When it comes to the plunge into value-based healthcare of all kinds—participating in value-based programs, both mandatory and voluntary; participating in accountable care organization (ACO) and other risk-based contracting involving population health management; participating in bundled-payment contracting; and all the rest—the leaders of patient care organizations are only beginning to fully realize how complex the strategic IT challenges involved really are.
For the leaders of the eight-hospital Orlando Health, based in Orlando, Florida, participation in accountable care/value-based healthcare has provided welcome opportunities—as well as some very practical strategic IT challenges.
As Jerry Senne, vice president of value-based care and population health, and Brandon Burket, director of value-based and accountable care, note, Orlando Health joined the Medicare Shared Savings Program (MSSP) on January 1, 2013, and, Burket reports, “We’ve been able to generate shared savings every single year. We’ve also generated some of the highest quality scores” in the program, he notes. And their organization, whose physician network encompasses 900 employed and 2,800 affiliated physicians, and a health system staff of over 18,000, continues to move forward in its accountable care/value-based care work.
Senne and Burket and all of their colleagues at Orlando Health have been making tremendous progress on all fronts, but it should surprise no one that data and IT issues have come to the forefront in this work. “At a more granular level,” Burket says, “we’ve had challenges with data, in terms of timeliness, accuracy, and the lag itself. We get the reports only monthly or quarterly with some payers, and often, that’s not often enough to be actionable, or it’s too old; so it’s more using the data as a compass than a roadmap.” Still, he adds, “The benefit from getting the data, even if it’s old—is that it creates a level of transparency and competitiveness among providers.”
Indeed, crucially, Burket notes, early on in the process as an MSSP-participating ACO organization, “We quickly developed what we called a data roadshow, and went around to physician practices, asking them which metrics would be useful for them. We developed 15 or 16 measures and developed our own homegrown data dashboard every month, and actually said, ‘Dr. Jones, you’re number 54 of 56 docs on this contract.’ We created a level of transparency that engendered a level of competitiveness that made people want to get better on these measures. And over time, the standard deviation actually got tighter; people were performing at a much higher level, and there was much less variation. That said,” he continued, “a lot of our commercial contracts are more claims-based, so it’s less onerous on the providers to report. But with regard to the MSSP, it’s quite a bear of an endeavor that many organizations struggle with every year. And it takes our care and coordination team quite a bit of time and effort to report on accurately.”
As Senne sees it, the ability to capture “population-based data—payer/claims data, pharmacy data, and to be able to distill that into reports, is important,” while at the same time moving physicians forward “to participate in medical governance and really participate in the journey, that is critical. Second,” he says, “to be able to achieve and maintain actionable data,” is crucial. “In early versions that payers came out with, they would say, your population health performance was ‘78’ on a scale of 1 to100; well, that was meaningless to physicians. On the other hand, if they gave you a list of the diabetics on your panel who were not controlled in their hemoglobin a1c, that was usable.”
Of course, all of these imperatives end up falling into the laps of CIOs and their colleagues in organizations participating in value-based care delivery and contracting. Rick Schooler, who’s been Orlando Health’s CIO for 16 years, sees all this clearly. “There are several must-do’s” for CIOs and other healthcare IT leaders in this venture, he says. “Number one, throughout your healthcare continuum, you’ve got to have integrated information, to the degree possible. And that’s a lot harder to do than to say. As your patients go through the continuum, those who are managing their care have to respond to things that do happen or don’t happen. You’ve got to have an EMR [electronic medical record] platform that’s generating and is capturing data, across that continuum. And you’ve got to have what a lot of people are calling population health platforms. And it’s not so much, ‘Hey, Rick has a lab result,’ but rather, ‘Hey, Rick didn’t get his lab work done.’ So there’s got to be a surveillance element going: are the things that should be happening, happening?”
What’s important, Schooler says, is to build the analytics capabilities to determine that “what should be happening is happening, and what shouldn’t be happening, isn’t happening. And you’ve got to understand physician performance: are physicians ‘in protocol’ in terms of managing their patients? You’ve also got to be able to look at claims data. And these are basically descriptive analytics. I don’t think you’ve necessarily got to have predictive analytics, though if you do have those, that’s great. But you have to make sure patients are compliant with their care plan.”
