Positive-deviant Health Data Journalism

We partner with newsrooms and data organizations to advance investigative techniques to surface new and different types of data to identify stories that are essential for advancing public understanding about improving health systems.

Photo of David Bornstein
11 10

Written by

Using Data to Tell the Whole Story
The reigning myth of journalism is that it is a reporter’s job to uncover problems.   Given a database, a journalist will typically hone in on the negative outliers, the worst performers, and pounce.  But this model of journalism gives a misleading picture of the world around us.   Journalism is more credible and engaging -- and society is stronger – when reporters also rigorously consider responses to problems and what we can learn from them.  
At the Solutions Journalism Network, we seek to legitimize and spread the practice of reporting on responses to problems.  Our organization was co-founded by David Bornstein, Tina Rosenberg and Courtney Martin -- three writers of the New York Times Fixes column, which looks each week at a different solution to a social problem.
To do this, we work with newsrooms around the country, training reporters and editors to integrate rigorous reporting about potential solutions into their coverage. We run story funds that catalyze such stories (specific topics have included a focus on climate change resilience and social and emotional learning). We are a few months into an 18-month partnership with the Seattle Times to produce an extended, in-depth series of stories on education, and we’re developing similarly intensive collaborations with other news organizations.
We also have partnered with the Institute for Health Metrics and Evaluation, which produced the comprehensive Global Burden of Disease (GBD) report in 2010. The GBD contains “comparative metrics for 291 different causes of premature death and disability across 187 countries, 20 age groups, and both sexes for three time periods: 1990, 2005, and 2010.”  The GBD data also has detailed breakdowns of U.S. data, so we can learn about how cities, counties and states rank in terms of risk behavior, wellness, disease and causes of death. 
Our collaboration with IHME highlights instances of positive deviance in the GBD data. To date, we have identified a wide range of positive deviant data points – story leads, which we have described on our website in a regular series. Each week, we uncover a distinctive – and often surprising or counter-intuitive -- health result; suggest possible explanations; and show the opportunity for journalists to explore this story at greater length.
This collaboration represents an important frame change, using investigative techniques to surface new and (for journalists) significantly different types of evidence to identify stories that better equip society to grapple with important health challenges. What’s needed now is to embed this practice in newsrooms, giving journalists access to positive-deviance intelligence and the tools to navigate it in ways that catalyze solutions-oriented health reporting.
Funding from the Knight News Challenge will support our efforts to build out and sustain collaborations with IHME and other credible data providers. More important, it will support partnerships with news organizations that provide the foundation for robust, data-driven journalism on health solutions.

Positive Deviance Analysis Health Reporting
The Solutions Journalism Network intends to establish positive-deviance health reporting in at least five newsrooms in the next 15 months. We are in discussions with a number of news organization to pilot this program and would welcome the opportunity to work with the Knight Foundation to identify ideal partners across its network in online, print, and broadcast media.  
For each newsroom partnership, Solutions Journalism Network will run a dedicated orientation for health reporters and editors (and other interested staff) and help frame an editorial strategy that incorporates positive health data from IHME and other sources. We will orient journalists about how to get at this data; how to identify positive outliers; how to form and test hypotheses suggested by the data; and how to craft richly reported stories that draw on instructive examples. As the reporting commences, we will provide an expert journalist-mentor to guide newsroom staff to relevant data, sources, and models that can inform stories.
We expect that this strategy will produce three key outcomes:
  1. Change the practice of health journalists in partner newsrooms. Reporters will be more likely to find and use positive-deviant data as a foundation for authoritative solutions-oriented stories. These reporters will, we hope, join our learning community and re-seed their practices with other journalists.
  2. The stories produced by journalists at partner newsrooms will influence public discourse around health systems, leading to stronger citizen engagement in problem solving. The stories also will serve as models for other journalists, reinforcing the spread of the practice change. 
  3. We will work with health data organizations to help them make positive deviance data more accessible – and appealing – to journalists. 

Problems Scream, Solutions Whisper
Why is this approach needed? Good solutions journalism is simply good journalism: reporting on what’s going on and what effects have emerged.  However, since problems scream and solutions whisper, there is always more attention directed to the former.  Journalists need to become more proactive, skilled and comfortable reporting on responses to rebalance the view of the world they produce. 

There is no need to overclaim, to pick winners or to speculate.  But journalists often hesitate to write about solutions.  They are afraid that such a story would be labeled advocacy or PR; they are particularly afraid that they will write a story calling a program a success and turn out to be wrong.  For reporters, excessive cynicism is a misdemeanor, but excessive gullibility is a felony. 
But using databases as the foundation for solutions stories removes much of the risk. The data show that according to an independent evaluation, something is working – the journalist need only investigate why and how.  
Here are some examples of solutions-oriented stories using health data.

