At the recent Health Forum–American Hospital Association Leadership Summit, much of the discussion by keynoters and learning-track faculty alike was on the coming transformation in health care toward paying for value. Newly announced mandatory Centers for Medicare & Medicaid Services programs for bundled payment in orthopedics added fuel to the volume-to-value wildfire.
But what exactly is the right vision? There is growing skepticism among many respected industry experts who question whether population health, providers at risk and accountable care organizations are really the right answer. They fear these models may all turn out to be a bridge too far. Instead, they argue, we should get the basics of health care delivery right first. Then we should use bundled payment–type models as our lead foray into financial incentives that promote improved care coordination and clinical performance delivered by focused, high-performing teams.
At the Summit, Michael Porter from Harvard Business School strongly argued that bundles are the right path to value. Similarly, my pals Nathan Kaufman and Jeff Goldsmith, for whom I have enormous respect and with whom I nearly always agree, clearly articulated in recent writings their concerns with the at-risk vision based on their thoughtful analysis of recent ACO performance and their decades of research and strategic consulting with health care leaders.
I get their point, based on the rocky performance of CMS's ACOs (pioneers and shared savings model alike) and the meager overall improvements in cost and quality they have yielded after the substantial investment occurred. I also wholeheartedly agree that providers bearing risk is not for the fainthearted. I call it the mutual disrespect problem: In health care everyone thinks that everyone else's job is easy, and anyone can do what a health insurer does. Actually, it's not that easy.
So, there is a plausible (and perhaps even likely) conclusion that meaningfully incenting providers to deliver care by taking financial risk for a defined population they serve (across the continuum of care) is an impossible dream that will end in failure. Therefore, we should settle back on bundles and other less grandiose improvement initiatives instead.
I am not yet throwing in the towel on providers at risk. On the contrary, I still believe that our best hope for sustainable health care may well come from large integrated systems of care competing on the basis of cost and quality for a defined population.
But, like my more skeptical colleagues, I completely agree that this is difficult work for policymakers, health plans and the entire health care delivery system.
One Strategy or Many?
If you believe all the PowerPoint presentations you see in health care, it would seem at first glance that there is only one strategy for health care systems: Get big, take risk, look like Kaiser, only better.
Yet, many systems that are large, growing and respected, such as Virginia-Mason, Cleveland Clinic, Mayo Clinic, Yale-New Haven and others, are not going to be in the population health business even though they are the de facto provider of most health services for a large and identifiable chunk of their local population. And that strategy is perfectly fine, provided they can secure patient flow that is profitable for the services they provide. It further assumes that large self-insured employers will be willing to suck up the inevitable cost shift from Medicare and Medicaid over the next decade. Maybe it all will be fine forever and they can manage their cost structure to deliver value in the form of quality and patient experience for the clinical services they provide.
At the other extreme, there are actors who are "all in" on the path to risk. Advocate, Intermountain, Presbyterian and many others are committed to having a majority of revenues from at-risk lines of business including Medicare Advantage, Managed Medicaid, their own employee health plans, direct contracts with self-insured employers, and a variety of commercial and public payment accountable care arrangements. They, too, have their challenges. But the commitment of the leaders seems real, as evidenced by the public statements of CEOs and the signing of blood oaths by members of the Health Care Transformation Task Force early in 2015 (a group of private sector payers and providers who committed publicly to have a majority of their revenues derived from value-based payment by 2020).
For example, at the MedAssets CEO conference that I moderated earlier this year, polling of an audience of 200 C-suite executives from large integrated delivery surveys showed that a majority believed most of their revenue would be derived from value-based payments by 2020. In addition, more than 80 percent said they would have either an insurance product co-branded with an insurer (53 percent) or have their own insurance license (30 percent) by 2020. In a proprietary Nielsen Strategic Health Perspectives survey taken in 2015, 8 percent of hospitals said they are "committed to moving the majority of revenues to fully at risk within five years," up from 1 percent in 2014.
