This study predicts that the demand for physician services will significantly outpace the supply of physicians within the United States of America from 2017 to 2030, causing many states to face severe physician workforce shortages (Fig. 2). By 2030, 34 out of 50 states will have physician shortages with a grade of C− and below. Most of these states are located in the South and West regions, which is consistent with our prior publications regarding the shortage of nurses [14, 15].
Examination of the healthcare workforce is difficult due to the complexity of factors (e.g., population growth, age, economics, healthcare policy, healthcare practice, geography, models of care, new technologies, and innovations) that affect supply, demand, and balance between supply and demand in each state. By 2030, the following states will have the largest physician shortages (the number of physician jobs): California (32 669 jobs), Florida (21 978 jobs), and Texas (20 420 jobs). The key reasons for high shortages in these three states are attributed to high physician demand from the growth of the total population, aging state population, and aging physicians without an adequate commensurate increase in physician supplies. Taking California as an example, the state population and elderly population will grow by 112% and 148% between 2017 and 2030, respectively. Meanwhile, 33.4% of all active physicians in California are over 60 years old and within 5 years of retirement, although at the national level, only 30.3% of all active physicians are 60 years old and above [20]. From 2006 to 2016, the number of medical students at Doctor of Medicine (MD) and Doctor of Osteopathic Medicine (DO) schools increased by 20.7% (7387 students), and the number of residents and fellows increased by 12.8% (10 429 students) [20]. On the basis of current trends, the number of new licensees per year is not large enough to replace these physicians as they retire, and physician demand will outpace the physician supply within the state of California [21].
Florida and Texas, the two other states with the largest physician shortage, are experiencing similar situations as California. These two states are on the top of the list of the fastest-growing states in the United States of America. In the next 10 years, Florida’s population will increase by 30% and Texas will increase by 22%. Furthermore, by 2030, the aging population will increase 73% and 55% in Florida and Texas, respectively. Physicians are aging along with the general population in these two states, which will have a profound impact on the physician workforce. Approximately, 34.3% of all active physicians in Florida [22] and 27.2% of all active physicians in Texas [23] are over 60 years old. It should be noted that Florida has the seventh oldest physician population in the United States of America [22] with a total of 16.6% of its physicians planning to retire in the next 5 years [24]. In the past 10 years, these two states rapidly expanded their number of medical students, residents, and fellows. From 2006 to 2016 in Florida, there was a 70.9% increase in the number of medical students and a 50.5% increase in the number of residents and fellows. The percentage changes in Texas are 31.6% and 18.6%, respectively [20]. It must be taken into account that the growth of medical students outpaces the growth of residents and fellows. Thus, continual efforts to increase the pipeline of physicians in these states are paramount.
As for physician shortage ratios (physician jobs per 100 000 people), by 2030, the three states with the most severe physician shortage ratios will be Mississippi (120), New Mexico (101), and Louisiana (100). The only state expected to have an F grade is Mississippi. Mississippi has a low physician availability of 118 active physicians per 100 000 people, 42% below the national mean of 203. Consequently, Mississippi will require an additional 3709 physicians by 2030, a 51% increase of the state’s current 3528 practicing physicians (as of 2017) to meet the national benchmark. Since 2006, the MD and DO schools in Mississippi have increased their enrollment by more than 130%, which ranked first in percentage change nationally. By 2022, Mississippi medical school graduates will likely increase to more than 265 students annually, which is more than 2.5 times the Mississippi medical student graduation rate prior to 2005 [25, 26]. In addition, Mississippi increased its graduate medical education (GME) first-year training positions at community-based GME sites from 14 to 56 and its residency slots from 42 to 108 [25]. These actions can help Mississippi to train and retain more medical professionals within the state to mitigate its physician shortages.
New Mexico has the second-worst projected shortage ratio and is also the state with the largest change in shortage ratio, rising 57.1 physician jobs per 100 000 people by 2030 (Table 2). Among most states, aging of the physician workforce is a significant factor contributing to future state shortages. New Mexico has the oldest physician workforce in the nation, with 37% of physicians over 60 years old and facing retirement in the next 10 years [27]. To maintain the status quo, New Mexico will require an additional 2118 physicians by 2030, a 40.4% increase of the state’s current 3128 physicians (as of 2017). The aging physician population reflects the difficulty New Mexico has in attracting and retaining young physicians.
Massachusetts and Vermont are the two states with the highest physician surplus. Both of these two states are located in the Northeast region, where GME training programs have historically been located. The distribution of residents is particularly important, given evidence that physicians tend to practice in geographic areas similar to those where they complete their GME training. Every year, federal GME spending of over $15 billion trains residents across the country [28]. The Northeast region received $5.47 billion (38%) of total federal spending, which is almost three times what the West received ($1.83 billion, 13%). Consequently, 31% of GME residents were located in the Northeast [28]. It should be noted that the states with a physician surplus (graded A or B) will face increasing pressures from states with physician shortages (graded C− and below) to attract and retain physicians from their state [29].
