Sophisticated systems that link payer, financial and clinical data will drive population health efforts. More IT spending and a learning curve lie ahead.

After spending big bucks on new or upgraded electronic health record systems that can meet meaningful use criteria and other benchmarks, trustees now are getting an unwelcome surprise. It turns out that their EHR investments are only one part of the information technology equation needed to succeed with population health.

“They’re just burned out on IT spending,” says Dana Sellers, CEO of Encore, a Quintiles company. “They feel like, ‘Hey, I gave you $200 million, and you told me that was going to be everything you’d need. Now I want to see value from that.’ ” Some trustees even thought they’d be able to cut IT costs because of software system consolidation and other efficiencies, Sellers says.

The reality is jarring: EHRs that support a fee-for-service world are not sufficient for the population health management world. “[EHRs] are great because they help with transactions, and they can improve the documentation and billing and coding,” says Anne Docimo, M.D., chief medical officer of Jefferson Health System, Philadelphia. “[But] now we need to pull that information out of that system.”

Existing IT systems do contain patient data collected for the clinical communication and analyses necessary to manage populations, says David Kim, a health care technology consultant with Encore. But new software that wraps around EHRs and their existing data collection functions will be needed to address the shift to value-based care. “To say that you are just done with your investments in IT,” says Kim, “is not a good strategy.”

Different Purposes

EHRs are meant to work in clinics and hospital settings. Care coordination and management, by contrast, require providers to “look across the blank space” the EHR does not reach, says Sellers. In an accountable care organization, where multiple entities jointly take on risk for a population, each provider should be “reaching out to manage a patient in the void between multiple organizations.”

Health care still is running on a fee-for-service chassis that supports the way care has long been delivered and reimbursed, Kim says. But as fundamentals of delivering care change, providers “need to make investments to address the shift.”

Think back to old-fashioned cash registers, Docimo suggests. Check-out staff used them solely for transactions with people presenting items for purchase, tallying the cost and maybe computing change. Today those registers help to keep inventory, predict customer spending patterns, and drive operations and margin. “We are just starting down that path [with IT],” she says. In health care, the aims are to improve quality and decrease cost, and to avoid the consequences of not meeting those aims.

To take any sort of financial risk — pay for performance, bundled payment, shared savings or full capitation — “you’re going to need to have an electronic data warehouse, payer data, financial data in terms of what your costs are, and your clinical data,” Docimo says. The technology then enables providers to build the necessary clinical actions into a workflow to hit performance and payment targets.

In short, to operate in a value-driven era, health care systems must be able to:

• Turn raw data into targeted, actionable information to identify and remedy care gaps.

• Combine EHR data with Medicare, Medicaid and commercial claims to see the most complete picture of a patient’s behavior.

• Create and operate a higher plane of electronic communication necessary for care coordination.

• Make full use of available data for timelier and more comprehensive management of patients.

An Underdeveloped Market

Health care’s new business model doesn’t have a ready-made technology solution. Systems for coordinating care, analyzing data and providing patient-centered care are in the early stages of development, with an array of vendor approaches that each cover only part of the functional range, says Sellers. Encore oversaw technology selection for a client recently and got 26 responses to its request for proposals — each with varying functionality and none able to cite more than three organizations that were using their products.

The value-oriented IT market is a promising opportunity for EHR companies that could sell into their existing customer base, but “the vendors have, in some ways, been a little bit too far behind,” says John Jenrette, M.D., CEO of Sharp Community Medical Group, San Diego. “They haven’t been keeping up with what the needs are.” Sharp, an 800-physician independent practice association with 60 percent of its contracts bearing financial risk, has spent significant time showing vendors a road map of what it needs from IT, “and some are doing better with it than others,” Jenrette says.

With the market in transition and significant consolidation occurring in the IT sector, “there shouldn’t be a rush to spend money on all of these technologies,” Kim advises. But providers are under the gun to move into value-based care, with the federal government and commercial insurers migrating into risk territory.

Health systems can’t ignore the need to invest in the value model, but identifying and purchasing the right IT will be difficult because the market for these solutions is immature. “There are organizations at the leading edge of that [value] spectrum that are saying, ‘We can’t wait, we have to move now,’ ” Sellers says. “And there are organizations, as in any wave of adoption, who come along a little slower.”

Data Deficiencies

In the meantime, health systems have to tackle the use of high-end analytics and decision support. At minimum, the data in EHRs and administrative systems have to be retrievable and usable. Consistent data definitions now become critical, along with the design and discipline to enter data in the right places to track and analyze quality.

That’s not happening today, experts warn. Patient data are captured in multiple places, the data elements are poorly defined and what is collected isn’t ready to be used to manage care across the continuum. Health systems think they can buy business-intelligence tools, load in data from EHRs, nursing documentation systems and other sources, and get great insights, says David Garets of Change Gang LLC, an IT consulting firm. But that can’t happen unless all the data are available and “normalized,” or engineered in a data warehouse to recognize differently represented elements of data as the same thing, such as heart attack, myocardial infarction and AMI.

