During the mid-October event, participants in the challenge searched for untried solutions to wellness needs ranging from access to mental health services to the cost and causes of emergency room visits.
It all began when Indiana’s largest state agency, the Family and Social Services Administration (FSSA), approached the state’s analytics solutions provider, the Management Performance Hub (MPH), for assistance releasing data from its new warehouse. MPH and its private-sector partner chose a different approach.
States are obligated by federal and local law to closely guard patient health-care data they collect, but high levels of anonymization and aggregation can make it usable.
To ensure they were releasing data that could help technical minds identify and solve unanswered questions, staffers at MPH and Indianapolis-based KSM Consulting (KSMC) had conversations with doctors, health-care providers and organizations around the states first to learn about their needs.
MPH and KSMC then tailored 25 sets of data — an amount both organizations believe to be the largest-ever Medicaid data release — from FSSA’s new data warehouse, using a Health Insurance Portability and Accountability Act anonymization technique, expert determination, which preserves granular detail. They then released the data sets on Oct. 16, five days before the challenge.
Indiana Chief Data Officer Darshan Shah emphasized how important it was for the state to “start with the key outcomes” and partners and “work backward” to identify helpful data.
Shah is also director of MPH, which supports state agencies and the governor on data and analytics but also spearheads the release of public data. MPH transitioned out of the Office of Management & Budget on July 1 to become a separate state agency.
Shah and FSSA CIO Jared Linder credited FSSA Secretary Jennifer Walthall with championing the value and release of medical data collected since its warehouse opened in 2013. Shah described MPH’s role as the “facilitator bringing together partners and agencies to tackle broad-based problems.”
“You’re looking to illuminate a problem, you’re looking to create a potential path for success. The idea here is to create ideas and generate next steps out of it,” Shah said.
Linder, whose agency administers programs including Medicaid for Indiana residents, described himself as a believer in open data, and said the state’s data release set a standard.
“The beautiful part about this was the model we created, about how you go to the market and ask them what they want and produce results. All that, conceptually, now it’s a portable package for (MPH) to pick up and apply to your next data source,” he said.
Attendance at the challenge on Oct. 21 was strong, with more than 150 participants in 30 teams. The timing — one day before the opening of the 2017 Midwest Fall Technology Conference of the Healthcare Information and Management Systems Society (HIMSS) Indiana Chapter — was strategic.
The state is home to pharmaceutical giant Eli Lilly and Co., headquartered in Indianapolis; and Warsaw, Ind., which likely has the largest concentration of medical device manufacturers in the world.
“We thought it was important to get the message out to all of those people that Indiana does view the data it collects and maintains on behalf of the citizens,” said John Roach, director of data analytics at KSM Consulting, which worked with providers to focus on particularly relevant data, then prepare and anonymize it for the state’s release.
“Really, those are the people that are going to move the needle with respect to Indiana’s health,” Roach added.
Officials announced challenge winners at the HIMSS conference on Oct. 23 and led a more detailed panel discussion about the event on Oct. 24, the conference’s final day.
More than 20 teams ultimately presented their finished ideas at the challenge. Their ranks included veteran health-care executives as well as graduate and undergraduate students from Indiana and neighboring states.
Judges awarded $1,000 prizes to members of the OurHealth team of professionals for their data visualization of emergency room visits that could have been better and more cheaply handled in primary care; and to the Random Variables team of grad students at Indiana University’s School of Informatics for their data analysis of populations that would benefit from better access to mental health services.
The four members of OurHealth, an Indianapolis company that works with employers and municipal governments to provide on-site and nearby employee wellness and acute care clinics, used a New York University algorithm and a classification system from the Agency for Healthcare Research and Quality to identify more than $150 million worth of emergency department visits over five years that would have been better served by primary care.
The team identified another $100 million worth of emergency room visits that could have been served by non-emergency acute care.
Brian Norris, vice president of data and analytics at OurHealth, said data challenges and hackathons are becoming “pretty common” in Indianapolis, but praised FSSA and MPH for the size and detail of their data release, along with FSSA, KSMC and the Indiana HIMSS chapter for partnering in the challenge. The program created relationships that otherwise wouldn't have existed, he said.
“It’s pretty unprecedented for a state to do this, and so I think it now creates this culture. I think from a cultural standpoint, it now sets a tone that there’s this partnership between private and public entities that are leveraging these assets,” Norris said.
Runners-up looked at how demographic and employment changes impact Medicaid spending and how geography, demographics and diagnosis drive emergency room visits.
The KSMC consultant said it’s still early to assess the challenge’s ultimate value to the state, and characterized organizations as in the “analysis or the planning stages” of ingesting the data released, understanding its meaning and “making organizational plans around it.”
Similarly, he and the CDO said it's rare, and not expected, that events like the challenge will quickly yield finished applications or solutions. That’s not their aim.
“You’re not going to do this over 24 hours and have an app that will change the world. The goal here is to be able to let this be a feeder,” Shah said.
However, there is at least one group putting the data to practical use right now. One critical care hospital in the southeastern part of the state is using the Medicaid data to identify patient concentrations as well as the services that high-use, high-cost populations use.
“They were looking to make an expensive capital decision — they’re going to build brick-and-mortar, a new facility — but they didn’t know exactly where to place it. This additional data is going to help them make sure they can make that decision,” he said.
