Gozu, Aysegul MD, MPH; Kern, David E. MD, MPH; Wright, Scott M. MD
Approximately 25% of practicing physicians in the United States are graduates of medical schools outside of the United States and Canada; they are commonly referred to as international medical graduates (IMGs).1,2 This number is likely to rise, given both the increased demand for physicians for the aging population and the fixed supply of physicians graduating from medical schools in this country.3,4 The United States is not alone in its large number of IMGs; the number of practicing IMG physicians is also rising in other countries including the United Kingdom, Canada, Australia, and New Zealand.5–8
The U.S. health care system is dependent on IMGs to fulfill the health services needs for health services of the American population.1,3,4,9 In particular, IMGs are more likely to practice in specialties in which there have been shortages (including primary care medicine and psychiatry)3,9–12 and to serve socioeconomically disadvantaged patients.10,13,14 More specifically, IMGs disproportionately care for vulnerable segments of the American population in the following settings: (1) rural, underserved areas,15–18 (2) not-for-profit, community health centers chartered to provide care to the indigent,19,20 (3) clinics in impoverished, urban settings,19,20 and (4) Critical Access Hospitals.21
IMGs make up a large constituency in the U.S. graduate medical education (GME) enterprise. For 4,500–5,000 IMGs annually,22,23 GME is the entry point into U.S. medicine from more than 120 different countries.2,24 IMGs have trained at medical schools or health systems with varying levels of sophistication and resources.24 Thus, each trainee arrives with unique personal, cultural, and professional experiences and perspectives.
Residency training is a demanding time for all physicians. Some researchers believe that acculturation to a new health care system and way of life makes this transition additionally challenging for IMGs.25,26 The goal of this study was to identify personal, professional, and psychological similarities and differences between IMGs and U.S. medical graduates (USMGs) who work together in the same internal medicine (IM) residency training programs. We hypothesized that IMGs would score less favorably than USMGs on scales measuring fatigue, stress, self-esteem, and personal growth because of the added difficulties associated with their acculturation into both a new health care system and the United States.
Study design and participants
We conducted a cross-sectional survey of IMGs and USMGs working together within the same community-based IM residency training programs in Baltimore, Maryland. We identified all of the accredited IM programs, including both university and community programs, in the city through the Accreditation Council for Graduate Medical Education Web page.27 We chose IM residency programs for our study because there were many such programs within Baltimore and because IMGs are commonly trained in this discipline. We limited the study to a single metropolitan area for feasibility issues related to institutional review boards (IRBs), response rate, and study coordination. We contacted all program directors to determine the proportion of IMGs among their house officers (interns and residents). We had decided a priori that we would include only training programs with at least 25% IMGs in our study because some of our secondary research questions and hypotheses related to how IMGs and USMGs interact with each other. We excluded two training programs in Baltimore for which we were unable to discover the proportions of IMGs.
We recruited all house officers at each of the six programs that met eligibility criteria to participate. Because U.S.-born IMGs (n = 9) share commonalities with both IMGs and USMGs, we decided that we would exclude them from the analyses.
Survey development and content
We chose content areas for the instrument based on study hypotheses and a literature review. Two of us (D.K., S.W.) have expertise in medical education, survey development, and educational research. One of us (A.G.) is an IMG who came to the United States for IM residency training. Together, we iteratively revised the instrument over time with input from education researchers in the division of general internal medicine. We conducted pilot testing on recent graduates of residency programs (fellows and junior faculty members, both IMGs and USMGs) to refine the instrument.
The survey collected information from the following three domains: demographic characteristics, prior experience in medical practice, and career plans after residency. Embedded into the survey were four scales (all previously validated for content, internal construct, and relation to other variables) that we felt to be particularly germane to residency training: the Iowa Fatigue Scale,28 Cohen’s Perceived Stress Scale,29 Rosenberg’s Self-Esteem Scale,30 and the Personal Growth Scale.31
We administered the survey at each of the six programs during April and May 2006. One of us, the primary investigator (A.G.), worked with chief residents at the participating sites to find conference times that would provide exposure to the most housestaff. If residents and interns were unavailable at such conferences, we sent mailings with interdepartmental return envelopes addressed to chief residents to distribute in order to encourage full participation. We included an accompanying recruitment cover letter, approved by the IRB, that did not delineate our research hypothesis. Through this letter, we made respondents aware that all data were confidential and that only the study investigators would have access to data. When house officers completed their surveys at the conferences or delivered them to chief residents, they removed the identifying cards (bearing their names) from the surveys and placed them in separate envelopes.
