Business intelligence, data governance challenge healthcare CIOs

As healthcare providers nationwide install and upgrade electronic medical records (EHR) systems in hopes of capturing their share of $27 billion in federal incentive money for “meaningful use” of EHRs, those who manage the underlying IT infrastructure are hard at work trying to manage all the new electronic data being generated. As attendees of WTN Media’s ninth annual Digital Healthcare Conference heard earlier this month, EHRs, data governance and business intelligence are all intertwined.

“BI is strategic in healthcare now,” said Alan Eisman, North American healthcare director for New York-based BI firm Information Builders. EHRs are great at capturing episodic data, Eisman explained, while BI is good at looking across visits, hospitals and systems. But BI technology is useless without the data from EMRs.

“I’m ready to declare the era of business intelligence,” agreed Galen Metz, CIO and IS director for Group Health Cooperative of South Central Wisconsin. Metz and Eisman joined Dave Lundal, VP and regional CIO for SSM Integrated Health Technologies, Dean Clinic and the Wisconsin Integrated Information Technology & Telemedicine Systems network, for a panel discussion on BI in the context of Accountable Care Organizations (ACOs).

Lundal took issue with the U.S. Department of Health and Human Services’ proposed ACO regulations, saying that measuring performance on 65 quality measures is just too high of a bar for most healthcare organizations right now. “That’s a no-go,” Lundal said, adding, “I use the word ‘daunting’ more than any other word.”

Metz noted that Group Health, which has its own health plan in addition to a care delivery network, has long prided itself on care management and coordination—which he called “the central foundation of ACOs.” The organization has a bonus program for employed physicians based on Press Ganey patient-satisfaction scores and on meeting many of the same National Committee for Quality Assurance standards in the meaningful use and ACO regulations. Yet, Metz still believes Group Health has not figured out some pressing workflow issues. “The productivity we’re still struggling with,” he said. According to Eisman, transparency and alignment of incentives with workflow are key in building an ACO, so it is important that everyone knows what they are accountable for. “That’s where the data and the business intelligence really help enable what we’re trying to accomplish with accountable care,” he said.
CIOs need to consider all these factors even as they prepare for the switch to ICD-10 coding by the federally mandated deadline of October 2013. But they should not fear this conversion, said Barry Mason, VP for IBM’s global healthcare payer marketplace.

“One of the things that sets [business intelligence] up is ICD-10,” Mason said during a keynote presentation about what he called “big data” and the “transformation to an evidence-centric health system.” Mason said that most of the people he talks to in healthcare are asking about how to derive value from all the extra data made available by ICD-10, which is 10 times more granular than the ICD-9 code sets it is replacing.

According to Mason, ICD-10 is much less about hitting the compliance date than at understanding the impacts beyond compliance. “ICD-10 will allow us to populate electronic health records and personal health records with much more meaningful data,” Mason said, noting that IBM has a great interest not only as a technology provider but as a large employer facing runaway health insurance costs. This will drive a greater patient-provider relationship than ever before and support ACOs and the patient-centered medical home model gaining favor among healthcare purchasers.

“The movement toward coordinated, accountable care is a strong one,” Mason said. However, there seems to be a negative backlash every time someone mentions payment reform. “How do we do this in a way that doesn’t come off as cost controls?” he wondered aloud.

The answer, Mason said, is for healthcare organizations to discuss market forces in the same conversation as legislative actions to “get ahead of the game.” In other words, health systems need to managing change before change manages them. “What it’s forcing is a completely different way of looking at change and how to manage it,” Mason said.

He said that maybe 20 percent of healthcare organizations are truly embracing change and trying to lead and innovate through it, while 30 percent are “still in denial.” Meanwhile, the other half the market is frozen, wondering how to proceed.

Mason reported seeing more healthcare CIOs hired from outside industries, which brings in a system view that long has been lacking. “In healthcare, we still need to knit ourselves together as a system,” Mason said.

In the case of IBM, enter Watson, the supercomputer that beat “Jeopardy!” champions in a widely publicized challenge earlier this year. IBM recently announced that healthcare would be the first industry to use Watson, and that is no coincidence.

Just like the game show “Jeopardy!” healthcare decisions are based on real-time hypothesis generation from reams of unstructured data. “It forces us out of the paradigm of thinking only in rows and columns,” Mason said.

Mason said that Watson’s “first pass” was 50 percent accurate in answering questions on medical board exams, which clearly is not good enough. “This thing is about 12-18 months from commercialization,” according to Mason, after IBM builds several use cases and tests the technology with payers and physician groups.
The key for Watson in for healthcare will be making the technology fit specific workflows, and IBM will be looking to answer one critical question: “How can I take natural English, both structured and unstructured … and bring it back with high confidence to support physicians?” Mason said.

That may be easier said than done.

Nancy Birschbach, assistant vice president for information technology at Agnesian HealthCare in Fond Du Lac, said that it’s difficult to optimize workflows because many users don’t know what’s in their workflows and “don’t know the impact of their workflow in driving data support” and, therefore, CIOs face a conundrum. “I don’t think people know what to lead,” Birschbach offered.

She said data governance would be easier today if the healthcare industry had better organized data when preparing for HIPAA compliance a decade ago.
Of course, that is all water under the bridge now, so those involved in data governance have to work with what they have, not what should have been. That may mean CIOs will have to shift their focus.

“Data governance is as much about business process as it is technology,” said Jim Orr, business development director for Information Builders. “In fact, the technology piece is probably the easy part.”

Orr and other speakers at the WTN Media conference recommended that IT leaders engage with not only clinical staff, but the business people within their organizations as well, so that everybody has the same objectives for managing all the new data.

“It has to be the executive leadership that runs the governance process,” added Philip Loftus, CIO and VP of Aurora Health Care, which has 15 hospitals across the state. He said that the IS Steering Committee at the Milwaukee-based health system reports directly to the COO and CEO.

Orr said he tells clients to establish an office of data governance or information management to give the idea legitimacy within the organization, as Aurora did. “We recommend that you think about a [leadership] council,” Orr added. A council will serve as an “enabler” so the office can do difficult the day-to-day governance work.

“Data assets are probably the toughest assets to manage because they’re fluid,” Orr said.