Sellers tells of a large system that, anticipating requirements in two years to submit inpatient quality measures electronically, went through a trial run. “The results all came out zeros,” she says. The lesson: Electronic measures require building and using consistent data, assuring that particular data fields are built and then always used for a given measure instead of capturing those elements haphazardly. For population health, providers must identify key elements for financial and care management, then enforce consistent, clean capture by all staff.

Building on the EHR

To be sure, EHRs and the data they produce are essential for the move into population management. “Everybody has to invest in the basic EHR just to be able to codify information, to eventually get data that become meaningful,” Jenrette says. The EHR also creates valuable registries of people by disease state and other characteristics, produces summaries of the gaps in recommended care for those people, and supports the right decisions in the care of patients, he says.

In any shared risk or performance basis for revenue, however, “you are responsible for the entire population that’s been attributed to you on the basis of whatever attribution model your payer partner is using,” says Anne Meara, associate vice president for network management at Montefiore Medical Center, Bronx, N.Y. “So the claims information becomes very important. And that claims information does not live in your EHR.”

Most organizations are identifying risk by analyzing large amounts of data and are measuring outcomes at the end. But between those two functions is where care management and care coordination take place, and the EHR is not sufficient. Enrolling members in a program, assessing them, developing a care plan and managing them through the care continuum requires different technology.

Care management technology, for example, helps to maintain a full picture of each person’s financial and clinical impact on the enterprise, says Encore’s Kim, a consultant on Montefiore’s care management project. Such a system houses all the points-of-care contact across all venues and what needs to happen next. Managers accountable for health care costs need something like that to track and intervene effectively, he says.

Part of the Sharp physician group’s IT plans include a set of data analysis tools that combine clinical and administrative data to determine cost drivers — that is, episodes of care put in financial context. “You have to look at quality and best practice, and you have to marry that with the most efficient and effective [options],” says CEO Jenrette. Stellar quality metrics, produced through EHR data, might not be as good a thing when subjected to cost analyses. “You could have the best care in the world and treat every one of your prostate patients with a proton beam, but you’re going to go under,” he says. Less costly treatments might produce comparably desirable results.

Workflow Drives Everything

Ultimately, population health management is all about execution.

Montefiore manages a population of 400,000, relying on technology that can track the people most in need of care, according to their health status, and in keeping with variable requirements of multiple risk contracts. Automating workflow is critical to keeping tabs on where people are in the care process, ensuring that things don’t fall through the cracks, and using health professionals to their highest level of skill. It’s essential as health systems’ business approach includes more sites and types of care.

“Workflow eats strategy for lunch every day,” Meara says.

Standardizing workflow starts with standardizing assessments, problem lists and sets of interventions depending on patients’ problems and the information collected for them. “How else do I measure whether or not we’re doing the right things, how much of it we’re doing, whether or not it’s working, if I’m not collecting that information and using it to drive the next step in the process?” she asks.

An automated workflow also must incorporate payer contract details. “When people say ‘population management,’ they assume everyone in the population arrives on your doorstep in the same way, subject to the same rules and requirements, but that’s just not the case,” Meara says. “Every contract we have has a different set of requirements, and we probably have a dozen contracts right now.

“A lot of provider organizations that are getting started along this path haven’t come to the full realization yet of what that really means to the staff on the ground,” she says. But physicians don’t need to be plugged into the particulars of a contract, she notes. “We try to keep that out of the line of sight of the physician. The physician should be taking care of the patient and not worry about all of this stuff.” 

John Morrissey is a contributing writer to Trustee.

Your Technology To-Do list

The information technology capabilities involved in sizing up an attributed population, analyzing large amounts of data and guiding care activity are the price of producing reimbursable value — and for many health systems, the expense may not stop there. Existing information systems may require enhancement or replacement, and the more sophisticated infrastructure may demand new expertise to build and run.

One likely purchase is a more advanced revenue-cycle information system, says David Garets of Change Gang LLC, an IT consulting firm. “The 20-year-old systems that most health systems have now are not equipped to do fee for quality; they’re equipped to do fee for service.”

In addition, the hospital should be part of a health information exchange, Garets says. “You’ve got to figure out how to get data from outside your environment, because you’ve got to be able to track patients across the entire continuum.”

Ideally, population health functions should work off the same platform as the electronic health record for all clinicians and coordinators, and be tailored to different roles, says John Jenrette, M.D., CEO of Sharp Community Medical Group, San Diego. “Physicians are not going to go to another [IT system], another website, to get information.”

And the enhanced breadth, complexity and functionality of new IT infrastructure likely requires a chief data scientist, someone with unique skills and experience who commands a sizable salary, Garets warns. But, he adds, “You’re going to have a really hard time getting access to accurate data if you don’t do this.

“There’s a lot of hard work that’s got to be done, there’s a lot of infrastructure that has to be built, and it’s not going to be cheap,” Garets says. — J.M.