ON THE COVERON THE COVER FREE

Boy Walking by Karen Shore, Boston University School of Medicine, Boston, Mass

Over the past 5 years, more than 20,000 graduates of foreign medical schools have entered residency programs in the United States to obtain graduate medical education. The vast majority of these international medical graduates (IMGs) have gone on to practice in the United States after completing their training. Consequently, the number of IMGs has risen steadily reaching nearly 25% of all allopathic physicians practicing in the United States.¹

Considerable controversy surrounds this influx of physicians from abroad at a time when the US health care system is undergoing dramatic change. Concern over a potential physician surplus has caused several prominent policy groups to call for a cap on the numbers of foreign national IMGs permitted to remain permanently in the United States.² With growing pressure to curb Medicare reimbursements to teaching hospitals, policy analysts have also questioned the allocation of funds for graduate medical education to pay for residency positions for IMGs.³

Supporters of the open-door policy toward IMGs have countered that these physicians practice in disproportionately high numbers in areas of the country that have been neglected by the US health care system. For some time, workforce experts have debated whether such IMGs provide a national service by alleviating the geographic maldistribution of physicians throughout the country.⁴ Others ask why the United States, with its overall wealth and burgeoning medical education system, is not able to provide care in so many of its communities using physicians trained in its own medical schools

At the same time, the widespread perception that IMGs receive inferior training in foreign medical schools continues to dog IMGs' efforts to prove that they provide quality care to these communities.⁵ The creation of the new Clinical Skills Assessment (CSA) exam as a requirement for IMGs seeking entry into a US residency program is evidence of concern that IMGs do not have adequate clinical and interpersonal skills to care for US patients.

This issue of MSJAMA explores some of the questions raised by IMGs' controversial presence in the United States. How have recent immigration restrictions on foreign national physicians changed the composition of the US physician workforce? To what extent do IMGs provide care to segments of the population that have traditionally been neglected by US medical graduates? What sort of cultural barriers must IMGs overcome in order to deliver care to patients in the United States? What role do clinical-based exams such as the CSA play in standardizing the quality of care provided by all graduates entering residency training?

Over the past 10 years, foreign national physicians have had widely varying success in applying for permanent residency status in the United States. Since 1990, one path by which these physicians have remained in the United States is to apply for national interest waivers to circumvent traditional immigration certification requirements.¹ An examination of the history of these waivers reveals some of the conflicting priorities that inform the United States' immigration policies toward foreign physicians.

US immigration laws contain certain provisions that enable foreign nationals to qualify for permanent residence under the sponsorship of their employers. In most instances, employers must undergo a complex recruitment and advertising effort to prove that qualified US workers are not available to fill the position in question.² This application process is generally complicated, time-consuming, and intense in that it requires employers to demonstrate that they are not decreasing the availability of jobs for US citizens by employing workers from abroad. Under this framework, foreign physicians are regarded primarily as a threat to the job security of US national doctors.

However, an alternate pathway to permanent residence exists on the premise that the immigration of a foreign national can carry broad benefits to the United States. A recent initiative arising from the Immigration Act of 1990, the "national interest waiver classification" pathway, has antecedents in previous legislative enactments intended to induce socially beneficial cooperative behavior.³ Under this framework, the United States recognizes that foreign physicians willing to work in medically underserved areas provide "gap-filling" services that contribute to the overall welfare of the population. Such physicians can apply to be excused from the traditional certification process if they provide documentation that it is in the "national interest" of the United States to admit them as permanent residents.

Starting in 1998, however, there emerged a growing pattern of denials of national interest waiver cases involving foreign physicians. These denials were based on the argument that foreign national physicians who worked in medically underserved communities were providing a local rather than a national service,⁴ or that these physicians needed to work more than the 3 years stipulated for J-1 exchange-visit visas in order to make a true and lasting impact on the community.⁵ This policy change coincided with the appearance of work force reports arguing that IMGs departed prematurely from at-risk communities, as well as a growing movement in the US medical community advocating more rigid policies toward foreign physicians.^{6 - 7}

This resistance eventually matured into a consistent denial of national interest waiver requests filed on behalf of foreign physicians as set forth in Matter of New York State Department of Transportation (NYSDOT). After NYSDOT, the Immigration and Naturalization Service stated categorically that "there is no indication in the legislative history, statute, regulations, or binding legal precedent that physicians as a group are exempt from the labor certification requirement."⁸ Thereafter, employers seeking to recruit foreign physicians were forced to go through the traditional application process with attendant delays, uncertainty, and complexity. This situation created a particularly difficult situation for employers with special recruitment needs, such as those located in designated medically underserved communities or in facilities under the jurisdiction of the Department of Veterans Affairs.