What's Happening in Kentucky?
1) The IHME newly released US health data includes information on obesity trends. When constructing a color-coded map of the United States about trends in “sufficient physical activity” between 2001 and 2011, they found a striking sea of blue and purple in Kentucky. 

What accounts for Kentucky’s marked improvement? The state has invested heavily in supporting physical education.  Schools have started placing a higher priority on phys ed.  Or the improvement may be in part due to Kentucky’s low starting point.  
Is Kentucky’s experience meaningful? Could something be learned from the state’s experience? These are questions that beg for a journalist to dig in.

What's Happening in Russell County, Alabama?
2) Scattered across the United States are counties that have seen big improvements in life expectancy rates for women. IHME found, for example, that Russell County, Alabama made big improvements between 1985 and 2011.
In Russell County, life expectancy for women jumped from 72.4 years in 1985 to 76.1 years in 2010, still lower than the state average, but much improved.
What’s going on? Investigation is needed to pinpoint the causes, but some hunting into the 2011 cause of death statistics for the county reveals that, in Russell, heart disease (178 per 100,000 people) is well below the rest of Alabama (250.2 per 100,000). Interesting. So now what’s the cause? That’s where the journalist is needed. This story is begging to be explained.

A Tale of Two (Very Different) Good Samaritans
3) New York State collects information about births in hospitals.  Journalists looking for an expose might look at data on Cesarean births at Good Samaritan Medical Center in West Islip, Long Island:

Good Samaritan has a C-section rate of 49.4 – one of the highest in the state.  A journalist at a Long Island or New York outlet might dig in to see why. Is someone profiting? Are doctors opting for convenience over patient benefits?
But there’s another Good Samaritan – Good Samaritan Regional Medical Center in Suffern, also just outside New York City. Its data on Cesarean births looks quite different:     

This Good Samaritan’s rate is 18.4 percent. Now the investigation story would be stronger – and far more interesting -- if the reporter also visited this Good Samaritan and found out why its C-section rate is so low. What is this hospital doing that the other one is not?  How did the staff do it?  What battles did they fight to get there?  And, of course, are there problems with what Suffern is doing?  What can other hospitals learn from this?
Using data in this fashion to identify and juxtapose different performance points on a spectrum can provide a fresh way to frame issues – highlighting the performance range, the known responses, and the options available to address a problem. This is enormously empowering and admits for a much more robust engagement potential. 

Who is working on the project? Who are your partners? 
Solutions Journalism Network co-founder Tina Rosenberg, director of this project, is a Pulitzer Prize-winning journalist who co-authors the “Fixes” column in The New York Times “Opinionator” section. SJN’s Sarika Bansal, who manages our partnership with IHME and other data providers, is a journalist who specializes in social innovation and global health.

The Institute for Health Metrics & Evaluation (http://www.healthmetricsandevaluation.org/), our founding partner in this initiative, is an independent global health research center at the University of Washington that provides rigorous and comparable measurement of the world's most important health problems and evaluates the strategies used to address them.

McClatchy Newspapers, ProPublica, The Center for Investigative Reporting, PRI’s “The World,” and National Public Radio have indicated their interest in becoming early-stage newsroom partners in this initiative.
How do you know there is demand for this project? 
More than two dozen newsrooms have asked for SJN’s help in making solutions journalism a part of their reporting. Our collaborations range from a brown-bag lunch with assembled staff, to year-long intensive projects, to having a solutions journalism mentor embedded in the newsroom.   
It is well established that newsrooms are interested in using data in their reporting. It is also clear that readers – and policymakers – want information about what works in health care. Readers are desperate to know: Which hospitals are safe?  What’s an effective program for increasing exercise? Governments, insurance companies and health care practitioners want to know:  What are good models for reducing hospital re-admission? How have hospitals successfully reduced MRSA transmission?   
The answers to all these questions are buried in health databases – most often found by looking to the positive deviants.  We plan to help journalists find them and turn them into stories that will make Americans healthier.
How is your project different from what already exists?
We propose a new way of thinking about covering health and health care. People have come up with solutions – with varying degrees of success – to almost every problem in health care.  Those responses can be found in databases, by looking for the positive outliers.
But despite many efforts to get journalists to use data more effectively, few daily reporters think about searching for positive deviants (many still don’t think about using databases at all). One problem is that most databases are too difficult-to-use to be of help, especially for short deadline reporting. We aim to change the way journalists cover health with custom guidance in using databases to give them the information they want. The stories they produce will be models for other journalists, as well, highlighting the distinctive elements of solutions journalism and the new public engagement possibilities it offers.
How will the data or information you use or create be made open? 
Our project will help newsrooms turn health data into stories – open by definition. These stories will translate data into news that readers can engage with.  They will publicize and link to the databases and research that uses them.

