Overview of a pathway to practice for foreign-trained physicians
Thank you to Joe Lonsdale, Judge Glock, Sebastian Caliri, and Clay Spence for their thoughtful feedback on this policy proposal. Their contributions were invaluable and strengthened the proposal. All errors are my own.
The United States faces a provider shortfall that is expected to grow worse in the years ahead. To address part of this shortfall, and alleviate the disproportionate impact it has in rural America, states should open up a pathway to practice for foreign-trained physicians willing to commit to practice for five years in rural areas.
Foreign-trained physicians earn significantly less than their American peers and would likely be tempted to practice in the United States if they were not required to repeat residency to do so. By allowing providers who completed equivalent residencies abroad to practice in the United States, states can improve access to health services, address health disparities, and ensure continued adherence to high quality standards.
States should ensure that graduates of foreign residency programs that meet the criteria used by the Accreditation Council for Graduate Medical Education to evaluate American residency programs have a path to practice. Reviews should be done on a residency program basis to ensure the particular program meets the required criteria. States can finance their reviews of foreign residency programs by charging fees for their reviews. States should accept programs accepted by other states rather than expend resources reviewing programs on their own. Total annual cost of reviewing foreign residency programs will likely total approximately $600,000 and return multiples of that figure in reduced health expenditures.
Policy proposal: Pathway to practice for foreign-trained physicians
The physician shortage has emerged as an intractable problem facing the U.S. While a variety of solutions should be considered, confronting this situation requires addressing supply-side restrictions currently in place throughout the U.S. health system.¹ As part of this effort, states should consider removing barriers preventing trained physicians who have completed residencies abroad from practicing medicine in the U.S. and addressing part of the projected physician shortfall. States should also liberalize their approach to providers licensed in other states and remove unnecessary scope of practice restrictions and supervisory requirements that apply to Physician Assistants (PAs) and Nurse Practitioners (NPs).
I. Literature review: Foreign-trained physician entry into the United States (U.S.)
A. Why would foreign-trained physicians seek to practice in the U.S.?
Foreign-trained physicians seek to practice in the U.S. for a variety of reasons, including: an income differential between the original country in which they practiced and the U.S.; the high-quality of life in the U.S.; work opportunities their partner has in the U.S.; vicissitudes of life introducing unexpected changes.
Physicians in the U.S. earn more than their foreign counterparts. Medscape’s 2019 international physician compensation report shows an average salary of $313,000 for U.S. physicians compared with $163,000 for German physicians, $138,000 for British physicians, and $108,000 for French physicians. This differential holds true even when comparing U.S. primary care physicians to their foreign counterparts as their average income is above the overall average physician income in other countries. While the differential with some countries is partially explained by cost-of-living differences, the difference also reflects different payment systems and supply restrictions in the U.S. health system (e.g., scope of practice limitations on PAs and NPs). Other contributing factors, such as differences in the cost of medical school, would not be sustainable barriers in the absence of supply restrictions barring foreign medical school graduates from practicing in the U.S.
Even those who come for non-economic reasons face barriers when they seek to practice medicine. Qualified physicians who are forced by unexpected circumstances to move have had to stop practicing medicine. This includes refugee physicians fleeing war, who are forced to go through the regular licensing process when they come to the U.S., including repeating residency programs.
B. How would the presence of additional foreign-trained physicians benefit Americans?
The American public would benefit from an influx of foreign-trained physicians as America has a provider shortage that will only grow more acute over time. The provider shortage has been recognized as a problem by the Health Resources and Services Administration (HRSA), which has dedicated resources to tracking shortage areas. The Association of American Medical Colleges (AAMC) projects that the shortage will become worse by 2030, with the shortfall ranging between 40,800 and 104,900 physicians.
This shortage poses a particular challenge for those who do not reside in large metropolitan areas, where primary care physicians are more concentrated. The shortage is even more acute when it comes to specialists, with 263 specialists per 100,000 people in urban areas and 30 specialists per 100,000 people in rural areas.
