Pay for Performance: Where is the Evidence?
One of the greatest accomplishments of the Affordable Care Act is that it has helped in shining a light on the disconnects of American health care. This huge, complex, expensive, stakeholder written behemoth is demonstrating our inability to lessen costs, improve the payment system, cultivate quality and provide equity in the delivery of health care.
The problems of the existing fee-for-service method of compensating physicians has been discussed in an earlier blog.1 So called “pay-for-performance” (P4P) is the least effective solution being offered by the health policy wonks.
Now we are involved in the old past-time of debating how to pay-for performance. The ACA is toying with Accountable Care Organizations (ACO’s), physician quality reporting, value bases purchasing and other cost control attempts masquerading as quality initiatives. In addition to the Centers for Medicare and Medicaid Services (CMS), state Medicaid programs and the private sector have all developed numerous P4P experiments.
As we shall see, the inherent weaknesses of the P4P concept are playing out in the thoughtless drive to pursue the latest iterations of this fad. We will look at the difficulties and dangers of trying to define and codify “quality” and “value.” An excellent article on this subject can be found at Health Policy Brief: Pay-for-Performance, Health Affairs, October 11, 2012.
In order to pay for performance you have to be able to measure it. You have to have numbers that you can translate into dollars. This immediately brings up the old adage (often falsely attributed to Einstein), “Not everything that counts can be counted, and not everything that can be counted counts.” When physician performance is being evaluated we most often see an old list of processes, such as laboratory tests ordered for a certain diagnosis, screening tests, blood pressure recordings, smoking counseling, aspirin, etc. These processes have very little real relationship to the totality of what a good physician does for his patients. They are items that are easy to monitor once you have forced the physicians to pay for an expensive electronic medical record system. But they are meaningless, costly, and do not refer significantly to a paradigm of individual patient care.
That being said, numerous studies have shown no lasting, meaningful improvement with various reward/penalty systems of this type. Rosenthal and Frank stated that “The authors review the empirical literature on paying for quality in health care and comparable interventions in other sectors. They find little evidence to support the effectiveness of paying for quality.”2 Ruand and Blustein reported “Despite offering substantial financial incentives, the MassHealth P4P program did not improve quality in the first years of implementation.”3 Campbell et al reported from Britain that a study of family practices showed no differences in performance between the controls and those provided incentive payments. In fact, continuity of care had decreased.4 U.S. hospital P4P was equally ineffective in the large (780 hospitals) Premier study.5 The respected Cochrane Organization did an evidence-based review of multiple studies and concluded that there is no good evidence that pay for performance schemes increased quality.6
In addition to process-directed efforts we occasionally see outcome oriented plans. These attempt to focus on more meaningful outcomes of medical interference such as death, strokes, heart attacks, etc.
They are even more difficult to measure for individual physicians and their patients because outcomes are so dependent on co-existing illnesses, multiple physician involvement, patient cooperation, socio-economic class, etc. These elements are the data of the risk-management discipline which insurance companies developed decades ago. If there are rewards for caring for high risk patients then many more patients find themselves categorized as high risk. If there is income loss because a system does not compensate for the high cost of high risk patients then those patients will face loss of access to medical care. There is no incentive in any system that depends on metrics that are not under the control of the physician. The latter includes many socioeconomic factors that decrease the effectiveness of medical care.7
Both process and outcome definitions of performance simply set new rules for gaming the system. They do not assure quality and they do not diminish costs. Cherry-picking is one of the numerous ways to stack the deck. For physicians this could mean continuously culling one’s practice to exclude patients who do not contribute to a high performance score. Up-coding is another well-known method of increasing payments. As Woolhandler writes, ”the system is too easy to game–too hard to set up.” 8
Complexity of Players
A 2007 study showed that Medicare patient visits revealed that a median of two primary care and five specialty physicians provide care for a single patient. It is difficult to see how a P4P system can assign responsibility and rewards for these patients. This doesn’t count the effect of urgent care and emergency room visits that effect the parameters being measured by someone’s incentive profile. 9
Woolhandler and Ariely 10 give an excellent discussion of some of the lessons of behavioral economics. Of special interest in the evaluation of the concept of P4P are the findings concerning motivational crowd-out. This refers to the numerous studies showing that tangible rewards (like money) have a strong ability to “crowd-out” or destroy intrinsic motivation (doing something for its own sake).11
With our present day complex, uncoordinated, multiple payer health system all of the above writings, research, arguments and theories will not lead to any method for lowering medical costs or improving quality. A single payer system would have the weight necessary to establish a payment system flexible enough to provide fair compensation for physicians without the inherent weaknesses discussed above. We are seeing a strong physician movement toward salaried positions as hospitals and large clinics hire an ever larger percentage of both new and older physicians. Over time this may turn out to be the key that frees physicians to concentrate on patient care. For primary care this will require vigilance to avoid gate-keeper roles with restrictive cost monitoring. Then the incentives can truly focus on quality. A single payer system would be able to devise an appropriate mix of payment systems to fit the various practice settings in which physicians work.
4) Campbell, S. M., Reeves, D., Kontopantelis, E., Sibbald, B., & Roland, M. (January 01, 2009). Effects of pay for performance on the quality of primary care in England. The New England Journal of Medicine, 361, 4, 368-78.
5) Ashish K. Jha, M.D., M.P.H., Karen E. Joynt, M.D., M.P.H., E. John Orav, Ph.D., and Arnold M. Epstein, M.D. , The Long-Term Effect of Premier Pay for Performance on Patient Outcomes.N Engl J Med 2012; 366:1606-1615 April 26, 2012
6) Scott A, Sivey P, Ait Ouakrim D, Willenberg L, Naccarella L, Furler J, Young D. The effect of financial incentives on the quality of health care provided by primary care physicians. Cochrane Database of Systematic Reviews 2011, Issue 9. Art. No.: CD008451
11) A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Deci, E.L., Koestner, R., Ryan, R.M. Psycol Bull. 1999 Nov;125(6): 627-68; discussion 692-700.