My take on these and other proprietary surveys is this:
Half are not moving. About half of hospitals are either doing little or taking a wait-and-see attitude. It is more prevalent to wait and see on the East Coast than on the West.
The movers are still early in the process. Of the half who are moving, most are pursuing clinical integration strategies and dabbling in ACOs and shared savings pilots so as not to be left behind. They are not all in, yet.
True risk bearers who are "all in" form a small but growing minority. I think the 8 percent number is right and reflects my experience that this is a real phenomenon, is rapidly growing and may account for a third of health systems in two or three years. And that would be enough to have a significant impact in the marketplace.
Most providers are not thrilled with payment innovations of any type. Nielsen surveys consistently show that almost all payment innovations — whether bundles, capitation, global payment or others — are not enthusiastically embraced by hospitals or physicians.
Finding Common Ground
I grew up in Glasgow, a city divided on sectarian lines in the '50s and '60s like Northern Ireland. As a Protestant, I didn't have a Catholic classmate until I went to university. Rangers and Celtic matches were tightly contested on the field and turned into running, bloody pitch battles in the streets afterward. It was senseless, stupid, counterproductive and energy-sucking. Sectarian division is not helpful.
I worry that we may be entering a sectarian battle in the health care system between bundlers and the true population health believers. I don't think a sectarian fight is a good idea, so let me propose some common ground:
Any big move away from open-ended, fee-for-service payment is good. Let's agree that any departure from unfettered fee-for-service that encourages providers to improve quality of care, advance outcomes and contain costs is a move in the right direction. We should celebrate all moves off the fee-for-service baseline.
It won't be either/or. The vast majority of hospitals and health systems will live in a world of mixed incentives for a long time. There will be bundles, pay for performance, gain-sharing and at-risk models mediated by insurers or by other intermediaries yet to emerge. Few institutions will have the luxury to be pure plays, so get comfortable with mixed models.
Respect each other's differences. There are legitimate advantages and disadvantages of bundles vs. risk-bearing models that I will touch on briefly, but let's just respect that each has something to offer. Let's not turn this into "my liturgy is better than your liturgy."
Develop thoughtful policy. The reason that many of the ACO and Pioneer programs had a rocky start is that there were flaws in the policies from the get-go. Some of us pointed them out in a tongue-in-cheek way at the outset (see my "Chasing Unicorns: The Future of ACOs"). Most notably, the fact that patients can go wherever they want without hindrance and that success was benchmarked against national norms, not relative performance improvement, each has been a major contributors to the pioneers' decision to turn back (see my column "Why Some ACO Pioneers Turned Back").
Bundles vs. Population Health Risk
Focusing on bundled payment may be a more acceptable strategy to many health systems because it is consistent with the central focus of most institutions, namely continued growth in clinical programs. More surgery is still better economically for hospitals in a world of bundles. When done right, incentives can be aligned to encourage improved care coordination across the continuum of care and, in turn, to improve quality and drive out some forms of waste and redundancy.
The good news is that, because you don't have to take any risk for the total costs of care for the population, you don't really have to worry about the frequency of services being provided. More is better. (One exception is the Wal-Mart/Pacific Business Group on Health center of excellence model where part of the bundled payment for cardiac surgery and joint replacement is for screening for the appropriateness of the cardiac and orthopedics procedures being contracted. This turns out to have a big effect on utilization, as 30 to 50 percent of cases are deemed inappropriate for surgery.)
The great advantage of bundles is that they can encourage improvement of teams on dimensions that they actually control and that patients care about.
The downside of bundles is that not all health conditions are as easy to bundle as joint replacement. Complex comorbid patients are trickier to bundle. But as Porter taught us all at the Summit, there are examples from other countries in which bundled payments have been designed to accommodate complex, chronic conditions.
Overall, my main beef with bundles is best exemplified by a quote from a health system CEO who told me: "Screw me on the bundle and I'll screw you on something else." This is a wise statement because it points to the potential for providers to raise charges on non-bundled items to accommodate for revenue lost.
