In health care, it seems that enterprise data warehouses typically are IT luxuries enjoyed by the largest systems and academic medical centers employing Ph.D. informaticists.
However, smaller facilities, down to tiny critical access hospitals (CAHs) and federally qualified health centers (FQHCs) may also find themselves scrambling to set up their data warehouses. By and large, their electronic health record (EHR) systems cannot aggregate and analyze data adequately for participation in quality-based reimbursement programs such as accountable care organizations (ACOs) or similar programs from private payers.
EHR inadequacies aside, there can be other justifications for small facilities to set up an enterprise data warehouse, said Heather Budd, COO of Blackstone Valley Community Health Care. Blackstone includes an FQHC among several Rhode Island facilities focused on primary care, including internal medicine, family practice and OB-GYN.
"You must have insight into your own data," said Budd, who engaged the services of Arcadia Solutions LLC to set up Blackstone's EHR, a data warehouse, patient data analytics and payer reporting mechanisms. The organization will defray the costs by offering its use to two other facilities as a shared service. That idea remains in the planning stages, but it may ultimately extend to more area providers.
"Trust is the underpinning of all these different initiatives," Budd continued, adding that, although Blackstone isn't currently a part of an ACO, it's setting up the IT infrastructure to be able to join one. "If you were to join an ACO, without your own analytics on board you've got to trust the way they're calculating everything. And right now it's not very well defined how it's calculated in the first place. "
She added: "So being able to do parallel calculations and have the negotiation leverage to be able to say, 'This is where we are with our performance, and this is the level at which we're contributing to your savings,' I just don't see any other way to do that without having your own data. It's worth every penny."
Budd also points out that having your own enterprise data warehouse grants a health care provider independence -- that is, the ability to leave one value-based reimbursement program and join another. "Cost metrics are really in flux at the moment, and until we agree on a set that's truly reflective of the care and efficiency of the care being provided, I think we're going to need that nimbleness to switch horses midstream. Ownership is, really, power."
'Long, expensive journey' to ACO-class analytics
Applications such as lab information systems (LIS) and EHRs do not provide the report-writing and data analytics capabilities that ACOs and similar programs will require, according to Al Shulski, IT director for lab information systems at the massive Geisinger Health System, headquartered in Danville, Pa. and serving 2.6 million residents in 44 counties.
While Geisinger has not joined the Centers for Medicare and Medicais Services' Pioneer ACO pilot, the organization is running analytics for internal purposes, both for clinical and operational data such as supply chain management and staff utilization. Only clinical information resides in Geisinger's enterprise data warehouse, so it employs a separate Altosoft Corp. analytics platform to gather data across several information systems to write the complex reports.
Soon, Geisinger will have to marry patient test results from the lab system to patient demographics contained in the EHR for meaningful use reporting. The data warehouse and analytics could make that process more straightforward, though at this point, depending on the final federal requirements, the health system might be able to do it from within the EHR.
"A lot of hospital [lab] systems, historically were very, very good at getting data into them, and not so good at getting data out of them," Shulski said. "That applies to EHRs as well…the native report writing tools from the vendors supply very limited information."
Personally, Shulski said, smaller providers will be challenged by ACO requirements and may have a hard time meeting them. Geisinger, he said, implemented its EHR system back in 1995 and has developed ACO-like best-practice care recommendations for its physicians in the form of clinical decision support via that EHR. That's been one "long, expensive journey," as Shulski put it.
"[ACOs] are going to be very challenging in the timeframe and in the costs that these smaller organizations are going to…incur," Shulski said. "I think that some of them will be able to pull it off, but I think others will find it too daunting to pull it off on their own and will need to partner with someone who can help them."
Altosoft CEO Scott Opitz said that even for smaller providers, the enterprise data warehouse will be a necessity for the new quality-reporting based health care payment models.
Through customers such as Geisinger, Opitz is finding that analytics tools like his company's offerings will have to reach beyond the enterprise data warehouses, since data points that lie outside the warehouses will be needed to assemble the reporting that payers will require.
Heather BuddCOO, Blackstone Valley Community Health Care
It will only get more complicated as some providers opt for cloud services -- although the more open and standards-based those vendors become, Opitz added, the more straightforward those analytics processes will become in turn.
Complicated or not, Opitz said he believes smaller providers will struggle with all the IT firepower needed to stay in business. "I don't think there's any way getting around" setting up an enterprise data warehouse, Opitz said. "Individually, these smaller hospitals are already terribly strained. They've had a hard time keeping up with the investments around implementing HIPAA."
Judy Hanover, research director at IDC Health Insights, concurred, adding that EHR systems do not always provide the best tools for necessary analytics or even collecting the data required for reporting. She pointed out that, with so many data transactions coming in real-time during the course of care throughout a health system, it's tough to get the application to take the required snapshots for quality reporting.
"That's why they'd need to export or extract the data into a warehouse, or at least a repository of some kind," Hanover said. Does she think that smaller providers will have to host their own repositories? "It depends on what their intention is, to participate in the market -- and then, going forward, it depends on what actually happens."
Advice for small providers considering the enterprise data warehouse plunge
FQHCs and other providers sized similarly to Blackstone who might be exploring their own enterprise data warehouse implementations should seek partners, Budd suggested. That can include working with vendors -- her organization outsourced the Ph.D.-level informatics and some of the payer interfaces to Arcadia -- as well as tapping association resources and joining with other local providers to share costs, create efficiencies and develop best practices together.
"Banding together to give yourself leverage in numbers is really critical," she said. "I just don't think that a one-to-two-doc shop is going to have the power that's truly needed to negotiate with big hospitals leading an ACO -- which, at the moment, is how it's been envisioned."
Budd also advised small providers to carefully consider the risks of joining a hospital-based ACO. Hospitals are feeling uncomfortable pressures to downsize, she said, predicting that some will fail. The ones who survive will have to shift more care to ambulatory providers, so those providers need to evaluate their would-be ACO leader before signing on the dotted line.
"Choosing an innovative and long-term-thinking hospital partner is really critical. Otherwise you're going to get swallowed up," Budd concluded. "The ambulatory environment is truly what's producing the cost savings…They have the ability to impact someone's health care before they get to the acute stage, which is extremely costly."
Another potential partner -- one that can perhaps help defray the costs of data collection and warehousing -- could be research foundations. James Szmak, CIO for JDRF (formerly the Juvenile Diabetes Research Foundation but now going simply by the abbreviation) said his organization is sharpening analytics processes to improve one-on-one donor relations. This customer service initiative brings together data from disparate information systems both onsite and hosted by cloud vendors.
While concentrating on the "sales" side at the moment with analytics, JDRF will begin to work with clinical data -- and publish the findings of its analytics projects -- from research it funds among health care providers.
More clinical data warehouse resources
From HIMSS 2012: Governance key to creating effective health data warehouse
From the Health IT Exchange: Is an enterprise data warehouse necessary for joining an ACO?
Szmak added that JDRF looks forward to analyzing public data gathered through meaningful use and other quality programs centering around diabetes management. The more physicians who enroll in these programs, the more robust these health data warehouses will eventually become.
"There's so many different data sources now that -- based on the research we're having done and data that's available -- we'll start now guiding the research we think is necessary to fill the gaps of knowledge we have," Szmak said. This contrasts with the recent past, when the foundation reviewed outside research proposals and picked the most promising ones. "Those data sources are critical, and being able to pull publicly available data sources into Altosoft [analytics system] will give us that ability."