Quite recently, Congress' concern about patterns of physician distribution has given rise to legislation entitled the Nursing Relief for Disadvantaged Areas Act of 1999, which addresses the shortage of nurses and physicians in designated areas of the United States.⁹ Signed into law by President Clinton on November 12, 1999, this bill is the culmination of a decade of debate on the role of immigration policies in serving the nation's health care agenda. As a result, physicians working within designated medically underserved areas and/or within the Veterans Affairs' system can again qualify for a national interest waiver.

The crux of this proposal is that physicians working in medically underserved areas provide a national service by alleviating the maldistribution of physicians in the United States. This measure does not create an unqualified entitlement to permanent resident status, as foreign physicians must fulfill a 5-year service commitment to a medically underserved area prior to obtaining permanent resident status. Nevertheless, it is the hope of workforce analysts that it will facilitate the recruitment and retention of physicians in areas that have been underserved by the health care system.

On July 1, 1998, the Educational Commission for Foreign Medical Graduates (ECFMG) implemented its Clinical Skills Assessment (CSA) as a new requirement for graduates of foreign medical schools seeking certification for entry into an accredited residency training program in the United States. The CSA is a day-long practical examination designed to assess graduates' ability to gather and interpret clinical data and to communicate effectively with patients and health professionals in English.

From the start, the CSA was conceived as only 1 of several assessment elements leading to certification by the ECFMG. Others include passing scores on the United States Medical Licensing Examination (USMLE) Steps 1 and 2, completion of an English comprehension test, and graduation from a medical school listed in the World Health Organization directory.

The CSA developed as a result of widespread concern among medical educators that basic clinical skills, including medical history-taking, physical examination, and doctor-patient communication, were not being adequately addressed in undergraduate medical education.¹ Within the United States, the Liaison Committee on Medical Education (LCME) responded to this concern by incorporating requirements on the teaching and assessment of basic clinical skills into its criteria for accreditation of US schools.

However, no international equivalent of the LCME was available to develop similar curriculum requirements for medical schools outside of the United States. Educators were concerned that medical schools around the world were variable in quality and not always subject to outside review.² Since neither the ECFMG nor any other agency could directly impact schools' curricula, the CSA was developed as a tool for evaluating educational outcomes.

To develop a process that yielded a reliable and fair assessment of clinical skills, a number of preliminary studies were undertaken under the auspices of the ECFMG. Two of the larger-scale studies involved both US medical students in their third and fourth years and graduates of foreign medical schools at varying levels of certification. The first, conducted in Baltimore, Md, involved a consortium of 6 schools; the second, conducted in Philadelphia, Pa, involved 2 local schools.

The introduction of this assessment is a milestone in US medical evaluation in that it is the first time basic clinical skills, including interpersonal skills and spoken English proficiency, are being assessed in a high-stakes environment. To ensure that the test is administered in a standardized manner, the CSA is offered only at the ECFMG headquarters in Philadelphia. The CSA uses standardized patients (SPs), many of whom are actors, to portray cases to the candidates and to score the encounters. SPs are recruited not for the presence of physical findings but for their ability to portray cases and score accurately; each receives 15 to 20 hours of training on assessing interpersonal skills and English proficiency, plus 6 hours of training for each case he or she portrays.

The ECFMG now has had experience with 8383 candidates who presented themselves for assessment at its center from July 1, 1998, through January 31, 2000. While the test is administered throughout the year, results are standardized through an analysis and equation of mean scores received on each case-SP combination. Analysis shows that 96.9% of these candidates received a passing decision. (CSA reports only pass/fail decisions, not numerical scores.)

Evidence suggests this high pass rate is a reflection of considerable self-selection on the part of the certification candidates. Describing their experience in questionnaires distributed after the examination, 80% of candidates reported making special preparations, including clinical observerships in the United States. Candidates also reported prescreening their spoken English proficiency by taking the Test of Spoken English, and only 3 candidates reported having scored below the level suggested in the CSA orientation manual as a prudent cutoff. Trial runs that were used to set CSA examinations standards involved test takers who had no investment in the results.

Successful candidates must meet or exceed standards for 2 separate components of the CSA: an Integrated Clinical Encounter (ICE) component assessing ability to gather data and compose a clinical note, and a Doctor-Patient Communications (COM) component based on spoken English proficiency and interpersonal skills. To date, 80% of those failing the CSA have done so on the COM side. Of the cohort described, 92% reported that SP portrayals in the test were realistic, and 90% felt that the SP format was appropriate.

