Pay for performance and physicians—an open question

Pay for performance as a payment strategy in health care has continued to gain substantial momentum over this past year. Pay for performance as a payment strategy in health care has continued to gain substantial momentum over this past year.

The recent push toward pay for performance has been spurred on by the rapid rates of increase in spending for health care, the continuing reports of large variations in spending for treating specific illnesses with little or no benefits associated with the higher cost treatments, and the all-too-familiar reports of quality problems in the United States despite higher rates of spending than any- where else in the developed world. Nonetheless, it remains to be seen whether pay for performance will be fully embraced by the U.S. healthcare system.

A First Step—Then a Left Turn?

H.R.6111, the Tax Relief and Health Care Act of 2006, passed just before Congress adjourned in December, provides an additional 1.5 percent fee increase for care provided by physicians under Medicare during the period July 1 through Dec. 31, 2007, provided they report on a set of quality measures. Pay for reporting is a useful first step and has proven popular when used for hospitals following the passage of the Medicare Modernization Act in 2003.

However, with the change of control in the Congress, and the substitution of Pete Stark (D-Calif.), a vocal opponent of pay for performance for physicians, in place of Nancy Johnson (R-Conn.) as chair of the health subcommittee of the Ways and Means Committee, further movement along the pay-for-performance track may be somewhat diminished. The future of pay for performance may well depend on the attitude of various physician groups, which has been at best, mixed, to date.

Some Challenges Ahead

As the 2006 Institute of Medicine report Rewarding Provider Perfomance: Aligning Incentives in Medicare made clear, pay for performance is only one of many steps that needs to occur in order for the U.S. health- care system to consistently deliver more efficient, high-quality, patient-centered care. Because so much is as yet unknown or underdeveloped, in terms of appropriate measures of quality and accountability, the IOM and most practitioners are adamant about the need to use a “phased” approach for pay for performance and to engage in what has been termed “active learning”—that is, adjusting measurements and payments in response to unintended or unexpected outcomes of changes in the payment system.

The question of the importance and appropriateness of starting this modification of the payment system with various institutional providers, such as hospitals, renal dialysis centers, and Medicare Advantage plans, has stirred up some controversy and resistance among the affected groups. But this reaction has been subdued compared with the resistance to the idea of using this approach to modify payment for physicians.

One reason for this greater resistance among physicians is easily understandable. Small rural practices face major challenges in providing needed measurements, and even larger practices that have some infrastructure to provide measurements may face reporting challenges.

Another possible reason for the greater resistance may be that, whereas small changes in payment can be expected to drive changes in behavior for institutional providers and therefore pose less risk, it is uncertain whether small changes in payment will affect physician behavior comparably.

This concern could be addressed, in principle, by making larger changes in payment than the 1.5 percent to 2 percent currently contemplated, along with bonus and reduced payments as part of a continuum of change. But the larger the change, the greater will be the demand for certainty about the likely effects of the payment modifications.

Therefore, starting with relatively small bonus payments for reporting remains the more prudent way to start, with the assumption that this approach will lead to small bonus payments for identified performance on quality and efficiency measures.

A Lesson from the Past

Past experience provides us with a reminder of what can occur if payment changes are made with little awareness of likely outcomes and indications of undesirable effects on the payment system. Consider, for example, the resource-based, relative-value scale combined with the volume performance standard (a variant of an expenditure cap), which was adopted without any preliminary trials and before any pilot or demonstration projects had been attempted.

The problems of the payment system prior to the RBRVS were well acknowledged:

  • High payments for some medical/surgical procedures
  • Low payments for many services provided by primary care physicians
  • Large differentials in payment between urban and rural areas
  • No strategy to moderate total spending on physicians

The concept of a relative value scale where all the Current Procedural Terminology codes used to pay physicians would be arrayed on a relative basis—sorted in terms of work effort, practice expense, and a geographic cost index—had some conceptual appeal. However, the legislation was passed, mandating implementation about two years after its passage, without the work even near completion and without any of the individual components ever having been tested in a demo or pilot project. Many of the problems stemming from the pre-RBRVS payment system could have been addressed with far less change to the overall payment than occurred with the RBRVS, and as a result, with far less difficulty—although few would want to return to the previous system at this point.

The Need for Measured Steps

As the experience with the RBRVS suggests, when introducing significant changes in payment, a good argument can be made for taking slow, measured steps. And indeed, the pace of change being proposed for pay for performance, particularly for physicians, is quite modest. In fact, along with those who have expressed concern about the uncertainty of pay for performance’s likely outcomes, there are some who have criticized the concept as being too insignificant or too narrow—even too modest.

I strongly agree that more needs to be done— particularly to encourage better care of chronic disease in the elderly population and to better align incentives between physicians and hospitals and potentially between physicians and other providers or payers. And there are other strategies that could be adopted in addition to pay for performance. Gainsharing is a promising strategy that is winning some prominence. With this approach, physicians share in the savings they produce through smarter prescribing and use of lab and imaging and through better treatment of serious illnesses or chronic disease. Because gainsharing is the subject of a Centers for Medicare and Medicaid Services demo scheduled to start in 2007, we can even expect to know something of the likely results of such activities.

But in the meantime, pay for performance is an idea that should not be too readily dismissed. A slow, gradual, voluntary (at least for physicians) introduction of pay for performance has enormous appeal for many—and if it brings in modestly more money, even potentially for physicians.


Gail R. Wilensky, PhD, is a senior fellow at Project HOPE and a commissioner on the World Health Organization’s Commission on the Social Determinants of Health. She was the HCFA administrator from 1990 to 1992 ([email protected]).

Gail Wilensky, Ph.D.

Categories