Ultimately, Schooler says, that means that “You’ve got to have integrated information across the continuum; you’ve got to have a population health or care management capability, that’s typically found in population health platforms; and then you’ve got to have analytics capabilities that allow you to look retrospectively, and potentially predictively as well. You’ve got to have the physicians and information to enable the patient-centered medical home. Do you need an enterprise data warehouse? No, but you need the ability to capture key information about patients.”
What about I.T. Strategy and Downside Risk-Based Contracting?
Things get even hairier when it comes to pursuing any contracts with downside risk, Schooler says. “The minute you get to downside risk, you’re going to need to have your clinically integrated network [CIN] fully developed,” he says. “And the more mature that clinically integrated network is, with primary care, the more successful you’ll be.” What’s more, he says, patient care organization leaders who want to become skilled enough to pursue downside risk over time, should start by mastering data analytics around their ACO work first. “ACOs are like a toe in the water. They’re basically upside reward,” he says. “But as you move into downside risk, you’ve got to have a more mature clinically integrated network, with mature IT. And these clinicians are not on the same platforms. So you have to be able to pull data in from all your points of care, and bring them into your database.” In other words, referring to the health information exchange phenomenon, he says, “You need an HIE platform. When you start taking on downside risk, everybody in that platform needs to know what’s going on. And when you expand that CIN, there’s no way you’re going to have all the providers on the same platform, so you need an HIE platform to bring all the data into the same place, absolutely including images.”
Christopher Longhurst, M.D., CIO of UC San Diego Health, agrees with Schooler with regard to the analytics tools needed to pursue value-based contracting. “With regard to value-based purchasing,” Dr. Longhurst says, “I think the critical unmet need from an infrastructure standpoint is self-service analytics tools. The more we can empower our physicians and administrators to query the EHR [electronic health record] analytics, the faster we will develop a data-driven culture.”
Physician Group Leaders Plunge into Intensifying Efforts
The healthcare IT leaders at hospitals and health systems would do well to heed the advice of the CEOs of physician groups that have already spent several years participating in accountable care organizations and value-based purchasing. Those medical group CEOs, as they plunge more intensively into continuous clinical performance improvement work, say that the slope only gets steeper over time.
Asked what the hardest thing is about using data in his organization’s ACO work, Jeffrey LeBenger, M.D., says, “I could go through a list.” LeBenger, M.D., chairman and CEO of the nearly-800-physician Berkeley Heights-based Summit Medical Group, a multispecialty physician group practice that covers a broad swath of northeastern New Jersey, and which is involved in several ACOs, says, “You need data aggregation, master-patient indexes, provider data management, and provider performance management reporting. You have to be able to do ad-hoc reports, you have to have a mechanism to monitor your quality metrics, you need to look at all your claims analytics, you have to put it into a clinical workflow model, you have to look at your attribution lists coming from payers, and your HCC coding lists, and on and on.”
Jeffrey LeBenger, M.D.
Is that a lot? Absolutely, Dr. LeBenger says. “There are all of these issues. And everything has its own IT solution. And how do you bring those IT solutions together? That’s where the cost really comes in. It’s a million dollars for every solution; and you could be talking about 12 different solutions. And sometimes you can’t aggregate the data 100 percent, sometimes you can only achieve 80 percent, and that has to be good enough. But I bring it back to the business model. If you have a good clinical practice system and you bring it back to managing things in the ambulatory environment, and looking at your high-acuity patients in a certain way, and looking at your skilled nursing patients, and how you manage your patients in a quality metric. It’s the clinical model that’s of the utmost importance, but then you need all these data points, to make sure you’re doing it right. So it’s really a conundrum here, because it’s so expensive.”
