|
TRENDSThe International Medical Graduate Pipeline: Recent Trends In Certification And Residency Training
International medical graduates (IMGs) represent a large proportion of the population entering graduate medical education (GME) programs. Many of these internationally trained physicians go on to practice medicine in the United States. To be eligible for GME, IMGs must be certified by the Educational Commission for Foreign Medical Graduates (ECFMG). The number of certificates issued by the ECFMG has varied over time and historically has exceeded the number of available training positions. More detailed longitudinal analyses are required to better understand the interwoven issues of physician supply, consumers’ needs, and the role of IMGs in the U.S. health care system.
PHYSICIANS WHO DID not attend medical schools in the United States or Canada, referred to as international medical graduates (IMGs), play an integral role in the U.S. health care system. Such physicians now represent approximately 25 percent of practicing doctors nationwide.1 Because successful residency training is a prerequisite for licensed practice in the United States, IMGs also make up a sizable portion of physicians in graduate medical education (GME).2 The large number of IMGs in the United States, in both graduate training and licensed practice, demands that this physician cohort be considered when studying present and future U.S. health care practitioners’ needs.3 To enter a U.S. GME program accredited by the Accreditation Council for Graduate Medical Education (ACGME), IMGs are required to achieve certification by the Educational Commission for Foreign Medical Graduates (ECFMG).4 Requirements for certification have varied over time, but they have typically included both a check of medical credentials and one or more examinations.5 Throughout the credentialing process, the ECFMG has maintained comprehensive records on all applicants, including detailed demographics, examination results, and, more recently, residency training status. In addition, starting in 1971, the U.S. Department of State designated the ECFMG as the visa sponsor for all J-1 exchange visitor physicians who participate in clinical training programs. To train in a residency program, physicians who are not U.S. citizens or permanent residents must secure an appropriate work visa. Unfortunately, while additional information is available for J-1-sponsored IMGs entering GME, the majority of IMG trainees neither need nor use this non-immigrant visa pathway to work legally in the United States. More important, other than J-1 sponsorships, the ECFMG has no responsibility for tracking IMGs’ visa status. Historically, a large number of available GME positions have been secured by IMGs. In fact, there are more entry-level positions than available U.S. medical graduates (USMGs). Depending on societal needs and immigration issues, including available visa sponsorships and associated waivers to obligations to return to the home country following GME, IMGs can remain in the United States once they complete residency training, contributing to both the patient care and academic work-forces.6 Knowing the characteristics and specialty training choices of these people provides an early, albeit somewhat imperfect, marker of the future composition of the U.S. physician workforce. Obtaining an ECFMG certificate marks the point at which it becomes possible for an IMG to enter the U.S. health care system. Therefore, from a workforce perspective, it is crucial to know the number of internationally trained physicians who have the potential to train in U.S. GME programs. Likewise, since many IMGs remain in the United States after this specialty training, knowing their demographic characteristics is also important. We have two goals in this study: first, to document ECFMG certification trends during 1980–2004, and second, to describe residency placement of IMGs during 1995–2003.
Data sources. For the description of certification trends during 1980–2004, we used ECFMG records. Available ECFMG applicant information includes demographics such as citizenship at medical school, sex, age, country of medical school, year of certification, and so on; performance data (for example, U.S. Medical Licensing Examination, or USMLE, and Clinical Skills Assessment, or CSA); and various international medical school characteristics (such as location). For this study, we used citizenship at medical school and certification date. Once IMGs started GME in the United States, they were all required to have their certificates permanently validated by the ECFMG by submitting a Form 246. Permanent validation was essential since two obligatory elements of the certificate (Test of English as a Foreign Language, or TOEFL, and CSA) lapsed if the person did not obtain a qualified residency position within three years of its issuance. Permanent validation was instituted in 1995–1996 and dropped in 2004 when both TOEFL and the CSA were discontinued as certification requirements. To describe the residency placement of IMGs, we used information from Form 246, including the institution entered, institution address, graduate program identification, program specialty, and the date the applicant started the ACGME-accredited program. Personal data were supplied by individual residents. Residency program information was supplied by an appropriate official in charge of the program. Form 246 data can be linked to ECFMG certification data through a common identifier. Since residency data for 2004 were incomplete, this part of the study was restricted to people who entered a program between 1995 and 2003. These data are based on IMGs whose certificates were permanently validated; thus, they do not necessarily include all IMGs in all programs, especially more recent GME trainees. Unfortunately, although visa information is known for certain people (such as those having ECFMG J-1 sponsorships) and can be linked to other ECFMG data, this group represents only approximately 20 percent of the IMG population in residency programs. Therefore, questions concerning visa status were not addressed in the analysis. International medical school graduates. Physicians who completed their medical education in schools outside the United States and Canada, regardless of citizenship, are considered IMGs; USMGs are physicians who have graduated from Liaison Committee on Medical Education (LCME)–accredited medical schools in the United States and Canada. Although most IMGs are not U.S. citizens, a sizable proportion of all ECFMG certificate holders were born in the United States or became U.S. citizens. For purposes of this study, a U.S. citizen IMG (USIMG) was defined as a person who was a U.S. citizen at the start of medical school. The remaining IMGs are referred to as non-USIMGs.
