What Does All That Healthcare Data Really Mean
At a conference on Monday at the Brookings Institution in Washington, industry and government officials cautioned that simply releasing data in bulk without context or meaningful comparative tools won't achieve the goals of lowering the costs of care, increasing operating efficiencies and, ultimately, improving care.
"We talk about transparency, and lately I find myself asking the question, what does that really mean" says Stuart Guterman, vice president of Medicare and cost control at the Commonwealth Fund.
Healthcare Data Without Context is Just Data
Guterman describes the release of large, unordered datasets as the "proverbial black box inside a Lucite box, and calling that transparency," while in practice those efforts do little to unlock meaningful insights about the working of the healthcare system.
"It's great to have a lot of data, but what we really want is information," Guterman says. "That means we need to have data in a usable form."
[ Related: How to Navigate Big Data in Healthcare ]
At the same time, panelists acknowledge that the push for open data in the healthcare system is only in its infancy, and give credit to the federal government for some of its recent efforts on that front, including the release of a trove of physician payment data under the Medicare Part B program this April.
At the federal level, the drive to release more healthcare data has gained momentum through the various open government directives promulgated by the Obama administration, a charge picked up by Todd Park, formerly the CTO of the Department of Health and Human Services who went on to hold that role for the whole of the federal government.
[ Related: Health IT Success Hinges on CIO-Business Collaboration ]
Niall Brennan, who holds the newly created position of chief data officer at the Centers for Medicare and Medicaid Services, credits Park as an early champion of open data at HHS, but admits that some of the early releases didn't immediately translate into the type of information that would support the efforts of the research community or help patients make more informed decisions about their care.
"I think how you have to view our various data releases and exercises in transparency is it's been an ever-evolving journey," Brennan says. "Those first couple of years were an interesting process, because the commitment to transparency was definitely there, but some of the resulting datasets -- they were good, but they left people wanting more."
What Types of Health Data Will Go Public Next
Brennan won't say what types of data the government is considering pushing out to the public next, though he acknowledges that the more contextual information is included in those datasets the more useful they will be. So, for instance, researchers combing through a bulk list of payment data could benefit from sorting tools that would enable them to classify physicians by geography or the types of patients they typically treat.
Some of that information was included in last year's release of billing data for some 3,000 hospitals, the first release of its kind, which offered side-by-side comparisons that unveiled wide swings in billing rates for the same procedures at comparable facilities and garnered significant attention in the popular press.
"It highlighted massive variation in hospital charges for the same procedures often in the same geographic area and certainly nationally," Brennan says. "I think it really resonated with people, frankly I think more than we thought it would."
The reaction within the community of care providers to broadening access to health data has been mixed, with some groups voicing concerns about the accuracy of the information and the lack of context leading to erroneous interpretations.
Charles Cutler, an internist and the immediate past chair of the American College of Physicians, warns that aberrantly high Medicare billing rates alone are not enough to indict a practitioner for bilking the system.
In some practices, Cutler argues, a physician's patient base might be disproportionately comprised of Medicare beneficiaries, and the cumulative charges reported in CMS data might be perfectly legitimate.
"Simply because costs are high, we have to be careful not to attribute that that is somehow fraudulent or that the high utilizer is wasting services," Cutler says.
"He or she may be," he acknowledges, but "we have to be careful that there's not guilt by association."