The Johns Hopkins University School of Medicine IRB and the IRBs at each of the hospitals of the participating residency programs approved this study.
We examined the data for evidence of skewed data, outliers, and nonnormality to ensure that the use of parametric statistical tests was appropriate. To compare IMGs and USMGs, we used t tests or chi-square analyses depending on the variable type. Because the majority of the variables were binary, we dichotomized the others to facilitate interpretation in logistic regression analysis.
For the four validated scales that were embedded within our instrument (fatigue, perceived stress, self-esteem, and personal growth), we calculated Cronbach alphas to assess their internal consistency.
We used logistic regression to produce unadjusted odds ratios (ORs, with 95% confidence intervals) to characterize individual attributes associated with IMG status versus attributes associated with USMG status. We then used multivariable logistic-regressions to identify the variables that were independently associated with IMG status—as compared with USMG status. We included only variables that were associated with the dependent variable (IMG status) in the bivariate analysis at P < .10 in the final multivariable model. In the multivariable analyses, we (1) used generalized estimating equations (GEEs) to account for clustering effects from the six training programs, so as not to allow any single program to overly influence the results, (2) applied stepwise forward selection to model building wherein we added variables to the model based on their levels of significance, and (3) examined variables included in the models for evidence of colinearity and interactions. To assess the goodness of fit of the model, we applied the Hosmer-Lemeshow method based on deciles of risk.32
In an attempt to fully understand the differences between USMGs and IMGs with respect to fatigue, stress, self-esteem, and personal growth, we analyzed the four scales’ scores as continuous variables, and we dichotomized each of them into “≥ mean” versus “< mean” as others have done previously.31,33 Because the results were similar for both analyses, we have elected to present the dichotomized versions for the sake of clarity. We constructed four separate multivariable models which included (1) one of these four scales and (2) the variables that would turn out to be independently associated with IMG (versus USMG) status from the multivariable model described above. We again used GEEs to control for influence from any single training program. We never entered the four scales into modeling together because they were all highly colinear. The colinearity we observed among fatigue, stress, self-esteem, and personal growth may provide an explanation to other variables’ validity evidence and support the intended inferences of these scales.
We analyzed data using STATA 8.0 (STATA Corp., College Station, Texas).
This study had participation from all six IM residency programs in Baltimore wherein we could establish that IMGs made up more than 25% of the house officer pool (range: 31–92%). We received surveys from 176 of the 225 house officers (78%). Our usable response rate, after excluding the nine U.S.-born IMGs, was 74%. Overall response rates at the different programs ranged from 73% to 88%. There were no statistically significant differences between responders and nonresponders in terms of gender, year of training, or program (all P values were nonsignificant).
Characteristics of respondents and comparisons between IMGs and USMGs
Table 1 displays demographic and other characteristics of the IMGs and USMGs. We put these statistically significant variables into a single multivariable model.
In the multivariable model shown in Table 2, three variables were independently associated with IMG status: IMGs were (1) more likely to have a native language other than English, (2) more likely to have practiced medicine before the current residency training, and (3) less likely to have debt >$50,000.
Performance on the four scales
Iowa Fatigue Scale.
This scale asks informants to consider the previous six months when responding. Respondents answer the 11 items on the scale using a five-point scale (“Not at all” to “extremely”). Scores of 40 to 55 indicate severe fatigue, 30 to 39 indicates substantial fatigue, and any score below 30 is considered normal. IMGs reported feeling significantly less fatigued compared with USMGs (Table 3); their mean overall fatigue score was 28.1 (SD: 6.9) as compared with 32.6 (SD: 5.6) for USMGs (OR: 2.67; 95% CI: 1.18–6.03). The internal consistency of this scale in our population was good, with a Cronbach alpha of 0.84.
Perceived Stress Scale.
This scale asks informants to consider the previous month when responding. The scale measures the magnitude of the stress of daily life with four questions assessed on a five-point scale (“never” to “very often”). Higher scores are associated with higher stress; there are no clear cutoffs defined for this scale. IMGs reported feeling less stressed compared with the USMGs (mean overall stress score was 5.9 [SD: 2.9] versus 6.5 [SD: 2.8]) (Table 3); however, this difference was not statistically significant (OR: 1.74; 95% CI: 0.71–4.25). The internal consistency of this scale in our population yielded a Cronbach alpha of 0.75.
Rosenberg Self-Esteem Scale.