The most widely heard complaint made by graduates of foreign medical schools regarding the CSA is that it is unfair for US medical graduates to be exempt from taking a similar national assessment examination. The National Board of Medical Examiners is currently piloting a Standardized Patient Examination for possible incorporation into the USMLE series, and, if adopted, this would be open to graduates of foreign schools in lieu of a separate CSA.

The United States' policies regarding international medical graduates (IMGs) have been hotly debated^{1 - 2} since the number of IMGs entering training more than doubled from 1988 to 1994.³ In the 1998-1999 academic year, IMGs added an additional 29% (5134) to the 18,000 US medical graduates entering allopathic and osteopathic training programs.⁴

This influx of IMGs into training programs comes at a time of widespread concern about a potential surplus of physicians and about rising costs of Medicare reimbursement to teaching hospitals for graduate medical education (GME). As a result, some workforce and professional groups have called for restrictions on the numbers of IMGs permitted to enter training and to remain permanently in the United States.⁵ However, others argue that IMGs have historically filled gaps in the physician workforce by practicing in geographic areas where US medical graduates (USMGs) prefer not to practice and by choosing specialties that USMGs do not wish to enter.^{6 - 7}

As a major center for GME in the United States, the state of New York provides a useful window into the experiences of newly trained physicians. Although only 7% of the US population resided in New York in 1999, nearly 15% of all allopathic residents and 29% of all IMG residents trained in New York during that year.⁴ A survey of residents completing training in New York therefore provides a meaningful perspective on the practice patterns and job market experience of both IMGs and USMGs.

Based on data from teaching hospitals and the American Medical Association's (AMA's) GME database, the Center for Health Workforce Studies at the University at Albany determined the number of physicians expected to complete an allopathic or osteopathic residency program in New York state in 1999. Excluding those in preliminary and transitional programs, approximately 4700 residents in 1140 programs were found to be in their last year of training in New York.

A survey with questions addressing the demographic characteristics, educational experiences, postresidency plans, and job market prospects of graduating residents was mailed to all teaching hospitals in New York State. The hospitals distributed the surveys to individual allopathic and osteopathic residency programs, with instructions to distribute them to residents during their last 6 weeks of training. When completed, the surveys were sent back to the Center by the same route.

The Center received completed surveys from more than 3400 residents for a response rate of 73%. The response rate for residents in primary care fields (family practice, general internal medicine, pediatrics, and internal medicine and pediatrics combined) was 77%. Of the respondents, approximately 48% were USMGs; 52% were IMGs.

The demographic characteristics of the survey respondents closely matched those reported in the AMA GME database for residents completing training in New York in 1999. In particular, the percentages of survey respondents who were IMGs versus USMGs were consistent with those documented in the AMA GME database. This led the authors to conclude that the survey respondents were representative of all residents completing training in New York during the year of study.

Four distinct groups of IMGs were distinguished on the basis of citizenship status: naturalized citizens/permanent residents (45% of IMGs); exchange visitor physicians with J visas (41%); physicians with temporary worker H visas (7%); and native-born US citizens who had obtained medical education outside the United States (US-IMGs) (7%). As US-IMGs and IMGs who are naturalized citizens or permanent residents enjoy similar immigration status, they were considered as a single group for data analysis. Physicians with J and H visas, both of which are US temporary immigration visas, were also grouped together for the purpose of analysis.

The results of the survey suggest certain trends in the practice and employment characteristics of graduating New York residents based on their citizenship and visa status. IMGs, particularly those with temporary visas, were more likely to train in primary care specialties, internal medicine subspecialties, and psychiatry than were USMGs. For example, 57% of temporary visa-holding IMGs surveyed were completing training in the primary care fields of family practice, general internal medicine, general pediatrics, or internal medicine and pediatrics combined, compared to 37% of USMGs surveyed (P<.001).

IMGs with temporary visas were more likely to subspecialize than were USMGs. For example, 62% of temporary visa-holding IMGs in internal medicine were planning to subspecialize, compared with 36% of USMGs in internal medicine (P<.001). IMGs who were US citizens or permanent residents subspecialized at a similar rate as USMGs (Figure 1). Of all graduating residents with J visas, 18% were planning to leave the United States immediately following completion of training. Excluding those going on to subspecialize, 28% were planning to leave the United States.

When the numbers of temporary visa-holding IMGs who were subspecializing or departing from the country were taken into account, the percentage of these IMGs actually entering primary care practice in the United States was approximately equal to that of USMGs (36% vs 38%, P=.409) (Table 1).

Of temporary visa holding IMGs with confirmed plans to practice in the United States, 84% were planning to practice in designated health professional shortage areas (HPSAs), compared with only 11% of USMGs (P<.001) and 6% for other IMGs (P<.001). However, when calculated as a percent of all graduating residents regardless of their plans (including graduates that were subspecializing or leaving the country), the percent of J visa physicians planning to practice in HPSAs dropped to 15%, and the figure for USMGs dropped to 5% (P<.001).