This physician shortage contributes to disparities in health outcomes and spending between urban and rural areas. Medicare beneficiaries in rural areas are 18% less likely to receive appropriate medication after hospitalization, 15% less likely to receive an indicated colorectal cancer screen, and 14% less likely to receive an annual diabetic eye exam. Rural Medicare performs significantly worse on 22 of 44 HEDIS measures, with severe consequences in disease burden and mortality.
Physician shortages also make it difficult for insurers to form adequate networks in rural areas, as recognized earlier this year by the Centers for Medicare & Medicaid Services (CMS) when it proposed and finalized relaxed network adequacy requirements for Medical Advantage (MA) plans in rural areas.² In endorsing a relaxed standard, CMS implicitly recognized that the physician shortage contributed to not only a lack of provider choice, but also a lack of insurer choice in rural communities and was a contributing factor to the low penetration rate MA plans have achieved in rural America.
The federal government has recognized the harm done by restricting foreign-trained physicians from practicing in the U.S. In a 2018 report from the Departments of Health and Human Services, Labor and Treasury, the departments wrote that:
[C]ertain policies relating to graduate medical education (GME), as well as significant restrictions on the ability of foreign-trained doctors to practice in the United States may … unnecessarily limit the supply of physicians available to provide care to Americans. Reduced competition among qualified physicians inevitably leads to higher prices for physician services and generally reduces the quality of care.
C. What are the barriers to foreign-trained physicians seeking to practice in the U.S.?
As is often the case with immigration, federal imigration limits play a role in preventing foreign-trained physicians from entering the U.S. However, the federal government clears an immigration pathway for foreign-trained physicians working in underserved areas under the Conrad 30 Waiver Program. Under this program, each state is allowed to bring in up to 30 foreign-trained physicians a year, for a total of 1,500 new physicians each year across the country. Significantly, most states do not use all 30 spots each year, with the average number of spots claimed ranging between 52% and 64% of spots depending on the year. There are thoughtful reform proposals to improve the program that should be considered at the federal level, including allowing states to claim unused waiver spots from other states and removing the requirement that waiver recipients show non-immigrant intent. However, even if federal reforms were pursued, the key barrier facing foreign-trained physicians would remain as state licensure regimes would still require foreign-trained physicians to complete residencies.
In order to practice medicine in the U.S., physicians must be licensed by the state in which they wish to practice. Licensure typically requires graduating medical school, passing the U.S. Medical Licensing Exam (USMLE), and completing a residency in the U.S. or Canada. These requirements also apply to international medical graduates.
Residency is a significant barrier to entry as it increases the effective cost of a medical education by creating a prolonged period with limited earnings, relative to physician’s level of education and the amount earned post-residency. A residency can lower the return on investment in a medical education below the return experienced by investment bankers or software engineers.
For foreign-trained physicians, the requirement to repeat residency again in order to practice in the U.S. effectively doubles the residency barrier and allows the income disparity between the U.S physicians and their foreign counterparts to continue. The physicians who currently come to the U.S. through the Conrad 30 Waiver Program come to complete residency training, as required under state law. This limits the appeal of the Conrad 30 Waiver Program, in effect subsidizing the lower salaries paid by other developed countries to their physicians.³ Of course, given the large payment differential between U.S. and foreign physicians, residency may not be a significant barrier in purely economic terms. After all, before a certain age it would be logical for a physician to emigrate and amortize the reduced earnings during their second residency over the course of their years practicing post residency. If states only wished to target younger physicians, residency would not be a substantial economic batter. However, accounting for human psychology, many physicians are unlikely to weigh future earnings the same way they count reduced earnings during their second residency in the prime of their lives. If a physician graduated undergrad at 21, medical school at 25, completed residency at 30, and then practiced for 2 years before considering relocating, accepting reduced income from 32-37 may seem like a significant sacrifice in the prime of their life. Additionally, per Vox:
US residency positions are also highly competitive and limited in number. In 2016, there were 35,000 applications for 27,000 positions, and those positions often favor graduates just out of US medical school.