This scale asks respondents to reflect on their current self-esteem. Higher scores indicate higher self-esteem. IMGs reported higher self-esteem than USMGs: 33.8 (SD: 4.9) versus 32.2 (SD: 4.1) (OR: 2.98; 95% CI: 1.18–7.53) (Table 3). The internal consistency of this scale in our population yielded a Cronbach alpha of 0.87.
Personal Growth Scale.
When using this scale, investigators must specify an interval of time for respondents to consider. For this study, we asked participants to consider how they have evolved since the start of residency training. Respondents assess themselves with respect to nine items related to personal growth on a five-point scale (from “much worse” to “much better”). Higher scores are associated with greater personal growth. IMGs reported significantly more personal growth over the course of their training as compared with the USMGs. The mean overall personal growth score was 32.2 (SD: 5.3) for IMGs and 29.8 (SD: 5.4) for USMGs (OR: 3.61; 95% CI: 1.58–8.20) (Table 3). The internal consistency of this scale in our population yielded a Cronbach alpha of 0.87.
The differences in performance on the aforementioned scales between IMGs and USMGs in the bivariate analyses were only minimally influenced by multivariable modeling that adjusted for the three variables (indebtedness of less than $50K at graduation from medical school, a native language other than English, and experience practicing as an MD before current residency) that represent the independent differences between the two groups (Table 3, footnote). The differences in scores on these four scales between IMGs and USMGs also remained consistent across each of the six training programs.
The findings of this study show that IMGs are different from their USMG counterparts in several ways. This study provides new information about how IMGs view themselves and how they experience their residency training as compared with USMGs. The IMGs in our cohort were older, more likely to be married with children, spoke English as a foreign language, and were less likely to have debt related to their medical school training.
The IMGs traveled from the corners of the globe to come to Baltimore for their IM residency training. Our hypotheses that the international move, learning the nuances of American medicine, and the reservations or biases toward foreigners34 would result in reduced self-esteem coupled with more fatigue and stress were not borne out by the data. Our data show that IMGs suffered less fatigue, felt higher self-esteem, and experienced more personal growth than their USMG counterparts. Perceived stress was not statistically different between the groups.
Prior work has suggested that, on the whole, IMGs are not as strong as USMGs in terms of clinical skills35 and communication skills.36,37 Yao and Wright38 found that IMGs were more likely to be characterized as “problem residents”—defined as “trainees who demonstrate significant enough problems that require intervention by someone of authority, usually the program directors or chief resident.” Investigators have also shown that attrition rates during the residency training period are higher among IMGs.39
The transition for IMGs as they begin to work as house officers in U.S. hospitals is not easy. Their inexperience with the new system of care coupled with adapting to a new culture may directly affect both their relationships with others at work and the quality of patient care that they are able to render.24,40,41 Some have presumed that these challenges faced by IMGs may predispose them to more fatigue, higher stress, and lower self-esteem than their USMG counterparts experience.42–44 Prior published articles about the training experiences of IMGs have recounted the following struggles: (1) residing in an alien environment,25,44,45 (2) working in an unfamiliar health care system,24,44 (3) struggling with language barriers,36,37,45–47 (4) experiencing incongruence between personal and American medical values or philosophies,46,48–50 (5) lacking knowledge of American culture,25,43,50 and (6) fearing discrimination.34
Why, then, do the IMGs seem to be coping better with residency training in this study? Several explanations warrant consideration. It may be possible that the IMGs in this cohort are particularly strong or that the USMGs studied are weaker than those described in prior reports. Both possibilities may be operating jointly because, at some community-based residency training programs, particularly in a less popular discipline like IM, the USMGs may matriculate from the lower quintiles of their medical school class, whereas most IMGs that come to the United States for GME have performed well in medical school (upper quintiles).51,52 The IMGs may feel more energized and enthusiastic about residency training, having persevered and overcome major challenges25,53 (such as obtaining visas and securing funds to travel for United States Medical Licensing Exams) for the opportunity to pursue their GME training in the United States. Also worth noting is that prior work comparing USMGs and IMGs has largely focused on knowledge, skills, and competence36,39,54 as opposed to the intrapersonal constructs of fatigue, self-esteem, stress, and personal growth investigated in this report.
At these residency training programs, IMGs constitute 65% of the residents and, as such, could in fact be construed the “majority” group. However, in reality, every single IMG is a minority unto himself or herself. As seen in Table 1, they represent many different countries from five different continents.
Several limitations of this study should be considered. First, we used self-assessed reports rather than objective measures of fatigue, stress, self-esteem, and personal growth. However, previous investigators have used these self-assessed scales in research with varied populations, and the scales have established validity evidence.28–31 Cronbach alphas in this study provided further evidence of the construct validity of the scales with our study population.
