To support the counterintuitive notion that the United States has too many physicians, Shannon Brownlee focuses on a series of strongly contested conclusions (“Overdose,” December Atlantic). A large body of research has established the need to expand physician supply, and prominent national organizations, such as the Association of American Medical Colleges and the American Medical Association, are calling for more physicians to be trained.
Brownlee cites the conclusion of Dartmouth’s David Goodman that some academic hospitals, such as those at NYU and UCLA, employ two to three times as many doctors per patient as others. In reality, academic medical centers cluster into two groups. Those in cities like Madison, Wisconsin; Salt Lake City; and Hanover, New Hampshire, use relatively fewer resources, while those in large urban centers, like Newark, Detroit, and Los Angeles, where extreme poverty and affluence co-exist, use more resources, particularly for the poor. Reducing the number of physicians in the latter cities could hardly lead to better care. (Los Angeles, Newark, and similar cities also have more police officers than Madison and Hanover do, but cutting the police force is unlikely to reduce crime.)
Brownlee also cites Elliott Fisher’s work concerning the relationship between expenditures and quality. Like Goodman, Fisher compares the noncomparable. His “high spending–low quality” category consists mainly of these same urban centers—Chicago, Detroit, Philadelphia, Newark, etc. In contrast, his “low spending–high quality” category includes all of Alaska, Washington, Oregon, Wyoming, Montana, Idaho, Utah, Minnesota, Maine, New Hampshire, and Vermont, and parts of neighboring states. How does one even begin to compare health care, or anything else, in such dissimilar geographic aggregates?
But most troubling is Brownlee’s quote of the Baicker-Chandra paper, which she says reports that “measures of quality … tended to fall in direct proportion to a rising ratio of specialists.” The Baicker-Chandra model looked at a hypothetical, computer-generated construct that predicts what would happen if a specialist replaced a family physician. Such a scenario never happens in the real world. It’s a statistical game. What happens when the outcomes of real physicians are examined? When states with more specialists per capita are compared with states that have fewer, it turns out that the states with the most have the best quality of care. Nonetheless, Brownlee suggests that “more doctors lead to worse care, and fewer doctors to better care.” This statement is categorically wrong and utterly irresponsible.
Richard A. Cooper, M.D.
Professor of Medicine
University of Pennsylvania
Shannon Brownlee asserts that the number of doctors being trained in the United States is increasing. That is not the case. While the number of medical-school graduates in the U.S. is increasing, the number of physicians coming out of residency training programs is essentially capped by limits on federal funding for physician training at the nation’s teaching hospitals. The growing number of medical-school graduates in the U.S. will merely displace international medical graduates who now fill residency training slots not taken by U.S. graduates. Unless the number of training slots is significantly increased—and there are no plans for such an increase—the number of doctors being trained in the U.S. will remain fixed where it has been for more than two decades.
Vice President of Communications
Double Oak, Texas
The Association of American Medical Colleges’ recommendation for a 30 percent increase in medical-school enrollment, cited by Shannon Brownlee, was never meant to be a cure-all for what ails the nation’s health-care system. We agree with Brownlee that any plan to improve our current system should include better-coordinated care, enhanced incentives to draw doctors to underserved areas, and a more efficient use of resources. But high-quality health care requires, first and foremost, that physicians be there for the patient.
It has been clear for some time that the United States is not educating enough doctors. The number of medical-school graduates has remained flat since 1980, while the U.S. population has grown by 70 million. To meet the public’s need for accessible health care, our nation has become reliant on physicians who obtained their medical education in other countries, many of which are less- developed nations where health-care professionals are also in short supply. Last year, almost 7,000 foreign-educated physicians entered our health-care system.
An acute shortage of doctors in the United States would have a profound effect on access to health care, including longer waits for appointments and the need to travel farther to see a doctor. The elderly, the poor, and rural residents would face even greater challenges.
The need to build the capacity of our nation’s medical schools is real, and will become more urgent in the decades to come. Any proposal to reform our health-care system must address this need to ensure that all Americans have access to the high-quality health care they deserve.
Darrell G. Kirch, M.D.
President and CEO
Association of American Medical Colleges
Shannon Brownlee replies:
Richard Cooper incorrectly characterizes the research I cite as “strongly contested.” The groundbreaking work by the Dartmouth group is widely acknowledged and honored, and it offers a robust argument for rethinking plans to expand the physician workforce. During two telephone interviews, Cooper offered no data to suggest that reducing the per capita number of physicians would lead to harm in cities like Los Angeles. His conclusion is based on the assumption that the market has set the number of doctors practicing in such cities in response to levels of sickness. In fact, prevalence of severe chronic illness accounts for less than 5 percent of the variation in the amount of care delivered to Medicare recipients in different regions of the country, one of several lines of evidence suggesting that the physician labor market is not particularly sensitive to the population’s actual demand for care. And physicians, unlike cops, are able to generate demand for their services.
Finally, Cooper’s characterization of the work of Harvard economists Katherine Baicker and Amitabh Chandra as nothing more than a “statistical game” seems odd, since they employed an analytical technique that is a staple of economics research. Baicker and Chandra asked what would happen if you held the number of doctors steady but changed the ratio of specialists to generalists. Their finding—that more generalists lead to better-quality care—has been replicated in a variety of ways by other researchers.
Phillip Miller correctly points out that increasing the number of graduates without boosting residency slots won’t change the number of doctors. The Council on Graduate Medical Education, which advises Congress on physician-workforce policy, specifically recommends that the number of physicians entering residency programs go up by more than 10 percent in the next decade.
I agree with Darrell Kirch that reducing dependence on foreign doctors is a worthy goal, and that a shortage of physicians would not be good for the nation’s health. What we’re arguing about here is the definition of shortage. Projections by the AAMC, COGME, and others seem not to take into account the growing body of evidence that too many doctors may be as harmful under our current fee-for-service reimbursement system as too few. A more effective strategy for ensuring our future health might be to reform the payment system; to encourage young physicians to go into primary care and to settle in regions of the country where their services are most needed; and to teach them to practice collaborative medicine. Simply increasing the number of physicians will accomplish none of those goals.