Subscribe if you want to be notified of new blog posts. You will receive an email confirming your subscription.
Today’s BFO: How can P4P Work W/O a QB?
Translation Todays blinding flash of the obvious (BFO): How can you expect pay-for-performance (P4P) programs in Medicare to work with out a designated physician quarterback (QB)?
Please allow me to elaborate.
P4P programs are based on two assumptions:
- Patients are assigned to a physician or a practice that will have primary responsibility for their care, and
- That a meaningful fraction of the care physicians deliver is for patients from whom they have primary responsibility
Wouldn’t you expect that this would be problematic for older (Medicare) patients who see multiple doctors over time? How can you assign accountability for performance to one doctor when the patient is seeing a number of doctors for a number of care episodes? (That’s the BFO part for me.) and wouldn’t you expect this to be even more problematic for patients with multiple,chronic conditions?
Shouldn’t somebody study this to figure out whether P4P is doable in Medicare? Well, they have.
An article in the latest New England Journal of Medicine concludes that the dispersion of patients care among multiple physicians will limit the effectiveness of pay-for-performance initiatives that rely on a single retrospective method of assigning responsibility for patient care.
Here are a few of the specifics that Pham et. al found in their research:
- The average Medicare patient saw a median of 2 primary care physicians (PCPs) and 5 specialists over the course of a given year, collectively working in 4 different practices
- Dispersion of care was even greater for patients with chronic conditions. For example, patients with diabetes saw a median of 3 PCPs and 6 specialists over the course of a year.
- The assigned physician billed for a median of only 35% of total physician visits for each beneficiary
- An average of 33% of beneficiaries had a change in their assigned physician from year 1 to year 2 and from year 2 to year 3; only 46% of beneficiaries consistently were assigned the same physician for all 3 years of the study.
Bottom line: how can you assign P4P bonuses to one physician when the patient is seeing a bunch of physicians for a bunch of conditions? You can’t.
In an accompanying editorial, Karen Davis of the Commonwealth Fund suggests that “pay for performance is unlikely to fundamentally alter the incentives in the fee-for-service payment system. Ideally it would serve as an interim in the transition to fundamental payment reform.â€
She discusses a number of potential alternatives to ensure a stable and responsible primary source of care and to improve care coordination. Alternatives include a prospective designation of physicians responsible for a patient’s care (e.g., a medical home model); and encouraging growth of integrated delivery systems, large physician group practices, or networks of physician practices.
P4P arguably makes sense when you can assign primary responsibility for an episode of care to one doctor (or practice). When I had a mountain biking accident last year and broke my shoulder, my orthopaedic surgeon clearly was the one physician who steered me through the system for surgery, physical therapy, another surgery, and continuing recovery. P4P could be pretty straightforward when episodes of care can be clearly defined. But Medicare is a different story.
The Pham et. al. study also is a bucket of cold water poured on Harvard Prof. Michael Porter. In his recent book Redefining Healthcare, Porter describes principles of value based competition. One foundational principle is that competition should center on medical conditions over the full cycle of care.
Phams findings about dispersion of care among Medicare patients also are a reality check for Porter, whose book has received mixed reviews. Arguably his principle might work for a clearly defined episode of care such as a broken shoulder. However as care is increasingly dispersed for older, Medicare patients a full cycle of care becomes progessively more difficult to define for patients with ongoing, chronic conditions.
So, back to the beginning. How can P4P work without a quarterback? It can’t.
This work is licensed under a Creative Commons Attribution-Share Alike 3.0 Unported License. Feel free to republish this post with attribution.
Discussing the merits and pitfalls of P4P in isolation strikes me as akin to determining how a finished cake is going to taste by sampling the recipe ingredients before baking; not very satisfying, in either case.
Once upon a time in a land not so far away, there began a grand experiment called the California model HMO, with large, multispecialty physician groups receiving capitation for provide comprehensive care for a defined population. Assumed by many today to be a complete failure and all but forgotten by pundits and politicos in current pre-election debates, there are many aspects of that model of care delivery that were quite successful. Notably, accountable organizations accepted responsibility for overall care of individuals, sidestepping the attributional concerns raised by Pham and David in your posting, at least at the organizational level.