II. Examining a reform effort: Assistant Physicians
A 2014 Missouri law enables medical school graduates to bypass residency and work as an Assistant Physician (AP). The law was prompted by a perception that qualified physicians were not getting residency spots and tailored toward medical school graduates without residency placements. While the law was not targeted toward foreign-trained physicians seeking to side-step residency, it may represent a model for future reform efforts as it creates an additional path to practice and is focused on care in rural areas.
Under Missouri’s approach, medical school graduates may only qualify as APs within 3 years of either passing the USMLE or graduating medical school (§ 334.036 R.S.Mo.). APs cannot have completed a qualifying residency and must be proficient in the English language (§ 334.036 R.S.Mo.).
APs are required to enter into collaborative practice arrangements with licensed physicians (§ 334.037 R.S.Mo.) in order to practice, and are limited to providing primary care services in medically underserved areas (§ 334.036 R.S.Mo.). Collaborative practice arrangements can delegate authority to administer or dispense drugs and provide treatment as long as the AP has the required skill set (§ 334.037 R.S.Mo.). Missouri deems APs the equivalent of PAs for CMS billing purposes and imposes no state supervision requirements on them (§ 334.036 R.S.Mo.). The state requires APs to meet the same continuing medical education requirements as practicing physicians.
The law as currently structured would seemingly only have limited appeal to foreign-trained physicians as APs have lower salaries than physicians and lack a pathway to practice outside arrangements with licensed physicians. A proposed amendment that failed to pass in Missouri in 2019 would have established a process for an AP to become a fully licensed physician after five years of practice and the completion of other requirements. As five years is a significant amount of time to practice under supervision for an already-licensed foreign-trained physician, a shorter pathway to licensure may be appropriate if the state legislature revisits the proposed amendment in the context of foreign-trained physicians.
There has been some criticism of Missouri’s efforts. Critics have examined Missouri’s effort from philosophical, quality of care, and implementation perspectives. Some have argued philosophically that “[p]romoting care to underserved by APs as a ‘fallback option’ devalues primary care and those patients.” Others have highlighted that APs may not be able to deliver the same quality of care as they “had significantly lower USMLE pass rates on all 4 Step examinations compared with the matching cohort of US medical graduates and for 3 Step examinations (except for Step 1) compared with international medical graduates.” Critics have also observed that in Missouri’s case, “[o]nly 25% of the licensees had secured collaborative agreements during the first year and thus were the only ones able to practice.”
Despite these limitations, similar legislation has been enacted in Arkansas, Kansas and Utah.
While an expanded version of the AP approach with a pathway to independent practice could be a viable pathway for some foreign-trained physicians, the salary differential in the interim years would likely deter many from pursuing this option and fail to alleviate the provider shortage the U.S. faces to the full extent the provider compensation gap between the U.S. and other developed countries suggests is possible.
III. Recommended reform: Foreign residency recognition approach
Based on a reform recommended by the federal government, this approach involves state legislatures authorizing their departments of health to deem residencies beyond those completed in the U.S. and Canada acceptable under the states’ licensure residency requirement. Under this approach, the requirements to graduate medical school and pass the USMLE would remain unchanged.
State’s could take a number of approaches to recognize acceptable foreign residencies. Part A of the section gives an overview of our recommended approach: Parts B through I go over different decisions that can be made in structuring this law on which states can choose differently.
A. Recommended approach
In order to attract foreign-trained physicians while ensuring the high level of safety and comfort provided by U.S. trained physicians will be maintained, states should set their initial decisions in statute to ensure the correct balance is struck. States should match the criteria the Accreditation Council for Graduate Medical Education (ACGME) applies to U.S. residency programs when evaluating foreign residencies. States should evaluate residencies on a program by program basis, rather than make decisions on a hospital or country-side basis. States should automatically accept residency programs accepted in other states.
