Deloitte, Intermountain Launch Healthcare Outcomes Tool

Healthcare Robotics: Patently Incredible Inventions

Healthcare Robotics: Patently Incredible Inventions(click image for larger view)

Deloitte has released a big-data analytics tool called PopulationMiner that it co-developed with Intermountain Healthcare, a provider organization based in Utah.

The tool is a further elaboration of the OutcomesMiner product the partners announced in July, with a twist: Instead of analyzing how a wide range of factors might have led to the outcomes of certain patients, PopulationMiner provides a broad population view that helps users decide what questions to ask about the relationships between treatments and outcomes.

Deloitte and Intermountain are targeting both healthcare organizations and pharmaceutical companies with PopulationMiner. In addition, the partners hope that the analytics tool will help providers and drug companies work together more closely to develop new medications and improve the use of existing ones.

"By providing the same tool to the provider and to the life sciences company, they can start collaborating in a much more meaningful way than before," said Asif Dhar, health informatics managing director for Deloitte, in an interview with InformationWeek Healthcare.

Dhar noted that a drug company can "start with a solution or medication and find out which condition they'll be useful for. They can also see the effects of medications on patients beyond the conditions they've been prescribed for."

Nevertheless, he stressed, PopulationMiner is not intended to help pharmaceutical firms research off-label uses for their FDA-approved drugs. "The intent is to start a process of meaningful science between life science companies and providers so that the layers of middlemen activity that distance the providers from those companies are eliminated," he said. Among these middlemen, he added, are "data consolidators," salespeople, and clinical trial organizations.

The analytics can help pharmaceutical companies identify patient cohorts they'd like to target in studies, he noted. The press release about the product launch clarified this point: "PopulationMiner can also produce hypotheses across complex, comorbid patient populations and a pathway for conducting follow on research with the same populations."

Regarding the use of the tool by provider organizations, the release said that PopulationMiner can help them "discover opportunities" to improve the quality of care and reduce costs. But the analytics are not designed for clinical decision support, said Katherina Holzhauser, assistant VP of IS commercialization for Intermountain, in an interview.

Instead of asking a narrow scientific question and doing an expensive study, she said, researchers can use the tool "to ask big, broad questions and play with variables very quickly." For example, if they were interested in women's health, they could look at women over 40 with three diagnoses and see their outcomes. Then they could add a fourth diagnosis to see how those outcomes change.

Like OutcomesMiner, PopulationMiner draws upon Intermountain's huge data warehouse, which includes a wide range of outcomes data. Altogether, the warehouse contains about 90 million patient records. For most purposes, Intermountain researchers are looking at 2.5 million records collected over the past 12 years, according to Holzhauser.

Organizations that license PopulationMiner can subscribe to a portal to view Intermountain's data and insights, or they can apply the tool to their own data, a Deloitte spokesman said.

Although few care-delivery organizations today have this kind of research ability, Holzhauser said she hoped these tools would help more institutions conduct outcomes research. Intermountain and Deloitte plan to build a collaborative of healthcare organizations to push the research further.

Dhar declined to reveal the names of pharmaceutical companies that want to use PopulationMiner but said there has been "strong" interest in the market. In addition, he said, the new tool was built partly in response to requests for this kind of analytics from customers who were already using OutcomesMiner.