Editor’s Note: Today, The Health Affairs Blog begins a series of four posts on trends in performance measurement and performance-based payment in health care. The series focuses particularly on the increasing emphasis being placed on measuring and rewarding cost-efficiency. Today, James Robinson (below) and Tom Williams contribute posts. On Monday, Arnold Milstein and Howard Beckman weigh in.

The U.S. health care system is in such desperate shape that every initiative, be it in policy or management, risks going through the cycle of illusion and disillusion, the inadvertent creation of excessive expectations and the inevitable failure to achieve the unachievable. (We tend to expect, for example, that initiatives simultaneously increase quality, reduce costs, avoid restrictions on consumer choice, and enhance pay levels to providers). The pay-for-performance (P4P) initiative faces just this risk as it moves from its hopeful birth as a bright shining idea (that America should pay for performance and not, say, for nonperformance) and begins to toddle out into the real world of health care, where unlimited expectations collide with limited resources. Skeptics have discovered that P4P has put a few dollars on the table and prompted modest improvements in information technology (IT) adoption and some even more modest improvements in preventive measures, but it has not cured diabetes or led to universal health insurance. Surprise.

The most important challenge facing the pay-for-performance movement is that the health polity has rediscovered the virtues of cost control (as a component of sustainable insurance expansion). P4P was born during the nation’s backlash against the cost control emphasis of managed care, and hence restricted its focus to quality, patient satisfaction, and, in some instances, adoption of IT. Today the talk is all about “value,” a felicitous term that means many different things to many different people but that, for present purposes, suggests that cost moderation is a virtue in addition to quality enhancement and that payment methods should encourage the former as well as the latter. The “old wine in new bottles” dimension of this is so obvious as to be beneath comment. In any case, P4P now is challenged to offer an alternative to the cost-increasing incentives in fee-for-service and not simply an alternative to the quality-endangering incentives in capitation (not that fee-for-service doesn’t endanger quality as well; type “anemia drug” or “spine surgery” or “radiation risk in radiology” into your favorite search engine).

The P4P Movement: Keeping The Focus On Payment Amidst “Binge Drinking At The Well Of Consumer-Driven Care” 

The best part of the pay-for-performance phenomenon, in my opinion, has never been its (modest) potential to directly influence quality or satisfaction or anything else but, rather, its ability to keep alive the debate over appropriate methods of payment during a decade of binge drinking at the well of consumer-driven health care. We are now almost at the decade mark of this mindless moment in which the pundits somehow convinced a large swath of policymakers and private payers that managing the health care muddle demanded only consumer cost-sharing incentives, without attention to how providers are paid and how they respond to payment incentives. Pay-for-performance kept the light trained on the physicians and the hospitals and, through its progeny “value-based purchasing,” on some elements of the drug, device, and ancillary sectors as well.

Having said these nice things about pay-for-performance, it’s time to say, Let’s move on to version 2.0 and not just celebrate (or castigate) the original template. I will leave the quality focus to the quality experts, but I have a few words to say about the cost or efficiency side of the value equation.

Most obviously — and I think that Sutton’s Law applies here — if you want to save money, you have to follow the money. P4P 1.0, bless its heart, was much about primary care. Version 2.0 needs to look at specialty physicians and hospitals and, alongside those categories, at the drugs, devices, imaging, and other technologies that drive most of the cost growth as well as some of the quality improvement in health care. Here, of course, P4P merges into the old-wine-in-new-bottles debate over the virtues of “episode pricing” as a payment method that bundles the physician, facility, and technology components of each course of clinical treatment without shifting to providers the epidemiologic risk inherent in capitation. It would also prompt a review of the historical split between physician payment and hospital payment for Medicare, a split that the managed care plans sought to overcome through capitation but to which they have defaulted after the successful provider counterattack.

Version 2.0 of pay-for-performance also merges into the debate over appropriate incentives for physicians in their role as choosing particular drugs and devices for their patients. This new-again debate encompasses Medicare’s “gainsharing” ban for physician-hospital partnerships, the “buy and bill” payment method for cancer drugs, the proposed two-part physician payment method for the “medical home,” the annual congressional override of the cost-controlling dimensions of Medicare’s “SGR” (sustainable growth rate) controls on the physician fee schedule, and the other peculiar fauna of the health care payment ecosystem.

P4P Version 2.0: Balancing Quality, Affordability, Consumer Choice, & Innovation

What is the long-term agenda for pay-for-performance, and for payment design more generally? Let’s start with the obvious. The goal of P4P is not to improve quality (only). It is not to improve patient satisfaction or IT adoption (only). It certainly is not to moderate cost growth (only). The desired structure of provider payment is that which best helps us balance the competing virtues of quality improvement, economic affordability, respect for consumer choice and values, and biomedical innovation. Provider payment is not our only tool for that job, so another desired element is its compatibility with consumer incentives, regulatory mandates, tort law, subsidies for capacity expansion and scientific research, and all of the other factors that influence health care performance. And did I say simplicity? The desired method of provider payment is one that providers can understand and that payers can administer at something less than the contemporary horrific administrative cost of health care.

And if we are allowed to dream, I would add that a well-designed, performance-based payment system could exert positive effects on the long-term dynamics of the health care system by creating a business case for innovations that reduce, rather than increase, the cost of care. Current reimbursement incentives reward new drugs, devices, and procedures that improve quality, but, with some limited exceptions (Medicare’s diagnosis-related group, or DRG system being one) do not reward innovations that reduce costs. Fee-for-service, in particular, imposes a 100% tax on innovations in physician practice style that reduce the need for visits, tests, and procedures. And don’t get me started on how hospitals pay for surgical implants.

America will always be willing to pay high (shall I say “value-based”?) prices for breakthrough innovations that raise the costs of care, even substantially, while creating dramatic improvements in life expectancy or functional ability for patients. But these desirable expenditure increases need to be offset, at least partially, by cost reductions elsewhere in the system. Pay-for-performance is based on the premise that incentives matter. I long to see a system in which entrepreneurs and providers can get rich by making the American health care system more affordable and hence more accessible to all. And even if incentives don’t matter, I would prefer a payment method that gave more to efficient, high-quality providers than to inefficient, low-quality providers. Imagine.