Reviews should be able to be initiated by foreign residency programs or by individual providers seeking to practice in the state. Priority should be given to reviews initiated by foreign residency programs, with the state committing to providing an answer within 180 days of receiving a fully completed application. States should not set any timeline obligation by which they commit to responding to applications from individual providers.
States should hire four FTEs to manage their review process, with annual expenditures of approximately $600,000. States should charge residency program applicants a fee of $25,000 per application and individual provider applicants a fee of $3,000 per application. The individual provider application fee should be waived if an application is turned down. Residencies should be required to undergo further review every five years for the purposes of more recent graduates seeking to practice in the U.S.
States should require providers admitted under this program to practice for five years in rural areas as defined by HRSA before the providers are allowed to practice anywhere in the state without restriction.
B. Deciding who decides
The determination of which approach is used could be made by either the legislature via statute or the department of health via delegated authority depending on political factors particular to the state. What is gained in flexibility to adapt to changes over time when a delegation is made to the department of health, may be lost in increased legal vulnerability and a lack of follow-through if the state executive has anti-immigrant tendencies.
C. Criteria for evaluating residency programs
States could choose to base their criteria on those used by the ACGME, the non-profit entity responsible for accrediting U.S. residency programs.
D. On what basis to recognize residency programs
States must decide whether to recognize residency programs on a program, hospital or country basis. As there is often variation in the quality of different residency programs within a country or hospital, a state may wish to accept graduates of one residency program without accepting graduates of a separate program in the same country. What this approach gains in flexibility, it loses in administrability.
A related approach with slightly less flexibility would involve accepting all residencies completed at a particular high quality hospital (e.g., Singapore General Hospital), rather than going by specific residency programs. This approach is broadly similar to the program by program approach but would involve less of a research burden and be correspondingly easier to implement. States may find this approach strikes a safe balance by eliminating the most concerning examples of variation in quality but not necessitating reviewing specific hospital programs.
Alternatively, a state may be willing to accept all foreign-trained physicians who completed their residencies within a particular country (e.g., England) as they recognize that a particular country has a consistently high standard across all its accredited programs or uses a sufficiently centralized system to prepare residents where it would not make sense to distinguish between programs. This approach may be more suited for some countries than others and a state may wish to use it in combination with a hospital or program based approach that applies to foreign-trained physicians from countries not deemed uniformly acceptable.
E. Accepting programs accepted in other states
States must decide if they will automatically recognize programs recognized by all or some other states. States may wish to leverage the work done by other states by automatically recognizing programs other states have recognized. This can be done on a reciprocal basis, perhaps even by pooling resources into a joint research team, or by a state acting unilaterally. As the benefit flows to the state with the broadest aperture, there is a strong incentive to act unilaterally.
This effort echoes the decisions made in the creation of the Interstate Medical Licensure Compact and a state’s decision to participate or not. Participating states may be likely reform candidates and may be comfortable accepting residency programs accepted by other states.
F. Initiating review of a residency program
States could opt to allow providers applying for licensure to initiate review of the residency program they completed and commit to providing an answer within 180 days, with an explanation of why their residency was not accepted if the state concludes it must reject the residency. The legislature could require states to have data to back up a rejection and default to a non-precedent setting approval of at least the applying provider in the event the state cannot produce a data-backed reason to reject a residency program within 180 days. In this situation, if the state cannot produce a data-backed reason to reject a residency program within two years of initial application, the residency program could be automatically added to the list of accepted residency programs.
As an alternative, the legislature could require residency programs to initiate an application for program approval allowing physicians to do so. This permutation would likely see lower uptake as foreign residency programs may lack a clear incentive to enable their residents to emigrate, though it could be used as a differentiator.
States could also not have an option for physicians or residency programs to initiate the approval process, and instead make approvals a top-down approach initiated by the state department of health. While this centralized approach may appeal to some regulators, it would be less flexible in practice and potentially fail to take advantage of untapped provider supply. This top-down approach may also codify implicit biases toward programs in particular countries. This approach may be most appropriate to start a recognition program up, with a plan to accept review applications after the review program is established.
