How Many Physicians? How Much Does It Matter?

The analysis of current estimates of the number of physicians practicing in the United States by Staiger and colleagues¹ in this issue of JAMA touches on a topic that is both arcane and central to policy debates. The arcane part is how physicians are counted; the policy part is how their contributions to the economy and to social needs are estimated.

The study by Staiger et al¹ examines a central source of data on the supply of physicians, the American Medical Association (AMA) Physician Masterfile. This is a longstanding inventory of the supply of all physicians, alive and dead, practicing medicine or not. Its origins trace back to 1906 when the AMA attempted to identify all practicing physicians in the United States, members and nonmembers.² The data file became widely accepted as the authoritative source of information about the number and location of physicians to the point that in 1966, the National Committee on Vital and Health Statistics, in a report entitled “United States Statistics on Medical Economics,” took note of the value of the data file as a reliable source for information on the “geographic location, age, sex and specialty” of physicians, but also noted that the data could be improved to better reflect the volume of patient care the physician provided.³ In 1968, the AMA Center for Health Services Research and Development revised its data collection processes and its method of classifying physicians. The file has since undergone various efforts to improve its quality over the years because it also is a source of revenue for the AMA,⁴ as well as reference data for public policy deliberations.

It is not surprising that in the study by Staiger et al,¹ 2 separate ways of estimating the number of physicians in the United States should result in different numbers. However, the importance of the study is not about the method but whether the actual number is too many or too few.

The current controversy over the accuracy of the AMA Physician Masterfile data is a continuation of a longstanding debate over the supply of physicians and whether that number meets the US health care needs and promotes economic efficiency.⁵ The supply of any labor resource, especially in the professions, is a critical component of market dynamics, as well as a mechanism to achieve social goals, including promoting a healthy population. The central policy question is how much public resources should be applied to regulating physician supply. The connection between physician supply and government policy has been tracked by economists⁶ and physicians.⁷ For example, physician supply is seen as an important determinant of costs in the Medicare program as well as a direct expense in the form of subsidy for graduate medical education.⁸ Elements of health reform bills being discussed in the current session of Congress include adjusting the physician supply and its distribution with incentives targeted to expand the number of primary care practitioners or to attract physicians to rural and underserved places. That policy attention to supply has expanded to specialists, including psychiatrists and general surgeons.

The debate over health reform under way in and outside Congress highlights how physician supply is connected to the more macro policy objectives of universal access to care and cost control. If access is expanded through subsidy to health insurance, the demand for physician services will increase potentially beyond what is available. If physician supply is in excess of what the market will bear, the prices of those services may increase to unsustainable levels. The first situation is seen as a cause for alarm and further subsidy of medical training by the government; the second, as a further threat to the stability of the health care delivery system as costs increase very rapidly, displacing other goods and services.

Establishing the right number of physicians is difficult if they cannot be accurately counted. It has been argued that knowing what is the “right” number is a futile quest,⁹ but enough is known about the distorting effects of an oversupply as well as an undersupply of physicians to provoke a search for accuracy. The margin of error of 10% attributable to sources as implied in the study by Staiger et al¹ may alternatively mean that such a difference does not affect outcomes as much as expected, because it has effectively gone unnoticed, or that the analyses must be reworked substantially and possibly uncover more potentially negative effects of undersupply and oversupply. That dilemma cannot be solved unless the supply is known—further argument for more accurate measure of physician supply.

The study by Staiger et al¹ provides one of several options for testing and comparing physician supply; in this case, US Census statistics and a cohort aging model. The model assumes consistent cohort effects over time when there may in fact be substantial changes in participation in clinical practice depending on the birth cohort—that is, physicians in the age range of 45 to 54 years may practice substantially more hours today than a similarly aged group in 2020. The lag in retirement reporting may be addressed with more timely reporting of activity from licensing agencies or third-party payers. For instance, development of better data collection and reporting mechanisms at the state level from licensing boards could be used to regularly update a central registry overseen by the Federation of State Medical Boards. What is surprising is that all of these are not used in a robust combination to provide a more consensus estimate of current and future effective supply of physicians. The federal government does, from time to time, issue a projection of physician supply,¹⁰ but it usually provokes more contention than agreement. The effects of this federal report on policy debates are often attenuated because the projection is often too late to really provide much guidance.

The physician workforce is one of the most critical factors that must be considered in current health care reform efforts and discussions. Having accurate estimates for determining not only the number of physicians, but also current and future physician workforce requirements and capabilities for delivering primary and specialty care, will be essential for achieving and sustaining effective health care reform.