Plunging Ahead: Advanced Physician Organization Leaders Move Forward to Manage Risk

Jan. 29, 2018
For the July/August Healthcare Informatics cover story, advanced medical group leaders share how they are analyzing and using data to feed continuous clinical and operational performance improvement

In the middle of the summer of 2017, many patient care leaders have understandably been focused on the June 20 release of the proposed rule for 2018 for the Quality Payment Program (QPP) under the MACRA (Medicare Access and CHIP Reauthorization Act of 2015) law, covering both the MIPS (Merit-based Incentive Payment System) program and the other side of the MACRA law, governing participation in advanced payment models (APMs). There is considerable complexity in the 1,058-page proposed rule; and much in the rule might change before it’s made final. But much of the discussion of the proposed rule, in the weeks following its release, was around the level of rigor and pace that should be embedded in forward progress around compelling physicians in practice more deeply into value-based healthcare payment.

Meanwhile, the leaders of more advanced medical groups say that the discussion around the details of MACRA’s requirements feels deeply secondary to the future they’re already building. Their organizations are already involved in taking on some level of financial risk in risk-based contracting, whether through federal accountable care organizations (ACOs) or ACOs sponsored by private health insurers, or through bundled-payment contracts (again, federal or private), or through any one or more of a broad range of population health-focused contracts of all kinds. And these medical group leaders have been spending the last few-to-several years learning how to collect and analyze data and use that data to feed continuous clinical and operational performance improvement, while at the same time building learning organizations, and transforming the physician cultures in their organizations. In other words, they’re already way ahead of the MACRA compliance game.

What’s exciting about that is that physician practice-based/clinic-based care remains at the core of the area of potential for change in healthcare, particularly when it comes to patients with chronic illnesses. Hospitals, health systems, health plans, employers, and governments all have a deep investment in efforts to improve patient outcomes and the patient experience, reduce costs and improve operational efficiencies; but it is in the physician practice setting that the most immediate interventions can be made—and the senior clinical and administrative leaders of the most advanced medical groups are making those interventions, and learning a great deal in the process.

One medical group leader whose attitude illustrates this perspective is Jeffrey LeBenger, M.D., chairman and CEO of the Berkeley Heights-based Summit Medical Group, a multispecialty physician group practice that covers a broad swath of northeastern New Jersey. Asked about the proposed 2018 rule for the QPP under MACRA, Dr. LeBenger quickly brushes the question aside. “I’m not concerned about that at all,” he says. “You’ve got to set up a model of care and make the economics work for you. If you can do that,” he says, there is a clear path to success. Indeed, he says, he and his colleagues have already zoomed past any of the core outcomes requirements mandated by the QPP.

Looking back at the last couple of years, LeBenger reports that “We’ve done very, very well in our population health process, meaning that we beat the market in PMPM [per member per month] costs in the state of New Jersey by almost 8 percent in the past year. What we found out” in terms of how to achieve success under risk-based contracts, “is that you must share data with your physicians, meaning that you have to scrub it, you have to fix it, and you have to work with financial and clinical data together. And it has to be timely data—you need timely data from the payers. And you have to put the right data into the right health information exchange. And you have to understand where your high-cost points are in your model of healthcare.”

Jeffrey LeBenger, M.D.

LeBenger further clarifies what he believes is and is not essential to medical group leaders wading into the ocean of risk-based contracting. “Everybody says it’s population health, population health,” he says. “To be honest, I think that’s a bit wrong. You have to look at your model of healthcare. We’re getting very close to 800 providers; we’re in seven counties. And what we’ve done is that we’ve created high-acuity urgent care centers with imaging and with specialists; and we’ve created high-risk clinics for our high-risk patients. We do transitions of care for our sickest 5 to 10 percent of patients. We have home healthcare. We have medical directors at 70 percent of the long-term facilities we’re in. We compensate based on outcomes. And we follow the patient-centered medical home concept; and we really manage the patients out of the hospitals. For us, a hospital is a center for tertiary or quaternary care. We look very closely at monitoring patients out of the hospital.” It’s all of those factors—essentially, creating a data-driven care management system around one’s covered populations, especially for all those patients with chronic illnesses—that is at the heart of success in this emerging world, he emphasizes.