G. Cost and funding for reviewing residency programs
Incremental costs involved in reviewing residency programs will vary, depending on if a state hires new FTEs and / or organizes a new division to take on recognition responsibility or tasks existing personnel with the responsibility. If the volume of residency programs that need review is large, as will ideally be the case, new personnel will likely be required. States may opt to reduce these costs by pooling resources with other states that have adopted similar policies or by charging applicants a review fee. If an applicant is a residency program itself this fee may fully offset the cost of review and become a profitable fee for the state. It will likely be worthwhile for residency programs to pay high fees to increase their desirability and attract top-tier residents. Willingness to pay may range significantly, but states can consider placing initial prices in the low five-figure range. If an applicant is a specific provider, fees would likely not be able to fully offset costs as too high a fee would deter providers from seeking to practice in the state. Willingness to pay may be closer to the low four-figure range.
Once the state determines that what information to request in an application, presumably including home country assessment results of the residency program in its accreditation process, state FTEs will need to verify the information provided is accurate with foreign counterparts and assess how the residency program compares with those accredited by ACGME. If a residency has submitted accurate information that shows it consistently produces residents that are equivalent to those produced by ACGME accredited residency programs it should be accepted.
As doing this will require FTEs able to work with foreign counterparts, and respected by personnel at either ACGME and / or academic medical centers, there will likely need to be clinicians on this team tasked with this responsibility. The team will likely need to be led by an experienced clinician and staffed with an additional two clinicians and one non-clinically trained support staff. Additional staff may be required if the program is successful and demand surges.
Based on the Physician Series within the federal General Schedule (GS), a physician with 5 years of graduate training qualifies as GS-15 while a less experienced physician qualifies as GS-12. Assuming the team lead is a GS-15, the other two clinicians are GS-12, and the non-clinically trained support staff in GS-9 annually salary costs total to $373,538. Assuming wages only account for 62.2% of total employer costs, with benefits accounting for the remaining 37.8% based on Bureau of Labor Statistics averages, the total annual employment costs is $600,543. This is a highly simplified assessment as: state pay scales vary from the federal GS; the federal GS adjusts for locality cost of living differentials; the analysis places all personnel at the highest step for their pay grade; the analysis does not account for facility costs.
H. How often to re-review residency programs
States must decide when to re-review residency programs deemed acceptable in the past. As a starting point, to balance administrative burden with the need to ensure continued high quality, states could initiate reviews if either: i. No provider has applied from a residency program over the previous five-year period, or ii. The state obtains data indicating there has been a statistically significant change in median licensing exams test scores for graduates of the program. Alternatively, to ensure safety standards are upheld, states could automatically rereview residency programs every five years.
I. Geographic conditions on foreign-trained physicians
While some states may require physicians licensed through this pathway to practice in rural areas, others will allow them to practice in any medically underserved area, and others will not put any geographic practice restrictions on these physicians. This decision will likely be driven by how severe of a provider shortage is projected in a given state and the specific provider distribution in that state. If geographic conditions are put in place, states will need to decide how long the conditions last before a provider is deemed fully qualified and able to practice anywhere in the state. This period of time will likely range anywhere from two to five years depending on state specific conditions.
IV. Back-of-the-envelope assessment of possible impact
While there is a high degree of uncertainty in how many physicians would take advantage of this opportunity, as their ability to do so will partially depends on exogenous changes in the U.S. immigration environment, a conservative estimate can be made assuming take up on Conrad 30 Waivers would raise from an average of ~60% to 100% as administrative changes in federal immigration law are unlikely to curtail the Conrad 30 program. It is important to note the ~60% average figure is based on incomplete survey data as only 36 states had complete data across all years surveyed. An increase from 60% to 100% represents 600 new physicians entering the U.S. each year, assuming all 50 states adopt this policy change. Over a ten-year budget window this respents 6,000 new physicians.