This context allowed for quality measurement (both process and outcomes measures), confidential quality reporting to providers, and eventually (1998, in the case of PacifiCares Quality Index) public reporting of quality performance at the group level, and paved the way for P4P in California, predominantly through the IHA collaboration among the majority of payers and physician groups in the state. Accountable health delivery organizations, coupled with prepayment, rigorous quality measurement, and public transparency of quality performance across multiple dimensions of care created a context for P4P to become a very meaningful program, not just a conceptual debate or token experiment. (See http://www.iha.org)
The IHA P4P program, which began in 2002, continues today and is rewarding quality improvement of provider groups in CA to the tune of $60 million per year, with an estimated cumulative payout over $200 million since inception. As implemented, P4P has fueled significant investments in infrastructure and process of care changes that have lifted all boats on a rising time of quality improvement across California specifically, 8 large health plans, and 225 provider groups with 40,000 physicians caring for 12 million HMO members are involved each year. Clearly, the debate about what to measure and who is accountable for P4P is largely over in this environment.
The argument for defining care at the accountable organization level, rather than individual physician level, was not only related to the many measurement, attribution, and small N dilemmas that still plague the latter; it was also deliberate recognition of the fact that in many cases, quality improvement needed to be a system property rather than the sole responsibility of each individual physician. The development of programs and systems of care ranging from hospitalist programs and same-day access to ambulatory care, to computerized disease registries and disease management programs designed to improve the effectiveness, the efficiency, and the experience of care delivery was the responsibility of the accountable provider organization. How those groups choose to distribute their P4P funding to infrastructure vs. programs vs. individual providers is up to them, and each undoubtedly uses differing logic for their distributional plans.
So in my view, our principal debate now should be not Whether P4P but rather How P4P? Incentive payments cannot be viewed as a panacea for all the frailties of our fragmented delivery system, but deserves a place in context with other important changes. Specifically, how do we couple reimbursement reforms like P4P with comprehensive quality measurement and reporting, greater HIT investment and deployment of EHR/PHRs, better consumer engagement and increased responsibility for self-care and healthy lifestyles, in order to align societal, consumer, and provider incentives, and reward behaviors individual and organizational that raise the value of healthcare? And how do we do all that in a manner that maximizes impact, while minimizing unintended consequences (gaming, perverse incentives, winner-take-all) and assuring reasonable distributive justice at the same time?
The California delivery model may seem to many a curious anachronism with few applicable lessons to todays healthcare reform issues. However, important insights have emerged from nearly two decades of experience with that model (which, incidentally, is alive and well in many West coast communities today). The takeaway lesson is not that somehow we can reconfigure all healthcare delivery in the U.S. to be conducted by large, multispecialty practices on a prepaid basis (though debating that proposition might be an interesting exercise), but rather that P4P context and P4P execution are inextricably bound together. Quality is a system property. In the current P4P debate, we ought to be as mindful of the context as we are of the specific measures and attributional challenges. Optimizing both is a necessary condition for P4P success, in my opinion.
Gordon,
Thanks for your thoughtful comments. Everything you say makes sense and makes me realize that my POV needs to be articulated more precisely.
Of course, you are right in pointing out that accountability can (and probably should) be designated at a system level, not that of an individual physician. This is also particularly important for patients with chronic conditions, whose ongoing care is more dependent on the system being there in the first place in place rather than on treatment by an individual physician.
Thus, my use of “QB” doesn’t convey quite the right picture; the QB should not be a retrospectively designated individual physician, but certainly could be a prospectively designated organization (e.g., group practice).
I’d also agree that the debate should be about “how P4P”, not “whether P4P”.
So the “aha” from the NEJM article is more about Medicare’s current attempts at P4P and the need to designate accountability prospectively, rather than retrospectively. The “aha” is also about a need fundamentally to rethink reimbursement, and that tinkering with P4P alone doesn’t address the root problems of an outdated reimbursement model.