Going at risk is different. The key distinction is that more is no longer better. Indeed, Kaiser and the capitated "delegated models" of California managed care learned decades ago that the primary goal was to minimize hospitalizations and procedures driving bed days per thousand enrollees to a level sometimes half of the normal fee for service population. When you are at risk, you are not trying to fill the hospital; you are trying to empty it.
But as we have learned in surveys and in the field, because there are very few pure-play, at-risk integrated delivery systems, the likelihood is that integrated systems will be in mixed payment models for some considerable time and will face conflicting incentives as a result.
While most systems will be in mixed payment models, many health systems (our surveys show about a third) are making significant investments in population health infrastructure and are experimenting and planning to take risk. When they run the math at the population level, they always run into the 5/50 problem. The reality is that, in any insurance pool, 5 percent of patients (the heavy users) account for 50 percent of costs; conversely, about 50 percent of patients use very little health care.
Hospitals have historic warm and fuzzy relations with the 5 percent. When you are an insurer, you'd prefer to hang with the 50 percent healthy folk.
As health systems pursue risk and dig into the analytics of whom those 5 percent of patients are and what conditions they have, they soon stumble across the following:
HONDAS. I was famously denied coverage before Obamacare kicked in (see "My Journey through Obamacare") because my insurer deemed me a HONDA (hypertensive, obese noncompliant diabetic alcoholic), none of which is true but is directionally correct. HONDAs show up on most lists of heavy users.
Behavioral health. In almost all insurance pools, from employer-sponsored coverage to Medicaid, the heavy users likely will have behavioral health issues.
End-of-life care. Particularly in the Medicare population, end-of-life care will be a big component of costs (approximately 27 percent); those patients at the end of life will be heavily represented in the 5 percent heavy users covered under Medicare.
Cancer. In any insurance pool, patients with cancer typically will be in the top 5 percent of heavy users.
Frail elderly. Again in the Medicare population, frail elderly will be among the heaviest users, and many of these patients are dual eligibles.
Specialty pharmaceuticals. The recent growth in new specialty pharmaceuticals with large price tags is having an impact on all insurance pools, from Medicaid to commercial. Indeed, the chief medical officer of one of the nation's largest insurers told me that this year, for the first time, any patient taking a biological likely would be in the top 5 percent of costs in the commercially insured population.
Social work, not medical care. As we have stressed in H&HN Daily many times before, the best management of many of these top 5 percent utilizers may end up looking a lot more like social work than medical care. Housing, transportation, income support, nutritional support and counseling may be more beneficial and effective than any form of conventional medical intervention.
If a health system is getting into the risk business, it has to be comfortable managing these issues. But it is not only the population health risk-takers who must confront them.
For example, expect more cancer care bundles from public and private payers that provide incentives to caregivers to manage the total costs of care, including specialty pharmaceuticals. Similarly, there will be considerable experimentation with dual-eligible payment models over the next decade as Medicaid expands massively in importance for the entire health care system. That experimentation might include bundled-payment pilots for the frail elderly or for patients with Alzheimer's.
In my view, it is not either risk or bundles; it's both. Let's get it right, though.
Final Observations on the Value Journey
Overall, my forecast can be summed up as follows:
Integrated systems with their own health plans, regional scale, direct contracting and Medicare Advantage contracts is the end game for some large players who are preparing for population health risk. Some, not all, will succeed.
Many hospitals will be caught between two paradigms for the next five years (at-risk vs. fee-for-service), but the direction is toward more risk-bearing on the basis of value through a variety of constantly evolving partnerships and risk-sharing arrangements.
Bundled payment for procedure-oriented care presents a major step toward promoting value and care coordination that does not require population health (frequency risk).
And finally: Value-based payment trends are not enthusiastically embraced by providers. So expect public payers to make more payment innovations mandatory, not just voluntary.