Hospitals use big-data platform to improve care

In treating diabetic patients at the Cleveland Clinic for many years, internist Dr. Anil Jain saw many with concurrent conditions such as hypertension, high cholesterol and depression. He was frustrated that there was no rapid way to collect aggregated data from the clinic's entire patient population to analyze the interplay of diabetic patients' various conditions and treatments and determine how to achieve best outcomes.

For instance, with depressed diabetic patients, he wanted to compare hospitalization rates between those taking antidepressants and those who were not to determine if more patients should receive psychiatric treatment to help them manage their health. But his system had no way to do that. “It's hard when we're not able to bring that together,” he said. “I was thinking there had to be a better way.”

With providers increasingly expected to manage the health of their enrolled populations, clinical leaders want software that allows them to quickly analyze population health through the use of big data and determine how they can improve outcomes and reduce costs.

Jain presented a proposal to Dr. C. Martin Harris, the Cleveland Clinic's chief information officer, to create analytics software that would dig deep into clinical, claims and billing data, device metrics, patient-generated information and other sources to create a rich analytics platform. After piloting a version of the software at the clinic, Jain and his colleagues worked with Cleveland Clinic Innovations, the system's technology development arm, to expand and commercialize the product, called Explorys. In 2009, it was spun off by the Cleveland Clinic as a separate company.

Explorys collects and organizes data from hundreds of sources and gives providers and researchers tools and applications that allow them to comb through the data to identify patient risk factors, track outcomes and evaluate treatment success. It digs deep into patient clinical data contained in any type of electronic health-record system as well as other sources. It has been used to speed up medical research.

The Explorys dashboard offers near real-time reports on a number of quality metrics the firm customizes for each health system. It can analyze a health system's entire patient population, subgroups with specific conditions, or even the patients of a specific physician. Four Web-based applications allow users to analyze relationships between diagnoses and treatments, establish benchmarks and quality scorecards, create data-driven patient population lists and manage providers' workflows.

Explorys has nearly two dozen health system clients representing about 360 hospitals and 55 million patients. Its software allows customers to analyze de-identified data from their own and other systems to benchmark their performance and search for insights about other treatment models. Together with its clients, Explorys has built a clinical data base that includes de-identified information on more than 50 million people.

Health systems that subscribe to the Explorys software include Mercy Health, Cincinnati; Baylor Scott & White Health, Dallas; Adventist Health System, Altamonte Springs, Fla.; and St Joseph Health System, Irvine, Calif.

“The bigger the dataset, the easier to find patterns,” Jain said. In April, computer giant IBM bought Explorys for an undisclosed price. IBM hopes to use Explorys and its huge dataset in conjunction with its Watson supercomputer and Health Cloud to provide medical insights and help health systems deliver better care.

One way that Mercy Health uses Explorys is to stratify diabetic patients by health risk and identify those who have not improved despite clinical intervention, said Dr. Amy Frankowski, senior medical director for Mercy's population health management program. By analyzing what has been effective with other patients, Mercy can work with its doctors to devise new care paths, better care coordination, or improved patient education approaches. “It's engaging our providers (by) giving them actionable data,” she said.

Since 2013, Mercy's diabetic patients have shown an 11% improvement in a comprehensive measure of diabetes management, which includes blood pressure, cholesterol, blood sugar, tobacco use and aspirin use.

Frankowski said Explorys has enabled Mercy to meet quality targets under value-based payment contracts. “We've done very well in our quality scores for the Medicare Shared Services Program,” she said. “Explorys was a tool that allowed us to track, take feedback, gain traction and improve.”