Trends in ECFMG certification. Between 1992 and 1998 there was a general upsurge in the total number of ECFMG certificates issued (Exhibit 1  ). In 1999 there was an appreciable drop in the number of certificates granted, followed by relatively stable volumes for the following five years.
Based on all certificates issued between 1980 and 2004, 13.9 percent have been awarded to USIMGs (Exhibit 1  ). This cohort represented approximately 20 percent of all certificates issued between 1982 and 1988. From 1989 to 1996 the number of certificates issued to U.S. citizens was much lower, averaging less than 600 per year. In 1997, USIMG certificate holders started to increase both in total numbers and as a percentage of the total certificate-holder population. During 2000–2004, 7,263 certificates were issued to USIMGs—22.9 percent of certificates issued.
Historically, ECFMG certificates have been issued to citizens of more than 170 countries (Exhibit 2
Concentrating on 1995–2004, we observed a pattern of decreasing numbers of certificates issued for many countries, including Pakistan, the Philippines, the former Soviet Republics, China, Germany, Egypt, Israel, and Australia. Based on percentages of all certificates issued, some countries, such as Syria, Iran, and Lebanon, increased their representation in the certificate-holder pool, albeit only slightly. In terms of overall growth, at least during 1995–2004, U.S. citizens represent an increasing proportion of the certificate-holder population.
IMGs in residency programs.
Between 1995 and 1998, the number of IMGs entering GME programs averaged about 5,400 (Exhibit 3
More than 61 percent of all IMGs certified between 1995 and 2003 obtained residency positions (Exhibit 4 ). As a group, more-recent certificate holders (1999–2001) had a higher probability of obtaining a residency position. Although only approximately one-third of 2003 certificate holders had obtained residency positions, this reflects the time it takes to get accepted to a graduate training program, the potential lag in filing the Form 246, and incomplete residency data for 2004. Overall, USIMGs were far more likely than others to have obtained residency positions.
For the five specialties in which IMG most often train, we calculated the percentage of IMGs by residency program entrance year (Exhibit 5 ). This exhibit provides information, both cross-sectionally and longitudinally, concerning IMGs’ specialty choices relative to all IMGs; it does not reflect the overall IMG composition of individual training programs.
Approximately half of all IMGs specialized in internal medicine during the years we examined. Family practice has seen relatively large growth in the number and percentage of IMGs. In 1995 only 6.3 percent of IMGs entering GME entered family practice residencies. In 2003, 15.8 percent did so. In contrast, between 1999 and 2003 the number of IMGs entering psychiatry residencies decreased from 8.5 percent to 5.8 percent.
In 1995, 4.7 percent of all residents entering programs in that year started residency programs in Pennsylvania (Exhibit 6
Over time, the ECFMG has issued certificates to a large number of IMGs. Although these people are required to meet strict, albeit changing, certification requirements, including having to pass many of the same licensure examinations as students who attended LCME-accredited medical schools, many still do not secure residency positions. This is certainly linked to a restricted supply of available positions, but it also could be related to changing individual preferences. From a workforce perspective, if the number of training positions should increase, this relatively large pool of certified but untrained physicians could be available to serve the American public in a relatively short period of time. For IMGs who secure positions, the certification requirements have ensured that from a basic and clinical science perspective, they are ready to enter supervised GME training programs in the United States.7 Overall, although there might be continuing controversy over the future role of IMGs in the U.S. health care system, this group of physicians makes up a sizable proportion of the workforce, caring for millions of patients each year.8 Changes in certification volume, especially for certain physician cohorts, will certainly affect GME training and physician supply, both in the United States and abroad. In 1992 the number of certificates issued by the ECFMG rose, probably reflecting IMGs’ desire to avoid the initial administration of the USMLE. In 1999 the overall number of certificates issued dropped considerably, likely at least in part because of changes in ECFMG certification requirements (that is, introduction of the CSA), computerized examination delivery, and examination cost. Regardless of the cause, changes in the number of IMG certificate holders, especially among people who are more likely to stay in the United States or who are more recent graduates, or both, could eventually have an impact on physician supply and, depending on their numbers, relative abilities, and distribution, the quality of care. In parallel with recent increases in number of ECFMG certificates issued to U.S. citizens, the number of USIMGs entering GME programs has also grown. They now account for approximately 20 percent of the approximately 6,000 IMGs in postgraduate year 1 GME training positions. Although this trend might simply reflect a lengthier process, often involving visa petitions, that non-USIMGs must endure to secure residency positions, USIMGs, as a percentage of the overall yearly certification volume, were far more likely than non-USIMGs to secure training positions. Here, certain selection-related factors (such as clinical experience in the U.S. medical care systems and immigration and visa issues) could favor the USIMG group. Regrettably, although