Temporary visa-holding IMGs expressed significantly more difficulty finding a satisfactory practice opportunity than did USMGs or permanent residents/US citizen IMGs. For example, 44% of temporary visa-holding IMGs indicated that they had to change their plans because of limited practice opportunities, compared with only 15% (P<.001) for USMGs and 24% (P<.001) for other IMGs.

The results of this survey indicate that international and US medical graduates in New York State differ significantly in their training and practice characteristics. In addition, significant variations in posttraining plans appear to exist among IMGs, depending on their specific citizenship and visa status.

While it may be that IMGs entering primary care fields are filling a gap created by USMGs, the high percentage of IMGs in certain subspecialties also reflects their willingness to enter specialties that allow them to remain in this country while in training, even if demand for their services is low.

The results of this study indicate that IMGs holding temporary visas are more likely than other IMGs to practice in health profession shortage areas. Given the higher proportion of IMGs with temporary visas who plan to subspecialize or to return to their native country after completing their primary care training, the contribution of IMGs to primary care in underserved areas is not as dramatic as was previously thought.

Although US-IMGs and IMGs who are naturalized citizens or permanent residents are more likely than temporary visa-holding IMGs or USMGs to enter primary care specialties without subspecializing, few of them appear to go on to work in designated medically underserved areas and thus may not contribute to primary care in those areas as may have been thought.

It is widely accepted that cultural differences between physicians and patients can shape medical encounters and affect quality of care. A large body of research documents the effect of patients' race or ethnicity on physician decision-making and demonstrates that health-related beliefs affect health outcomes.^{1 - 2} While many training programs have focused attention on teaching US medical graduates (USMGs) to overcome cultural barriers, little attention has been given to the effect of cultural difference on encounters between international medical graduates (IMGs) and their patients.

As a framework for discussing the major cultural challenges facing IMGs in providing care to patients in the United States, we use a model³ that describes 3 core tasks of the health care provider—gathering data, developing rapport, and educating and motivating the patient—each of which can present difficulties to IMGs who are not fully acculturated to US society.

Residency programs trying to improve IMGs' data-gathering skills have focused primarily on teaching essential English-language skills to IMGs educated in non–English-speaking countries.⁴ Considerably less attention has been given to teaching IMGs to recognize regional patient dialects, colloquial speech, body language, and speech inflection, yet studies show that even IMGs who are proficient in standard English may find it difficult to understand patients' more subtle or informal means of communication.⁵ Formal English classes may not prepare an IMG to recognize a patient's use of the phrase "high blood" to refer to hypertension, for example, or to understand a substance user who says "one day at a time" to allude to the philosophy of Alcoholics Anonymous.

Developing rapport and responding to patients' emotions may also be a challenging task for medical graduates from non-Western countries. A 1999 study of the role of race and gender in the patient-physician relationship indicated that patients report more satisfaction in a given medical encounter when their physician is of the same race and/or ethnicity.⁶ Both IMGs and US medical graduates who belong to ethnic or racial minorities may find it difficult to establish rapport with patients who do not share their appearance or background.

Furthermore, many IMGs come from developing countries where epidemic disease, physician shortages, and disparities in education leave little room for exploration of the patient's story as a focus for clinical training. As a result, some educators have suggested that communication skills are not a primary concern of these countries' undergraduate medical curriculum.⁷ Graduates may find it difficult to shift priorities once they arrive in the United States.

Further complicating IMGs' efforts to establish rapport with their patients is the widely held but infrequently substantiated perception that IMGs are not as well-trained or as qualified as their US-trained counterparts.⁸ A 1997 study of residents in a primary care training program identified "fear of patient bias" as a major factor shaping IMGs' approach to their US patients.⁹ These fears of rejection or of being singled out may cause IMGs to be appear more aloof in their interactions with patients, or to insist more unyieldingly on their own authority as physicians.

Finally, differences between IMGs' and patients' health-related beliefs may affect patients' adherence to these physicians' medical advice. A 1990 study of foreign-born IMGs found significant disparities between patients' and IMGs' attitudes toward health-related issues such as family involvement in health care, the meaning of facial expressions, and the use of life-sustaining technology.¹⁰ Although it is not clear whether these differences were owing to residents' foreign upbringing or their education in a foreign medical school, they had the potential to change patients' trust in their physicians' instructions.

In the future, medical educators will need to take cultural differences into account when training IMGs for practice in the United States. Tools such as mixed IMG and USMG support groups, international student retreats, cultural-sensitivity training, and standardized patients may prove useful in achieving this goal.