There are a few related studies in the literature which examine the status of the physician workforce. HRSA produced a model of patient demand for primary care services that also incorporated the sizable challenges of an aging and growing population [30]. This study projected a 23 640 primary care physician shortage in 2025. With delivery system changes and full utilization of mid-level healthcare providers, including nursing practitioner (NP) and physician assistant (PA) services, the projected shortage of 23 640 primary care physicians can be effectively mitigated. More recently, AAMC published a report on the supply and demand of physicians projected through 2030 [8]. This study supports the notion that an increasingly older population will result in an increase in the demand of physicians and cause a greater shortage of physicians. Using multiple different scenarios to reflect different assumptions, such as the use of NPs and PAs to assist staffing problems, AAMC predicted that the physician shortage will be between 42 600 and 121 300 in 2030. Based on our models, we predict that there will be a shortage of 72 472 physicians in 2025 and 139 160 physicians by 2030. These findings are consistent with the aforementioned reports under the scenario without mid-level healthcare services. In many states, mid-level healthcare providers are present to mitigate the shortage of physicians. With the assistance of these services, the healthcare industry is able to utilize a new patient care model that is more reliant on a team-based care delivery in order to care for an increasing number of insured and elderly patients. This team-based care model helps healthcare organizations work together more effectively and efficiently [31]. The utilization of mid-level healthcare providers may help to mitigate the physician shortage, it cannot completely replace physicians [32].
Recently, the New York University School of Medicine announced that it would eliminate tuition to encourage people to pursue medical careers [33]. Without the prospect of overwhelming financial debt, more people will pursue medical careers. People view this decision as a positive step forward, but they also caution that it might not be a silver bullet for America’s worsening physician shortage. Presently, the United States of America is not facing a medical student shortage, but rather, a residency shortage. Because of the 1997 cap on Medicare to support GME, the necessary commensurate increases in residency training have been stymied [34], creating a bottleneck for the physician supply. In 2018, there was a record-breaking 37 103 US and international medical school students and graduates competing for only 33 167 positions, a shortfall of about 4000 residents-to-be [35]. To address this issue, the Resident Physician Shortage Reduction Act of 2017 (H. R. 2267) was introduced in Congress to increase by 3000 the annual number of residency slots from 2019 to 2023 [36]. States such as Arkansas, Kansas, Missouri, and Utah have also passed legislation to provide provisional licenses to some medical school graduates who have not been able to find residency spots [37]. With these provisional licenses, they can practice primary care under the medical license of another physician, but only in medically underserved areas. Some institutions have also created their own medical-school-to-residency pathways. For example, Kaiser Permanente currently trains 600 residents annually and provides continuing medical education (CME) to another 22 000 medical professionals [38].
With the rapidly increasing demand for physicians, many US healthcare institutions turn to foreign-trained doctors to supplement their physician workforce [39]. Foreign-trained doctors have long been an integral part of the US healthcare system, contributing substantially to primary care disciplines and providing care in underserved populations. According to the Organization for Economic Cooperation and Development [40], in 2016, there were more than 215 630 foreign-trained doctors practicing in the United States of America, a number far surpassing any other country. It should be noted that the global physician supply is finite and the competition for these providers exists across national boundaries [41].
As technology evolves, its impact on the medical field increases. For example, artificial intelligence (AI) algorithms will shift the roles of physicians from a knowledge-based role into more of a skills-based role [42]. All of that information will be available in an AI-driven database that can not only bring up the information at a moment’s notice, but also help to diagnoses. Technology enables physicians to spend less time testing samples and recording data and spend more time providing quality care to their patients. Technology also helps medical institutions operate more effectively and efficiently, which may alleviate some of the burdens of the physician shortage.
There were several limitations to our study involving the construction of demand and supply models. We made a crucial assumption in the demand model by using the current (2017) national ratio of physicians to the overall population (203 physicians per 100 000 people) as a baseline. We assumed that no shortage exists in this ratio, since a relative zero value was needed to calculate future shortages. This study does not comment on whether the country currently lacks physicians, but asserts that shortages are imminent, based on current trends in supply and demand. Therefore, if our assumption that the 2017 baseline without shortage is incorrect, our projections may underestimate the true nature of the shortage. In addition, the national slope was used in converting change in PHE to physician jobs to avoid state variations in physician workforce responsiveness to health expenditures. Further analysis indicated that change in PHE translates to a larger change in physician demand in some states than represented by the national slope. The responsiveness of physician jobs to PHE also may vary depending on the work setting for physicians.
The primary assumption in the supply model is that the average likelihood or propensity of an individual to choose to be a physician at a certain age is the same across every state and will be the same in the coming years. This does not address individual states’ differences in their ability to recruit young people into the medical profession and the capacity of their medical schools, including the number of faculty present. Moreover, we used an average propensity value over the past 10 years, which does not account for the increasing enrollment rates in medical schools. If enrollment continues to increase as it has for the past decade, the physician propensity value for those between ages 21 and 34 will be an underestimation. Another limitation that may underestimate the physician supply is the exclusion of future arrivals of foreign-born physicians. According to a study in 2015, almost a quarter of residents across all fields, and more than a third of residents in sub-specialist programs were foreign medical graduates [39]. This means that we are reliant on physicians trained outside the country to fill the gap.