Now that the final rule has been released for the Centers for Medicare & Medicaid (CMS) Shared Savings Program, what’s in store for the industry? HCI Associate Editor Jennifer Prestigiacomo spoke with American Medical Group Association’s John Cuddeback, M.D., Ph.D., chief medical informatics officer of Anceta, AMGA’s collaborative data warehouse [AMGA is based in Alexandria, Va.], to see what was in store for development of accountable care organizations (ACOs) this year and what care coordination tools will be necessary to lay the groundwork.
What do you think about the ACO final rule? Will it be enough of a catalyst to encourage organizations to go down this path?
In thinking about how the industry will change models over time, this is only one component. There are a lot of our members that are actually talking with their commercial payers about things like an ACO. It may not be exactly the same model, but there is some element of shared savings.
When we talk to medical groups, we hear lots of different models being considered, but they’re all in the same direction. I think the direction of focusing on and managing population health effectively, and sharing savings that result, is very much behind the movement we see in the market.
John Cuddeback, M.D., Ph.D.
Do you think CMS is doing enough to encourage ACOs?
This is a big battleship we’re turning around here. If you look at what the U.S. spends on healthcare, it’s the world’s fifth largest economy. So, it doesn’t change on a dime. It has a lot of infrastructure needs. I think if we look at conceptually at where the direction of where the market is going, I think the recognition that we need to focus on managing population health and paying for population health, as opposed to paying for individual services, is a very clear direction. I think most of our members are very comfortable with that, and that is precisely what AMGA has been advocating for almost a decade.
Much of the toolset available in the industry is focused around the care of an individual patient putting clinical decision support effectively into an EHR, which we’ve all found is a lot harder than it looks, because of the problem of alert fatigue and all of those subtle issues. That is the toolset that much of the industry is still trying to implement. More than 90 percent of our members are already using EHRs, and the majority of those are doing e-prescribing. I think a good model for an ACO was the Physician Group Practice Demo Program. It was a tremendous learning experience that CMS facilitated for the industry as a whole. Of their 10 participants, nine were medical groups, and all were AMGA members.
People talk a lot about care coordination and patient-centered medical home, which is a model that we absolutely support. If you look at the typical Medicare patient, who sees seven physicians a year, two different primary care physicians and five specialists, [you realize]that this is a fragmented system. The number of physicians a typical primary care physician needs to coordinate care with is 229 physicians and 117 other practices; these numbers were published in the Annals of Internal Medicine. And if you just look at the one-third who have four or more chronic conditions—because those area the really expensive patients where care coordination is really important—the median number is still 86 physicians and 36 practices. So the net of all of this is that AMGA members are not simply attacking this with just informatics, they’re putting people in place whose job it is to coordinate care. It could be something as simple as planned visits, thinking through what’s going to happen at this next visit for this patient. When we talk about design of a care system is not only having the primary care and specialist physicians under the same roof, but being able to have those tools in place for people to use, and most importantly, having the people there. HIEs are a very valuable technology, but simply connecting EHRs—even if we solve the EHR adoption problem—is not going to automatically coordinate care.
How many organizations do you think are going to sign up for the Shared Savings Program in the next three years?
I think to some extent it will depend on how the commercial market evolves. Because some groups feel they can get their feet wet in population health management with maybe a little lower risk and less exposure if they approach it on a smaller scale with some of their commercial payers. The other thing that a lot of our groups are doing is working with employers in their regions. Those are people who tend to be interested in and be able to effect fairly broad interventions, including workplace health interventions and initiatives. I think that our groups are looking at all of those opportunities. There’s no question of long-term direction. The question is what is the best opportunity that each group has in its local market to learn how to do this using comparative data like that we’re providing and the shared learning. The model for how this team-based process works is still evolving. There are some very good best practices out there. For example, you can integrate with an employer and have a connection between the primary care that a population is receiving and a workplace health initiative. That’s a great opportunity.
What do you predict that organizations are going to be focusing this year and next to create ACOs to coordinate care?