Given that the AAMC projected physician shortage ranges from 40,800 at the low end to 104,900 at the high-end, this policy change could address anywhere from ~15% to less than 1% of the shortfall. Of course, if physicians are able to obtain other immigration visas under other categories, including those targeting the highly educated, takeup rates may be significantly higher. As there are 1.8 million practicing physicians in the European Union alone, there is a significant talent pool abroad that may be tempted to practice in the U.S.
Increasing the number of practicing physicians in the U.S. will likely impact HEDIS scores and access to care in rural areas. As the current physician shortage, and resulting lower concentration of physicians per capita in rural areas, have contributed to ~$300 more in per capita Emergency Department (ED) spending and ~$200 less in per capita primary care spending on Medicare beneficiaries in rural areas than in urban areas, it is likely that addressing the shortfall would decrease rural ED spend per Medicare beneficiary by ~$300 in affected areas and increase rural primary care spend per Medicare beneficiary by ~$200 in affected areas. This translates into net per beneficiary Medicare savings of $100 per capita. As there are currently 11.6M rural Medicare beneficiaries, closing the care gap for 15% of them will result in annual overall net cost savings to Medicare of ~$170M in the tenth year after implementation, and total savings over than ten-year period of ~$938M (calculated by adding an additional 600 providers each year over the course of the ten years to reach a total of 6,000 new providers in year ten). This rough calculation does not account for improvements in health outcomes and quality of life likely to result from improved provider interaction.
This calculation was made assuming that the care gap between urban and rural areas would not increase as the physician shortage increases over the next decade. This is likely unrealistic. If the quality of care delivery in rural areas will become progressively worse as the gap increases, this calculator understates the impact of accepting graduates of foreign-residency programs.
Another way to think about this, is to estimate the impact if a single state adopted this change and tapped into the foreign-trained physician talent pool to address its entire projected shortfall. It is realistic that a state would be able to attract sufficient foreign-trained physicians given pay differentials.⁴ For example, as the Texas Department of Health and Human Services project a shortfall of 3,375 providers by 2030, if Texas alone adopted this policy change and attracted 3,375 additional foreign-trained physicians in the first year, it would close ~8% of the projected 40,800 provider shortfall, representing ~$928M dollars saved for Medicare over the ten-year budget window.
If we assume all states implemented this change at once, and could recruit 40,800 providers in one year, the entire $1.1B annual gap could be closed, leading to ten-year budget savings to Medicare of $11B. All told, the total benefit from this change will likely be significant.
V. Next steps
This paper serves as a conceptual explanation of the policy proposal. It justifies the proposal by explaining qualitatively how it will remove a barrier to entry and provides an overview of decisions a state would need to make to implement a foreign residency recognition program. It also provides a high-level quantitative assessment of the impact of this change on Medicare expenditures.
As a next step, a non-exhaustive review of the top foreign residency programs should be initiated by a non-profit or academic organization to assess which residency programs meet ACGME standards. Such a review would enable a state that chooses to implement this reform to implement an initial list of accepted foreign programs quickly.
Additionally, states should consider adopting legislation establishing foreign residency recognition programs.
 The author contributed to the report cited while serving at the National Economic Council and Department of Health and Human Services.
 The author contributed to development of the Executive Order that directed CMS to make this change while serving at the Domestic Policy Council and the Department of Health and Human Services.
 Incidentally, this parallels the drug pricing subsidy the U.S. is in effect providing to other developed countries, as recognized in the International Pricing Index Advanced Notice of Proposed Rulemaking (ANPRM). Reducing the amount Americans pay for drugs while continuing to encourage innovation by encouraging other developing countries to pay more, is the key goal of that ANPRM. The author contributed to the development of this ANPRM while serving at the National Economic Council and the Department of Health and Human Services.
 See supra literature review for a discussion on why foreign-trained physicians seek to practice in the United States. Changes in federal immigration policy may prevent some from entering, making this analysis slightly less conservative than the one performed at a national level focused on the excess Conrad 30 Waiver Program spots.