHarlem has a lot of professionals on their board-and professionals have a culture that is different from the culture of nonprofessionals and, in my estimation, professionals relate better to other professionals than nonprofessionals and professionals tend not to be populist in their orientation. As I work in Harlem and know professionals i know what they can be like. In any event, I would go to individuals in community and using dynamic system development model and adaptive coaching model thoroughly assess what individuals in community want with regard to community resources and how they wish to engage with such resources. At first, the information you may get may reflect experiential and anchoring biases and so its impirtant to do so reimagineering in that regard. Best regards for creating something of value to bring people in the community closer together and sharing together.

Photo of David

@David, great input! One way we address groupthink is by not only showing the latest and greatest as determined by the community, but also "trending" and "latest" sections as well as a search function so that users can find content regardless of populist opinion. We're also looking at community reputations or rankings of sorts, so that users can bestow individuals and organizations with a kind of karma for their contributions online and perhaps even accreditations offline, thereby making their future contributions carry more weight. As well, we're constantly changing the view of content so users are discovering more than less. It's an algorithm that we're constantly evolving, but the aim is simple: help people discover the most relevant content from the community, so that everyone can get in the know, on the same page, and involved faster. Let's continue any conversation about the Woven entry on that entry page itself. Be well!

View all comments