LeBenger’s perspective is shared with all of those interviewed for this story, who agree that:

  • The leaders of medical groups pursuing risk-based contracts, or considering doing so, need to develop core strategies that risk-stratify their covered populations and shift very strongly towards a care management-driven system of care that focuses on upstream interventions that avert ED visits and inpatient hospitalizations to the extent possible.
  • Medical group leaders need to develop highly sophisticated data and information systems that accurately bring together and analyze clinical information from their electronic health records (EHRs) and claims data from the payers with whom they’re contracting, in order to identify any gaps in care or care management; and they need to dovetail those information and analytics processes with multidisciplinary care management processes.
  • The health risk assessment and population health analysis process, and the process of using data analytics to empower multidisciplinary team-based care management processes, must in turn be tied into physician incentivization, including compensation.

In the case of Summit Medical Group, LeBenger notes that “How we get doctors to manage things better is that we provide up-to-date monthly dashboards,” which he and his fellow leaders use to incent their physician colleagues towards optimizing clinical and financial performance within the group. With support from those metrics, he says, “We compensate our doctors 20 percent based on quality, outside of direct productivity. And that’s true for our primary care doctors and specialists alike.”

A few hours north of northeastern New Jersey, in New York state’s Hudson Valley, the leaders of Crystal Run Healthcare have been on a trajectory similar to that of Summit Healthcare’s. The 400-plus-physician multispecialty group, based in the town of Middletown, has been succeeding at ACO and population health work for several years now. And what have Crystal Run’s leaders learned in the past couple of years? “We’ve learned, number one, that it is a lot easier to improve quality than it is to lower cost, because for the most part, quality is under the direct control of the organization,” says Scott Hines, M.D., the group’s chief quality officer. “So we’ve built a lot of systems around closing gaps in care for cancer screenings, immunizations, well-child visits, etc., and we really are in control of that. When it comes to cost, we’ve done a lot to monitor and reduce variations in care that are provided within our four walls and have resulted in greater standardization of care according to best practice guidelines, eliminating unnecessary or duplicative services; but the problem is that when the patient goes outside our walls to the ED or some other venue, we can’t control that. So trying to keep people in the most appropriate site of service has been our greatest challenge. The knee-jerk reaction for many patients is to go to the ED.”

Like their counterparts at Summit Medical Group, Crystal Run Healthcare’s leaders have been focusing to a considerable extent on driving clinical performance improvement through data, including through physician dashboards. “One dashboard that we’ve spent a lot of time updating and tweaking is what we call our payer quality scorecard,” Dr. Hines notes. “We look at our performance across the different metrics we’re accountable for—either a pay-for-performance-based metric, or a shared-savings or shared-risk incentive with potential quality gains.” Crystal Run is a participant in the Medicare Shared Savings Program for ACOs (MSSP); in addition, it has risk-bearing contracts with Aetna and with Blue Cross Blue Shield in its area, and a shared-savings arrangement with Cigna; and it is currently negotiating additional commercial risk contracts.  It also participates in three Medicaid managed care contracts in the state. A propos of all that activity, he reports, “We’ve built a payer quality scorecard that breaks it down by payer, shows the measures for each payer, and shows what we have in terms of data.”

Scott Hines, M.D.

Hines notes the importance of that analytics tool. “When we were relying solely on the payers’ claims data, there was at least a three-month delay/lag,” he says. “So we asked for specific numerators and denominators from the payers, for each of their measures.” In this context, the denominator is the patients included in a measure, and the numerator is the subgroup of patients that the health plan involved has determined are compliant with a particular measure. “And we take that information and feed it into our payer quality scorecard, and that captures any internal data, and shows us where we are in real time on those measures. We’re trying not to have any surprises, so that the payer says, for example, your benchmark was 75 percent for breast cancer screening, and you only hit 72 percent—that kind of thing.”

And what’s been learned from that process? “In the past, Hines, says, “we’ve asked for specific populations for each measure. We used to take each attributed population—say, 8,000 patients in a contract—and we’d feed those numbers into the scorecard, and the prediction was off by 15 to 20 percent for each measure, for two reasons. First, many measures had continuous enrollment criteria—the patient had to be enrolled for 12 months continually to be measured, and we had patients who weren’t, so that was distorting that number. And second, when you looked at the measure specification, we were applying the current year’s HEDIS (Healthcare Effectiveness Data and Information Set) specifications, but some of the payers were using specifications from, say, 2014 or 2015, so there would be mild differences in the inclusion or exclusion criteria, the lookback period, etc., so again, it was a source of discrepancy.” In other words, constant refinement of analysis and processes, is essential for success under increasingly-demanding risk contracts.