Second, we studied only IM residency trainees in Baltimore, and the results may not be generalizable to all other house officers across the United States. However, our sample did include both university-affiliated and non-university-affiliated community-based programs. Third, this cross-sectional study collected data at a single point in time and, therefore, did not capture changes that may have occurred over time during the residency training period. Finally, this study used quantitative methods only. Qualitative methods may have also resulted in the emergence of rich understandings of differences between IMGs and USMGs.55
This study has described important differences between IMGs and USMGs working together in U.S. residency training programs. Understanding and considering these differences may allow educators and training program directors to more effectively support and encourage their diverse cadre of trainees—perhaps thinking creatively about the strengths and needs of individual trainees. Pairing IMGs and USMGs to create a learning community in which house officers can support one another on the basis of their individual talents and cultural perspectives would represent an innovation that fosters respect, humanism, and professional development.
Dr. Wright is an Arnold P. Gold Foundation Professor of Medicine. Dr. Wright is also a Coulson-Miller Family Scholar; the support is associated with the Johns Hopkins Center for Innovative Medicine.
The authors are indebted to Mr. Ken Kolodner, Ms. Cheri Smith, and Dr. Gerald Whelan for their assistance. They would also like to thank Dr. Suzanne Caccamese, Dr. Morteza Farasat, Dr. Shaili Gupta, Dr. Nicolai Mejevoi, Dr. Raez Siddiqui, Dr. Izukanji Sikazwe, and Dr. David Zolet for their support.
1 Mullan F. The metrics of the physician brain drain. N Engl J Med. 2005;353:1810–1818.
3 Akl EA, Mustafa R, Bdair F, Schünemann HJ. The United States physician workforce and international medical graduates: Trends and characteristics. J Gen Intern Med. 2007;22:264–268.
4 Cooper RA. Weighing the evidence for expanding physician supply. Ann Intern Med. 2004;141:705–714.
7 Birrell RJ. Australian policy on overseas-trained doctors. Med J Aust. 2004;181:635–639.
9 Cohen JJ. The role and contributions of IMGs: A U.S. perspective. Acad Med. 2006;81:S17–S21.
10 Steward DE; Association of Professors of Medicine. The internal medicine workforce, international medical graduates, and medical school departments of medicine. Am J Med. 2003;115:80–84.
11 Rao NR. Recent trends in psychiatry residency workforce with special reference to international medical graduates. Acad Psychiatry. 2003;27:269–276.
12 Rao NR. Psychiatric workforce: Past legacies, current dilemmas, and future prospects. Acad Psychiatry. 2003;27:238–240.
13 Blanco C, Carvalho C, Olfson M, Finnerty M, Pincus HA. Practice patterns of international and U.S. medical graduate psychiatrists. Am J Psychiatry. 1999;156:445–450.
14 Howard DL, Bunch CD, Mundia WO, et al. Comparing United States versus international medical school graduate physicians who serve African-American and white elderly. Health Serv Res. 2006;41:2155–2181.
16 Baer LD, Ricketts TC, Konrad TR, Mick SS. Do international medical graduates reduce rural physician shortages? Med Care. 1998;36:1534–1544.
17 Baer LD. A proposed framework for analyzing the potential replacement of international medical graduates. Health Place. 2003;9:291–304.
18 Hart LG, Skillman SM, Fordyce M, Thompson M, Hagopian A, Konrad TR. International medical graduate physicians in the United States: Changes since 1981. Health Aff (Millwood). 2007;26:1159–1169.
19 Rosenblatt RA, Andrilla CH, Curtin T, Hart LG. Shortages of medical personnel at community health centers: Implications for planned expansion. JAMA. 2006;295:1042–1049.
20 Baer LD, Konrad TR, Miller JS. The need of community health centers for international medical graduates. Am J Public Health. 1999;89:1570–1574.
21 Hagopian A, Thompson MJ, Kaltenbach E, Hart LG. The role of international medical graduates in America’s small rural critical access hospitals. J Rural Health. 2004;20:52–58.
22 Brotherton SE, Etzel SI. Graduate medical education, 2006–2007. JAMA. 2007;298:1081–1096.
24 Whelan GP. Commentary: Coming to America: The integration of international medical graduates into the American medical culture. Acad Med. 2006;81:176–178.
