One of the most frequently asked questions in discussions of health care reform is whether health care organizations and the country as a whole are reaching the “tipping point” for meaningful change in how we pay for and deliver care.
It has become conventional wisdom among policymakers and health care leaders that around 30% is the magic figure. For example, it is argued that unless providers receive around 30% of their revenue from risk-based, value-based contracts (versus fee-for-service), there is little motivation for them to spend the time, effort, and resources to fundamentally change how they deliver care. Anything less does not send a strong enough signal. Thirty percent begins to get everyone’s attention that something real may be occurring.
But where does this 30% figure come from? It may have originated from two large-scale studies of integrated delivery systems that I and colleagues conducted in the mid-to-late-1990s. In this research we examined the strategies that ten leading systems were using to achieve greater functional, physician, and clinical integration. As a result of numerous interviews with clinical and administrative leaders within these systems, “we found as a general rule of thumb that approximately 30% of a physician’s practice needed to come from a single source before a physician would consider adopting that source’s recommended care management practices.” We went on to note that “thresholds” appear to be as important as withholds in modern medical management.
We subsequently used the 30% figure in presentations across the country and found it appeared to resonate with a variety of audiences. Apparently if you say something often enough, it takes on a life of its own; a self-fulfilling prophecy! At approximately the same time, two of our subsequent colleagues, Larry Casalino and Jamie Robinson, conducted field interviews of medical groups in California and came up with the same figure of around 30% before those physicians would make significant changes to develop their managed care capabilities. So 30% was on a roll!
But aside from these observation-based field studies, is there any systematic empirical evidence for the 30% threshold number? To determine a figure, ideally one would randomize stratified practices to different payment threshold conditions and observe their behavior over a given period of time. Or, in a weaker design, observe a group of practices exposed over time to increasing risk-based payment and see if there is a significant inflexion point on various behavioral change measures (such as systematic use of patient engagement strategies) at around 30%.
Lacking such studies, one might compare the behavior of practices belonging to accountable care organizations, which have inherent incentives to deliver efficient value-based care, with those practices not a part of an ACO. Data from the National Survey of Physician Organizations (NSPO3) revealed that those which were part of ACOs indeed scored significantly higher on an index of Patient-Centered Medical Home processes, reflecting better care than those not belonging to an ACO (53 points out of 100 versus 32 points). But we do not know the exact percentage of revenue at risk under such contracts. To approximate this figure, we asked all 1,398 practices in the survey to estimate the percentage of risk that each absorbed for primary care, specialty care, and hospital costs, and then compared these percentages to their score on the PCMH index. We found the biggest change occurred in the 21–30% range, where the PCMH score increased significantly — from 36 points in the under 20% category to 43 points in the 21–30% range. Interestingly, there was no further increase beyond 30%.
So while there may be some validity to the 30% figure, it is important to recognize the varying contexts in which it may occur. Important considerations include:
- The amount of incentive involved — that is, 30% of your revenue may come from value-based savings, bonuses, or incentives, but if the amount of dollars you get to keep is small, it may be insufficient to motivate changes in behavior.
- Transparency of external data reporting, in which one’s medical practice is publicly compared with others on quality and cost metrics serving as a motivator for change.
- The presence of individual comparative physician data feedback on quality and cost metrics that appeals to physicians’ “intrinsic motivation” to improve care.
- The extent to which the practice is located in a more competitive market, which may induce more motivation to change to retain current patients while attracting additional patients.
Realistically, these elements are likely to combine to generate the most extensive and sustainable changes.
So here we are, 20 years later from the likely origin of the 30% rule, with CMS announcing that 30% of Medicare payments will be based on alternative value-based payment models by the end of this year, and 50% by the end of 2018. If implemented, the Medicare Access and CHIP Reauthorization Act (MACRA) will provide a merit-based payment system and alternative payment models for all physicians, thereby reinforcing these target figures. They ultimately may be off by a little — but as Sir Archie Cochrane, the renowned British epidemiologist after whom The Cochrane Collaboration is named, once said: “It is better to be roughly right than precisely wrong”!
This article originally appeared in NEJM Catalyst on February 25, 2016.