When AMGA started its advocacy for the change from volume-based to value-based payment, they also recognized that medical groups need some help in getting and managing population health. This is a new competency for the healthcare system as a whole. A lot of medical groups are in need of help and opportunities to work with each other. One of things that has been really valuable for AMGA is the opportunity to bring people together to learn from each other. We have currently a collaborative for managing patients with multiple chronic diseases— what are the tools you need to have in place to manage these complex patients? We have an advisory steering committee for that collaborative and an interesting story came out of that first meeting when they were trying to develop the framework. And at the end of the meeting, the participants observed that no [organization there] had all of the processes in place that they had collectively defined as being necessary to care for these complex patients. That I think is a big part of what AMGA has done, bringing people together to learn from each other.
Over the past few years we have made that more data-driven by developing a collaborative data warehouse [Anceta], which is focused on the problem of managing population health and giving medical groups very detailed comparative data that they can explore for any sub group of their patients on how they are doing relative to other AMGA members and finding where there are members doing a better job and taking the opportunity to learn from them.
We have a hypertension collaborative that was originally chartered for two years, and the groups didn’t want to stop meeting. They felt they were getting a lot out of it, so that collaborative ended up running for four years. So it’s that sort of opportunity for people to come together and learn from each other about the little nitty gritty details. One of the things someone commented on was it’s not so much that we learn exactly what the right model is from each other, is that we learn what hasn’t worked from other people and we can avoid a false start.
What care coordination tools will be instrumental this year to lay the groundwork for ACOs?
The major emphasis that I see and hear about when I talk to the folks in the medical groups is, as organizations are coming together with heterogeneous systems, it’s figuring out how to get all of that together, whether it’s an organizational-level HIE [health information exchange], or some way of doing data integration. It is clearly important, whether it needs to be at the transaction level as the HIEs would enable, or whether it can be at the data integration level, which several of the groups were working at with at Anceta. We’re getting data from different parts of their organizations. We’re actually we’re bringing the integrated picture of their own data, but also the comparative data. So that’s one, bringing things in for data integration.
Another is reaching out to patients, whether it’s through smart pill bottles, or automated scales for renal impairment and congestive heart failure, or whether it’s behavioral interventions, or a personal health record. The idea of taking advantage of a whole broad range of opportunities we have to interact with patients outside of a traditional visit or phone call is a second theme. A lot of our groups are talking explicitly about how they’re designing some of their care process around encouraging patients and how they’re using psychometric tools to measure patient engagement.
The third theme is population health analytics, because everything we’ve got has been focused around the care of an individual patient—all the EHR work extending, all of the clinical decision support. So I think the ability to extend clinical decision support, whether it’s through comparative effectiveness research or something a little less rigorous, but to understand what really works for patients to manage complex patients. It’s much more holistic—what should we do for this patient, what engagement activities, what social support systems, understanding what the constraints are about this patient interacting with the health system. I think that is the value around the collaboration around the comparative data, understanding what those opportunities are and be able to optimize those very expensive patients’ care.
What are the biggest barriers for organization to participate in an ACO program? What do you think organizations will have the most challenge with?
I think probably the biggest challenge is managing in a mixed reimbursement model. If we could switch overnight or if we could switch at a known date in the future and everyone could prepare for the date and if everyone was going to be in a pay-per-value mode, then I think it would be clear what your roadmap was. There would be various pathways through that transition, but people would have a pretty clear roadmap. We lost a lot of steam with draft regs, and people we’re very excited about this transition in payment. Somehow providers need to get some exposure to the overall population cost of care.
You have to change the system so that the tools to achieve the efficiencies are in place at the front line where care is actually delivered and the decision-making is made at the front line. I think the biggest barrier is getting from here to there, and the fact that markets are going to evolve differently. We’ve got markets like the Pacific Northwest, whose markets are in a very different position than the same market in the D.C. region, and these were folks in the PGP Demo who were already up the curve and they’re still struggling. And then there are other markets that are just continuing fee for service and it’s hard to get traction for anything else and people are even struggling to gain airtime to talk about anything other than fee for service. It’s that heterogeneity of market that exists and the transitions that will occur individually, but I think the direction is clear.