The leaders of the Mt. Auburn, Mass.-based Mount Auburn Cambridge Independent Practice Association, or MACIPA, have been intensifying their ACO development over the last few years, reports CEO Barbara Spivak, M.D. Dr. Spivak, who has been CEO of MACIPA since September 1997, helped lead the organization into the federal Pioneer ACO Program, which MACIPA participated in for three years; MACIPA is now in Track 3 of the Medicare Shared Savings Program (MSSP) for ACOs. The challenge in Massachusetts, Spivak says, is that payment levels simply do not fully reward the amount of work required of physician groups to succeed in ACOs. Nonetheless, she and her colleagues are persevering; and they’re finding that the more deeply they plunge into clinical performance improvement work, the more significant the data and IT needs become.
In fact, Spivak and her colleagues made a significant decision a couple of years ago, even as they were barreling forward with their ACO work. After years of MACIPA’s physicians practicing on a very wide range of electronic health records (EHRs), they moved to integrate onto a single EHR platform, as they found their clinical performance improvement work being hampered by the heterogeneity issue. “A majority of our docs were on eClinicalWorks,” she reports, “but because we had implemented in 2007, the government was saying that even though we had taken on shared risk, we weren’t legally clinically integrated—you had to be officially labeled clinically integrated. We had 89 different instances of eClinicalWorks, and we had some on athenahealth and Centricity, and Allscripts, and the hospital [Mt. Auburn Hospital] was on Meditech.” As a result, Spivak and her colleagues decided to unify onto a single EHR, from the Verona, Wis.-based Epic Systems Corporation; as of July, all the practices had migrated to Epic.
Meanwhile, asked about the dashboards and analytics tools she and her colleagues use, she says, “We use a wide variety of dashboards and analytics. Beginning back in Medicare Advantage,” she says, “we began to engage heavily in analytics-facilitated work.” One of the key things that she and her colleagues are doing right now is using analytics to monitor which nursing homes MACIPA patients are discharged to, following inpatient hospital stays. This is combined with MACIPA’s narrow-network program, which explicitly favors the five skilled nursing facilities in the area whose cost-effectiveness and clinical outcomes the MACIPA physicians have found to be the best. “So we have preferred providers in our area,” she says. “Obviously, we give them choice. But we also let them know that we have very intensive systems in these five SNFs, with nurse practitioners, and trusted physicians. There are hundreds of nursing homes in the area, and we picked five we trust the most. And we monitor admissions, lengths of stay, discharges to home, and what happens after 30 to 90 days after patients are discharged from the facility.” All of those processes are very important when it comes to succeeding as an ACO, she notes—and IT and data are essential to success in those processes.
Broader Development Needed for the Journey Ahead
Industry leaders are looking to broad strategic IT development, especially around infrastructure and interoperability, in order to solve some of the fundamental problems facing the leaders of patient care organizations, as they take their organizations further into value-based healthcare contracting.
Don Crane, president and CEO of the Los Angeles-based CAPG, which describes itself on its website as “the leading association in the country representing physician organizations practicing capitated, coordinated care,” with “close to 300 multispecialty medical groups and independent practice associations (IPAs)” as members, says this: “Your IT folks will need to keep developing better data warehouses, and find ways to make IT more interoperable, and provide better informatics for stratifying populations of patients, and physicians, and managing care, and getting the right providers in place. All of this, to me, is a dream come true for IT people, because you’re talking about a massive shift taking place, and one that needs to be tech- and specifically, IT-enabled. There’s lots of opportunity involved, and that’s very exciting,” he says.
Asked what some of the early stumbles have been in the IT sphere, Crane says, “One stumble has been around interoperability, and the lack thereof. All of these systems that have been designed and developed and sold, don’t work well with other systems. So this plethora of ways to interface and integrate data and systems—that’s been one of the barriers, for sure.”
Ultimately, Crane emphasizes, it will be a highly sophisticated blend of efforts—IT- and data-facilitated performance improvement work—that will successfully lead patient care organizations forward into the emerging world of value-based healthcare. “You need a clinical intelligence running this, but also one connected to a P&L and a balance sheet. So it is this dyad, this blend, of the clinical and the administrative,” that will determine the future, he says. “And the intelligence of the system is this blend of the clinical and administrative; that dyad produces the best results or the potential for the best results,” going forward.