Advanced Medical Groups Seen Tilling New Ground

The kinds of work that Drs. LeBenger and Hines and their colleagues are engaged in at Summit and Crystal Run—the constant reiterations and refinements of strategic planning, data collection and analytics processes, and continuous clinical improvement processes—really is at the core of their success under risk-based contracting, those leaders agree. And so do industry experts and observers, including Matthew Cinque, executive director, performance technology at the Advisory Board Company, the Washington, D.C.-based advisement and research company. “The medical groups and physicians who have been involved in population health for any length of time, given that the ACA [Affordable Care Act] has been live for more than seven years—have evolved forward,” Cinque says. “The types of questions they’re asking and the needs they’re developing as they pick off the low-hanging fruit, are changing now. Early on, they focused on things like reducing utilization by frequent fliers in the EDs and elsewhere, and were asking questions like, which patients have been to the ED at least three times in the past year? Five years ago, that was very valuable information, and they’ve done a fair bit with that. Some of the questions people are asking now are around getting a better understanding of the predictive models used today, including asking about the social determinants of health, and using that data to factor into prediction of utilization; also around other forms of the psychosocial determinants of health.”

Matthew Cinque

Cinque goes on to say that “There’s been more success in identifying high-need or high-risk patients, with some opportunity for actionable intervention—either passing that data or list off to care managers, or circulating it among physicians. On the latter point around getting the physicians to actually use the data to improve outcomes—when you get down to the level of the individual physician who’s seeing 30 patients a day—success in actually changing behavior is still in its infancy. The vast majority of physicians are still operating primarily under fee-for-service” payment arrangements. The challenge on a practical level, as he puts it, is that “Not only do you have 100 diabetics who have uncontrolled A1c, but you’ll say, I want you to focus on these 40 who are in our risk-based contract—and in terms of focusing on the higher-compliance versus the lower-compliance patients—there are not many places that have cracked the nut on that.” As he sees it, one major element in success will come out of “wrapping care management functions and operations around physicians,” in order to lift some of the weight of trying to get patients to change their health-related behaviors, off the physicians.

"One Foot in the Boat” Continues For Now

All those interviewed for this article agree that one of the fundamental challenges in all of this remains what many in the industry are referring to as the “one foot in the boat, one foot on the shore problem” in U.S. healthcare, as even the most advanced patient care organizations still have a significant proportion of their revenues coming from fee-for-service payment models, even as they try to move forward to broaden their risk-bearing contract portfolios.

In making that shift over time, one of the core challenges remains convincing physicians in any physician organization to begin to shift forward, even as the incentives coming from the purchasers and payers of healthcare remain mixed, and sometimes not strong enough yet to compel revolutionary change. “It’s never easy; it’s always hard work,” says Mark DeRubeis, CEO of the 100-provider (85 physicians, 15 advanced-practice clinicians) Premier Medical Associates group practice in Pittsburgh. “But understanding what we need to change, and the kinds of leadership that are needed, and the necessary elbow grease, we’ve pretty much had the right attitudes among our physicians. And we are a physician-led organization, and we’ve found that that carries a good part of the weight,” he says, but quickly adds that “The hardest thing is that the healthcare system at large wants to effectuate a lot of change. Sometimes, federal policy leaders hit the mark in what they want to accomplish what they want, and sometimes, they just end up adding administrative burdens.”

He goes on to say that, fundamentally, “What’s been very difficult is financing the changes. Usually, the money isn’t sufficient, meaning that any money we’ve ever earned through quality is usually just plowed back into the organization for the next big thing, so to speak. It’s always up to the organization to figure out how to finance things; and it’s always, in healthcare, perform first and get paid later, so you really have to sell your physicians on the ROI” of moving forward on these initiatives.

The complexity of moving forward into risk while also managing fee-for-service payment-based operations at the same time is considerable, agrees James Whitfill, M.D., chief medical officer at Innovation Care Partners in Phoenix (formerly Scottsdale Health Partners). “For physicians who are paid by work RVU [relative value unit], whether in an employment model, or if they’re billing fee-for-service, it’s very hard to take them away from their core incentives to provide things that drive operational improvement, and it’s still common to see a work RVU-based productivity focus,” Whitfill notes. And he says, “Even though more advanced [physician] organizations are comfortable with their revenue cycle and EHR data, having the claims data coming in continues to be a brand-new space for providers, and getting your hands around that, particularly when you have multiple payers, is hard.”

James Whitfill, M.D.