Beyond seeding specific reporting projects, the Solutions Journalism Network seeks to teach a model of doing journalism that will spread and last.  We plan to use these projects as ways to build muscle memory into newsrooms that will make this kind of reporting a reflex in the future.

In this initiative, we want to codify two processes in ways that make them easily replicated. First, based on our work with IHME, we want to model a “concierge” function at data providers, showing them how best to connect reporters to positive-deviant data. Second, we want to codify best practices in newsrooms: How do reporters and editors find and analyze this data; and how do they deploy the analysis to develop stories that meet newsroom needs?
What will you make or do in this project? 
Over 15 months, SJN will work with both health data providers and newsrooms to seed collaborations that lead to productive use of positive-deviant data in health reporting.

We will build on our existing partnership with IHME to define and codify a “concierge” function charged with making data easily accessible and with creating a human interface to guide journalists into and through that data.
At each of three to five leading news organizations, we will train health reporters and editors (and other interested staff) in the positive-deviant data approach and help frame an editorial strategy that incorporates health data from IHME and other sources. As the reporting commences, we will provide expert journalist-mentors to guide newsroom staff to relevant data, sources, and models that can inform stories. Importantly, these mentors also will harvest intelligence about best data mining and reporting practices that will inform SJN’s codification of these processes for application in other newsrooms.

To catalyze high-quality reporting based on positive-deviant data, SJN will create a story fund that makes make grants of up to $2,500 to participating newsrooms to support travel and related expenses for reporting projects that promise to leverage positive-deviant data in extraordinary ways.

How can others learn from/build on what you do?
Our goal for this project is to make routine a sort of journalism that’s particularly well aligned with the goal of this challenge – reporting on outcomes and practices that promise to improve the health of a community. In its early stages, we’ll realize small victories as newsroom reporters discover that seizing on positive-deviant health data helps them see the world in a new way that leads to distinctive and high-impact health reporting.

Our “home run” will emerge as those journalists begin sharing their successes and best practices, learning from each other’s experiences. That’s the bigger idea driving the Solutions Journalism Network: We aspire to become a network of working journalists who collaborate to iteratively advance and share intelligence, spreading adoption of solutions journalism practices.
How much do you think it will cost?
How would you use News Challenge funds? 
News Challenge funds will support three streams of work:

1. Codification of data providers’ role:
SJN staff: $40,000
IHME staff: $25,000
2. Newsroom training and codification:
SJN staff: $25,000
Contract newsroom mentors: $25,000
SJN staff/mentor travel: $15,000
3. Newsroom partner reporting grants:
Up to $2,500 per grant: $25,000
Administration and overhead: $15,750

Describe your project in one sentence.

We will collaborate with newsrooms and data providers to mine health databases for evidence of positive outliers – hospitals with the lowest rates of infection, cities that are healthier than their demographic twins – that fuels authoritative stories about how systems can be made to work better.

Who is the audience for this project? How does it meet their needs?

It has three audiences:
-Newsroom journalists in partner newsrooms will benefit by gaining an analytic method of identifying under-reported, yet highly instructive, stories relating to public health innovation
-The general public will benefit from the reporting and a view into a much broader range of response and ideas of constructive value to public health challenges
-Journalists will benefit more broadly as we systematize and make available the positive deviance analysis methodology honed through partnerships.

What does success look like?

Five newsroom partnerships in which journalists have adopted the positive deviant analysis method, and are successfully and systematically using data sets, to identify and produce rigorous solutions-oriented journalism that is helping communities to grapple with and engage around key health challenges.

The positive deviance analysis lends legitimacy and practicality to the imperative among other news organizations to report systematically and rigorously about responses to social problems. It becomes a simple and doable prospect that enables this pattern to spread more broadly in the field of journalism.

Your Location

New York, NY, USA


Join the conversation:

Photo of Jared

I really like this model and hope that you're successful in implementing it. Your focus on positive deviants - backed by data - should create enough incentive for journalists to report on the positive aspects of health care and help find solutions, and in effect, help to change the paradigm of reporting. As you put it, "excessive cynicism is a misdemeanor, excessive gullibility is a felony" rings true throughout virtually every news outlet. Hopefully this method will spread to other areas of reporting, not just in health care. Congratulations on winning the challenge.

View all comments