studies have been completed to document where USIMGs obtain their medical degrees, specific information concerning the quality of their training is limited.9 If USIMGs are going to continue to make up a large proportion of the IMG residency population, and assuming that these people are more likely to eventually become part of the practicing U.S. physician pool, physician workforce projections, including practice patterns and locations, need to account for this trend.10 Similar to results from previous research, our data show that IMGs tend to be selected, at least initially, to primary care residency programs.11 Training in a primary care specialty does not preclude eventual subspecialization; however, the majority of IMGs who finish internal medicine training do not subspecialize.12 Furthermore, their subspecialization rate is only marginally higher than that of U.S. graduates. Therefore, if the number of IMGs in GME positions were to change, and specialty choices remain steady over time, access to generalist physicians could be affected. For some specialties (such as family medicine), the overall number and proportion of IMGs in programs have been growing steadily. Interestingly, the percentage of U.S. students matching to primary care specialties (internal medicine, pediatrics, and family medicine) peaked at 53.2 percent in 1998 and declined to 44.2 percent in 2002.13 Some primary care functions might be being met by nurse practitioners and physician assistants; however, increases in the number of IMGs training in the associated specialties would suggest that this group is filling gaps left vacant because of changes in U.S. medical students’ career preferences. For IMGs who are not U.S. citizens or permanent residents, this might reflect their willingness to enter specialties that could increase their odds of being able to remain in the United States because of societal need for their services.14 We also found that IMGs’ residency locations tend to be concentrated in relatively few states, with a large proportion training in New York. This finding in itself is not remarkable, in that a large percentage (13.4 percent) of the approximately 8,000 residency programs are located in this state.15 More important, for states such as California, the number of IMGs entering training programs has dropped considerably over time. This trend could be the result of USMGs’ residency choices or the disapproval of certain international medical schools by the licensing division of the Medical Board of California, but it highlights the nonuniform distribution of IMGs in residency training, both longitudinally and geographically.16 These data are important from a practice perspective, since IMGs tend to remain in the same state as their GME training, practice where IMG networks already exist, and locate in communities with large numbers of people of the same ethnicity.17 Given the overall disparity in the racial/ethnic distribution of the training physician population relative to the U.S. population, tracking the distribution of IMGs in residency training programs—especially those who constitute underrepresented minorities—will certainly be worthwhile.18 Changes in the characteristics of certificate holders and IMG residents underscore the complex nature of physician migration. Unfortunately, although changes in ECFMG certification rates point to patterns of migration to the United States, data are not available to classify the reasons for relocation or, for the non-USIMG cohort, the complex and ever-changing immigration pathways that must be navigated to secure employment. Although not a particular focus of this study, world events and U.S. immigration policy could certainly affect the flow and composition of foreign-national physicians coming to the United States. Even without more specific research, recent increases in the number of certificates issued to U.S. citizen IMGs combined with comparable decreases in other country-based cohorts (such as Pakistan) suggest that an array of political, economic, and cultural factors influence physicians’ migration to the United States. Here, from both the policy and planning perspectives, it would certainly be valuable to gather and analyze data related to the specific visa pathways that non-USIMGs follow to gain entry into the U.S. medical system. More specifically, for physicians who stay in the United States, especially those who were initially sponsored with an obligation to return home (those using J-1 visa) and subsequently received a waiver to that obligation, a detailed analysis of their practice patterns and the associated benefit to rural or underserved areas is required. Taken as a whole, additional focused national studies that relate physician supply, GME training, immigration policy (for example, changes in the apportion of and regulations concerning work-based visas or use of visa waiver programs to stay and practice medicine in the United States), and societal health care needs are required to address these and other workforce issues.19
John Boulet (jboulet{at}ecfmg.org) is assistant vice president, Research and Evaluation, at the Educational Commission for Foreign Medical Graduates (ECFMG) in Philadelphia, Pennsylvania. John Norcini is president of the Foundation for Advancement of International Medical Education and Research (FAIMER) in Philadelphia. Gerald Whelan is vice president, Assessment Services; James Hallock, president; and Stephen Seeling, vice president, Operations, of the ECFMG. This research project was supported by the Educational Commission for Foreign Medical Graduates and the Foundation for Advancement of International Medical Education and Research. However, the findings and conclusions do not necessarily reflect the opinions of these organizations.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Study Data And Methods
Study Results