25 McMahon GT. Coming to America—International medical graduates in the United States. N Engl J Med. 2004;350:2435–2437.
26 Maudsley RF. Assessment of international medical graduates and their integration into family practice: The clinician assessment for practice program. Acad Med. 2008;83:309–315.
28 Hartz A, Bentler S, Watson D. Measuring fatigue severity in primary care patients. J Psychosom Res. 2003;54:515–521.
29 Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav. 1983;24:385–396.
30 Rosenberg M. Society and the Adolescent Self-Image. Rev ed. Middletown, Conn: Wesleyan University Press; 1989.
31 Wright SM, Levine RB, Beasley B, et al. Personal growth and its correlates during residency training. Med Educ. 2006;40:737–745.
32 Hosmer DW, Lemeshow S. Applied Logistic Regression. 2nd ed. New York, NY: John Wiley & Sons Inc.; 2000.
33 Diaz A, Simantov E, Rickert VI. Effect of abuse on health: Results of a national survey. Arch Pediatr Adolesc Med. 2002;156:811–817.
34 Fiscella K, Roman-Diaz M, Lue BH, Botelho R, Frankel R. ‘Being a foreigner, I may be punished if I make a small mistake’: Assessing transcultural experiences in caring for patients. Fam Pract. 1997;14:112–116.
35 Whelan GP, Boulet JR, McKinley DW, et al. Scoring standardized patient examinations: Lessons learned from the development and administration of the ECFMG clinical skills assessment (CSA). Med Teach. 2005;27:200–206.
36 Hall P, Keely E, Dojeiji S, Byszewski A, Marks M. Communication skills, cultural challenges and individual support: Challenges of international medical graduates in a Canadian healthcare environment. Med Teach. 2004;26:120–125.
37 Mylonakis E, Mega A, Schiffman FJ. What do program directors in internal medicine think about international medical graduates? Results of a pilot study. Acad Med. 1999;74:452.
38 Yao DC, Wright SM. National survey of internal medicine residency program directors regarding problem residents. JAMA. 2000;284:1099–1104.
39 van Zanten M, Boulet JR, McKinley D, Whelan GP. Attrition rates of residents in postgraduate training programs. Teach Learn Med. 2002;14:175–177.
40 Fiscella K, Frankel R. Overcoming cultural barriers: International medical graduates in the United States. JAMA. 2000;283:1751.
41 Kinchen KS, Cooper LA, Levine D, Wang NY, Powe NR. Referral of patients to specialists: Factors affecting choice of specialist by primary care physicians. Ann Fam Med. 2004;2:245–252.
42 Haveliwala YA. Problems of foreign born psychiatrists. Psychiatr Q. 1979;51:307–311.
43 McClain TA. IMGs and cultural diversity training. Acad Med. 1996;71:1138–1139.
44 Cole-Kelly K. Cultures engaging cultures: International medical graduates training in the United States. Fam Med. 1994;26:618–624.
45 Heal C, Jacobs H. A peer support program for international medical graduates. Aust Fam Physician. 2005;34:277–278.
46 Rhodes P. Some hurdles for the overseas doctor. Br Med J (Clin Res Ed). 1983;286:1136–1137.
47 Rolfe IE, Pearson SA. Communication skills of interns in New South Wales. Med J Aust. 1994;161:667–670.
48 Searight HR, Gafford J. Behavioral science education and the international medical graduate. Acad Med. 2006;81:164–170.
49 Kales HC, DiNardo AR, Blow FC, McCarthy JF, Ignacio RV, Riba MB. International medical graduates and the diagnosis and treatment of late-life depression. Acad Med. 2006;81:171–175.
50 Kramer MN. The educational needs of international medical graduates in psychiatric residencies. Acad Psychiatry. 2005;29:322–324.
51 Garibaldi RA, Subhiyah R, Moore ME, Waxman H. The in-training examination in internal medicine: An analysis of resident performance over time. Ann Intern Med. 2002;137:505–510.
52 McDonald FS, Zeger SL, Kolars JC. Factors associated with medical knowledge acquisition during internal medicine residency. J Gen Intern Med. 2007;22:962–968.
53 Shafqat S, Zaidi AK. Unwanted foreign doctors: What is not being said about the brain drain. J R Soc Med. 2005;98:492–493.
54 Whelan GP, Gary NE, Kostis J, Boulet JR, Hallock JA. The changing pool of international medical graduates seeking certification training in US graduate medical education programs. JAMA. 2002;288:1079–1084.
55 Fossey E, Harvey C, McDermott F, Davidson L. Understanding and evaluating qualitative research. Aust N Z J Psychiatry. 2002;36:717–732.