Uwe E. Reinhardt is an economics professor at Princeton. He has some financial interests in the health care field.
Medicare pays physicians for services rendered to Medicare beneficiaries according to a fee schedule that was enacted when President George H.W. Bush signed the Omnibus Budget Reconciliation Act of 1989. The schedule took effect on Jan. 1, 1992, and is in force today, after a series of modifications over time.
This fee schedule contains about 7,000 distinct nonsurgical and physician services, classified under a nomenclature based on the Current Procedural Terminology to which the American Medical Association holds jealously guarded intellectual property rights.
Readers may find it a bit odd, as do I, that the nomenclature of a government fee schedule for physicians is privately owned, which may restrict its use for research outside government and definitely does so for commercial purposes.
One would have thought that the nomenclature of the government’s fee schedule would be a pure public good, freely available to all potential users.
History of the Fee Schedule
The introduction of the Medicare fee schedule was a reaction to the horizontal inequities that
had resulted from Medicare’s previous reliance on the “usual, customary and reasonable” payment method then used by private insurers
for all patients and now mainly for services procured from physicians outside the insurer’s network.
Such a system naturally leads to different payments for the same service provided by different specialties and to physicians in the same specialty practicing in different geographic areas. These differences were purely haphazard and had no relation to different levels of quality of these services or their value to patients.
While such inequities have been routinely countenanced in the private insurance market, where price discrimination is the norm, they violate the constraint that government payments to providers must, above all, be seen as fair among providers and fully transparent to all.
To own up to these requirements, Medicare, under the Reagan administration, initiated in 1985 the development of a new, horizontally equitable and more easily defensible fully transparent fee schedule.
The first step was financing a large study, jointly conducted by researchers at Harvard University and at the American Medical Association, to estimate the relative amounts of “work” physicians contribute to the services they render. The definition of “physician’s work” took into account the physician’s time, mental effort, judgment, technical skill, physical effort and psychological stress.
The results of the Harvard-A.M.A. study, published in 1988, laid the groundwork for what is now known as the resource-based relative value scale, or, in common parlance, the R.B.R.V.S.
How Physician Fees Are Determined
Basically, the R.B.R.V.S. breaks down the total cost of providing a particular physician service into three components expressed in relative value units, commonly known as R.V.U.’s:
1. the physician’s own “work,” which accounts on average for 52 percent of the total relative costs for each service;
2. the physician’s outlays for any practice resources other than his or her own time (office space, supplies, clinical and administrative staff), which accounts for an average of 44 percent of the total relative cost of a service;
3. professional liability insurance, accounting on average for 4 percent of total relative practice costs.
For a particular market area, each of these three cost components is adjusted by a geographic practice cost index that accounts for variations across market areas in the cost of living, the prices of nonphysician practice resources and malpractice premiums.
The sum of these geographically adjusted R.V.U.’s for a particular service then constitutes the total R.V.U. of that service.
Finally, to convert this schedule into a fee schedule expressed in dollars, the total R.V.U. of a given service is multiplied by a “conversion factor” – a dollar amount per R.V.U. applied to all services in the relative value schedule.
The original conversion factor, in 1992, was $31. The current one for 2010 is $36.0791. In 2001 it was as high as $38.2581 (for details, see here).
The conversion factor is updated annually according to a formula that takes into account growth in the Medicare Economic Index (a price index for medical practice inputs), a projected productivity gain thought to be achievable by all physicians and growth in gross domestic product, thought to represent the nation’s ability to pay. The latter linkage is expressed by the “sustainable growth rate” formula, or S.G.R.
A more detailed description of this process – replete with a flow chart — is posted on the Web site of the Medicare Payment Advisory Commission.
Criticism of the Fee Schedule
Just like Medicare’s payment system for hospitals, described in last week’s post,
the Medicare physician fee schedule is frequently criticized for being based on relative costs, rather than on relative values to patients.
It is a fair criticism, but it also is facile.
It is facile, because typically that criticism is not accompanied by a practical – and I stress practical — alternative. After all, no other payer has yet based physician fees on the value of services to patients. If pricing physician services according to their value to patients were easy, why would not private insurers have used that approach long ago?
Indeed, the lack of a workable alternative can explain why, for the most part, private insurers actually use Medicare’s cost-based relative value scale as the basis for negotiating fees with physicians in their networks, negotiating a conversion factor relative to Medicare’s.
Suppose, however, one tried to base Medicare’s fee schedule on the value of services to patients. How would one determine the relative value of each of the 7,000 or so services in the schedule?
Economists might argue that, in principle, every service in a particular market area should be priced at a level that elicits from physicians collectively just the quantity of the service that is demanded collectively by users of that service in that market area. That’s how markets work for bananas, haircuts and even some highly elective services produced by physicians – e.g., purely cosmetic surgery or Botox injections. But with day-to-day health care that is problematical.
First, who actually “demands” most of the 7,000 items in the physician fee schedule: patients on their own, or physicians urging often anxious patients to “demand” the services in question – e.g., diagnostic tests, imaging or surgical interventions?
In economic theory, demand for a particular good or service depends on the potential buyers’ willingness to pay for it, given their budgets and ability to pay. But whose willingness to pay should we use for Medicare? Should we use an estimate of what Medicare beneficiaries would be willing to pay for physician services in the absence of Medicare coverage? Or should it be the taxpayers’ willingness to pay, as expressed by their political representatives? Should it perhaps emerge from a consensus of some panel of experts convened by Medicare to represent patients or taxpayers?
While no one likes the cost-based Medicare fee schedule, its critics should be challenged to suggest a workable alternative based on “value to patients.” For starters, the critics should explain concretely how they would define and measure value in this context, in monetary terms, keeping in mind the administrative cost of any such system.
In future posts I shall comment on the highly controversial role the American Medical Association and intra-physician politics play in the periodic recalibration of Medicare’s fee schedule – a process that appears to have contributed to the current shortage of primary care physicians.
I shall also describe how the “sustainable growth rate” formula for updating the conversion factor for Medicare’s physician fee schedule has created, since 2002, an annual Washington ritual that can best be described as theater — with much drama in between, but with a predictable outcome.