Is There a Better Way to Produce a Health Workforce for the Population?

The article by Frederick Chen, MD, MPH, and colleagues in this issue of Family Medicine reminds us of the importance of residency training in family medicine. ¹ Their study focuses on the use of the Supplemental Offer and Acceptance Program, formerly known as the Scramble, to add residents who didn’t match in the formal matching program into open residency slots. There is a declining interest in primary care by US medical school graduates. According to the Association of American Medical Colleges, only 6% of US or Canadian medical school graduates are currently in family medicine residencies out of all of the residency slots for Accreditation Council for Graduate Medical Education approved specialties.² The proportion of US graduates choosing family medicine has been low for many years, with family medicine residencies relying on international medical graduates to fill slots.² In 2020-2021, among active residents, 26% of family medicine residency slots were filled by international medical graduates.

The policy position of creating more family physicians regardless of how they end up in residency slots has a relatively noble, population-based rationale.³ Based on the recommendations of the Council on Graduate Medical Education for the past 30 years, we need more family physicians.⁴, ⁵ Having a robust primary care system in the United States is a good thing because it is associated with better health outcomes and lower costs.⁶, ⁷ We need to keep the workforce pipeline filled with trainees and graduates to maintain a sufficient primary care workforce for the population of the United States. The unstated theme in the Chen et al study is that it is important to fill unfilled slots.¹ It is hard to argue with a position that will provide good health for the population. Every filled slot, regardless of how it gets filled, helps with our health workforce goals, yet, it is clear that this strategy has not achieved the proportion of primary care physicians that we say we need. Is there a better way to design a strategy to meet the health workforce goals?

Unfortunately, these residency slots and the goals of filling them are not designed in a truly structured way to achieve these health policy goals. In fact, residency slots are market commodities and tools for financial gain within hospitals and health systems. The reimbursement from the Centers for Medicare and Medicaid Services (CMS) for a residency slot is between $100,000 and $120,000 while the salary for a family medicine resident is approximately $65,000. That differential between revenue and costs makes residency slots financially attractive. However, that money reaches the hospital only if the residency slot is filled. An unfilled slot does not get the direct or indirect funds.

A different illustration of residency slots as financial commodities is exemplified by the bankruptcy of the Hahnemann University Hospital. In 2019, Hahnemann University Hospital went bankrupt and in the process concluded that their residency slots were assets to be sold.⁸, ⁹ Hahnemann sold their residency slots at auction for $55 million. CMS considered the strategy of selling medical residencies for profit to be illegal. A federal judge blocked the sale but the general idea that residency slots have monetary value was clear. The situation at Hahnemann and the desire to sell these valuable commodities will likely happen again to financially strapped health systems.

The present system for filling residency slots is complicated even further and strays from a rational strategy for workforce goals because the system is based on making the specialty or location appealing to graduating medical students. There have been many studies of specialty choice by medical students and strategies that would hopefully make them more likely to choose family medicine.¹⁰ As previously noted, with only 6% of US medical graduates choosing family medicine, it would appear that strategies based on trying to appeal to medical students has not achieved the workforce goals. The current system is based not on what society needs but rather on what medical students are attracted to. Keeping society healthy should be the paramount underpinning of the health workforce development strategy.

The question to consider is whether there is a more rational way to create a robust primary care workforce than competing for specialty choice and filling unfilled slots outside of the formal system. Perhaps a more proactive stance on workforce development needs to be taken. It may be that the best way to allocate residency slots is to link the number of residency slots and the specialty of those residency slots to the needs of the population. If, for example, we want 50% or even 70% of the physician workforce to be primary care, or more specifically family medicine because of the declining interest in ambulatory general internal medicine, then it might make sense to designate 50%-70% of the available residency slots as family medicine slots. At the same time make the number of residency slots nearer to the number of medical school graduates to eliminate surplus unfilled slots. Implementing such a plan would be a disruption to the current status quo and would be a departure from a competition between specialties to attract students to a system designed to meet workforce needs for the population.

It is important to keep in mind that the vast majority of residency slots are funded by CMS. It is their money and if the US government wants a larger proportion of their physician workforce to be primary care this would be a way for them to allocate their money in a way that achieves their stated goals. This new strategy would require a rebalance in the slots allocated to different specialties to increase the number of primary care positions. There would be winning specialties and losing specialties in this rebalancing and it must be noted that residency slots are a valuable market commodity. But if the overall goal is to create a workforce that is the best one for the community then this is a rational and strategic way to achieve it.

The advantages to such a rebalance would be clear. First, the needs of the population would be driving the workforce planning decisions. If we want 50% of the workforce to be family physicians we need to design a system to achieve that. This would move us away from the financial incentives surrounding filling residency slots. Second, the workforce would not be dependent upon an open market of competition for specialty choice. Third, it will be clear to everyone that graduate medical education is not a commodity or a health policy afterthought but rather a means to create a healthier population.