That said, Whitfill reports that, “We’re seeing that in our clinically integrated network, physicians are starting to understand this model. They’re understanding two primary mechanisms. First, we’re doing face-to-face quarterly meetings among our CIN leaders, of which I’m one, and our primary care physicians. It’s myself and other leaders with each practice, once a quarter. And after about three quarters, they begin to understand what this is about. But it takes those meetings, one-on-one or one-on-two. Also, we front-load revenue to practices in value-based contracts as we move into a contract, while letting them know that those payments might have to stop. That helps. So those two things, we feel, have gotten the attention of providers and driven engagement perhaps more than in some other organizations.”

Hospital-Based Organizations Moving Forward, Too

The landscape looks a bit different for the leaders of physician organization divisions of hospitals and integrated health systems, but the journeys certainly parallel each other. At the vast 21-plus-hospital UPMC health system in Pittsburgh, Oscar Marroquin, M.D., the health system’s chief analytics officer in its health services division (which includes its hospitals and its physician groups), notes that “We’ve been using our analytics infrastructure and teams to leverage the information that lives in our electronic medical records as well as in our billing systems, to understand patient flow, what the phenotypes are of the patients our PCPs care for, what the distribution is of risk—when we manage patients on the outpatient side, to describe the level of health risk. We’ve done a fair amount of work in that space, that has led to identification of opportunities related to which practices are the best-performing ones in terms of their ability to keep patients healthier and out of hospitals, versus others, and that has allowed us the opportunity to dig deeper to understand why,” Dr. Marroquin says.

One very interesting example at UPMC is around the inflammatory bowel disease (IBD) patient-centered medical home that’s been created at the health system, and which has brought together UPMC gastroenterologists, psychiatrists, social workers, and other clinicians, to help optimize the management of IBD. The program, led by an IBD-specializing gastroenterologist, has reduced both inpatient hospitalizations and ED visits; and its PCMH (patient-centered medical home) focus has allowed it to track patients longitudinally in ways that had not previously been possible, Marroquin notes.

Even at hospital organizations that are near the beginnings of their journeys into risk, learnings are emerging. For example, the 208-bed St. Joseph Hospital in Nashua, New Hampshire, is participating in the Transforming Clinical Practice Initiative (TCPI) out of the Center for Medicare & Medicaid Innovation (CMMI) at the Centers for Medicare and Medicaid Services (CMS). And some of the innovative work that leaders at St. Joseph are pursuing had already been moving forward when the hospital was participating briefly in the Pioneer ACO program, reports Becky Williams, R.N., the hospital’s care coordination manager. For example, Williams notes, she and her colleagues have helped to universalize depression screening on the part of the 80 physicians in the hospital’s physician network, which has improved depression diagnosis and early intervention. She is hoping that the hospital’s participation in the TCPI program yields many additional advances in the near future.

Advice for Healthcare IT Leaders

When it comes to offering advice to healthcare IT leaders, including CIOs and CMIOs, those interviewed for this article are in agreement: all of the IT infrastructure, analytics, dashboarding, and other work that physician organizations implement must be done as part of broader, consensus-driven strategic efforts. “Information technology supports operations, as opposed to the other way around,” says Premier Medical Associates’ DeRubeis. “What I’ve seen many organizations do is to turn that over to IT and say, OK, you put it in for us, you plug it in, and we’ll come along afterwards. And in our experience, you don’t get the type of engagement from the physicians, when you do it that way.”

Crystal Run’s Hines adds that “One of the first things to acknowledge is that the data’s never going to be perfect. And particularly early on in the journey, when you’re just starting dashboards for the physicians, for example, emphasize that this is the first version of this, and it will be tweaked over time, so encourage the providers to look at their data and assess their data, and find any faults in the data to make it stronger in the end. And second, try to make the data of delivery as simple as possible—whether it’s a speedometer type of thing, or color-coded like green-yellow-red, so that literally, in one screenshot, providers can see their performance.”

Finally, says Innovation Care Partners’ Whitfill, “With regard to the one-foot-on-the-dock situation, it’s going to be hard to work this out” in the next several years, as physician groups continue to function partly in the fee-for-service world, even as they move further into risk-based contracting. In that regard, he says, “Any IT-facilitated improvements you can make to efficiency and workflow will be a benefit to your physicians, regardless of whether you’re being paid fee-for-service or fee-for-value. And the same thing is true of analytics: being able to navigate both worlds will increasingly require data scientist-level expertise. And we’ve made investments in the EHR over the last 10 years. The next step has to be making efficiency in workflow, and analytics to make sure we’re going in the right direction.”