Mok, Philip S. MD; Baerlocher, Mark O. MD; Abrahams, Caroline MPA; Tan, Eva Y.; Slade, Steve; Verma, Sarita LLB, MD
There is a substantial physician shortage in Canada, including an estimated five million Canadians without family physicians.1 A potential solution to physician shortages in Canada and other countries around the world suffering from physician shortages is to recruit physicians who have received their training abroad. International medical graduates (IMGs), defined in Canadian terms, are foreign-trained graduates who received their medical education abroad; these individuals include Canadian citizens, permanent residents, and foreign graduates on employment visas.2 Having a thorough understanding of physicians training at home and abroad may have important implications in the design of future health care policy and physician recruitment strategies for academic institutions in Canada and for other developed countries, which tend to pay physicians more. Also, changes in physician immigration policies in one country can have a profound impact on policies and trends in another, whether concerning physician immigration between developed countries, such as Canada and the United States, or between developed countries and developing countries, such as Canada and South Africa.
With the above concerns in mind, we carried out the study reported here to compare IMGs and Canadian-trained medical graduates (CMGs) who had completed residency or fellowship training in Canada from 1989 to 2007, in terms of age, gender, specialty, and practice location.
Data from the Canadian Medical Association published in 2008 demonstrated that IMGs make up 22% of physicians in Canada, which has not changed significantly over the past few decades.2 For IMGs to obtain their medical practice licenses in Canada, they must pass a rigorous set of qualification exams, language proficiency tests, and postgraduate training and assessment and, occasionally, sign return-of-service contracts.3 However, there have been many factors that prevent IMGs from obtaining medical practice licenses in Canada, including significant financial, social, location, and language barriers.4 Because of the challenges that IMGs need to overcome, the majority of IMGs wishing to practice in Canada choose to go through residency or fellowship training at a Canadian faculty of medicine to obtain their medical practice licenses.5
We obtained the data for our study from the Canadian Post-M.D. Education Registry (CAPER), which was established in 1986 through the cooperation of several Canadian national organizations and collects individual data for all residents and fellows from all 17 Canadian faculties of medicine. The national organizations that advise and support CAPER include those that represent medical schools (Association of Faculties of Medicine of Canada), residents (Canadian Association of Interns and Residents), physicians (Canadian Medical Association), medical licensing boards (Medical Council of Canada, College of Family Physicians of Canada, Royal College of Physicians and Surgeons of Canada), the federal government (Health Canada), and provincial/territorial governments.
The CAPER database is the most comprehensive database tracking postgraduate education in Canada. Data are published from CAPER annually and are available publicly at the CAPER Web site. CAPER publishes and shares only aggregate-level, nonidentifiable data. Ethical approval was not needed for this study because (1) none of us had access to individual-level, identifiable data, and (2) we did not observe or survey any participants.
CAPER provided us aggregate-level, nonidentifiable data on all CMGs and IMGs who completed residency or fellowship training in Canada from 1989 to 2007. To carry out our study, we
* compared data from the earliest and latest time periods to evaluate whether there has been a change over time between the characteristics of CMGs and IMGs;
* pooled data for five years from 1989–1993 and 2003–2007 to ensure a sufficient sample size for comparison, as CMGs substantially outnumber IMGs in each year of graduation;
* used descriptive statistics to compare CMGs and IMGs, as CAPER does not provide raw data for further data analysis;
* compared the CMG and IMG cohorts through descriptive statistics in terms of choice of specialty, age, gender, and practice location, following completion of post-MD training; and
* analyzed the practice locations for CMGs and IMGs using practice location via postal codes two years following completion of training. However, postal code information for the 1989 cohort of CMGs and IMGs was not available from the CAPER database. The practice location data for both CMGs and IMGs were therefore pooled from 1990 to 1994 and from 2003 to 2007.
Results and Discussion
A total of 8,501 CMGs and 1,828 IMGs completed post-MD training (i.e., residencies or fellowships) at Canadian institutions between 1989 and 1993 compared with 7,734 CMGs and 1,879 IMGs who completed such training between 2003 and 2007 (Table 1). Between 1989–1993 and 2003–2007, the proportion of CMGs who completed training in family medicine decreased from 54% (4,568/8,501) to 38% (2,921/7,734), whereas those training in medical or surgical specialties increased from 40% (3,402/8,501) to 55% (4,260/7,734). In contrast, the proportion of IMGs who completed training in family medicine increased from 19% (344/1,828) to 37% (699/1,879), whereas IMGs who pursued medical or surgical specialties decreased from 61% (1,123/1,828) to 51% (958/1,879) in the same time periods. From 1989–1993 to 2003–2007, the average age of CMGs and IMGs increased from 29.8 to 31.1 years and from 36.1 to 37.0 years, respectively (Table 2).
Between 1989–1993 and 2003–2007, the percentage of female graduates increased from 41% (3,471/8,501) to 52% (4,016/7,734) and from 28% (509/1,828) to 42% (791/1,879) for CMGs and IMGs, respectively (Table 3). The women-to-men ratios were calculated for CMGs and IMGs in each specialty. The specialties with the largest changes were family medicine and pediatric specialties. From 1989–1993 to 2003–2007, the women-to-men ratio in family medicine for CMGs and IMGs increased from 0.87 (2,123/2,445) to 1.70 (1,841/1,080) and from 0.58 (126/218) to 1.55 (425/274), respectively. The women-to-men ratio in pediatric specialties for CMGs and IMGs increased from 1.37 (193/141) to 2.21 (289/131) and from 0.52 (83/160) to 1.14 (75/66), respectively.
As shown in Figure 1, the top five countries from which IMGs completed their medical degrees in 1989–1993 were India (10.3%; 189/1,828), the United Kingdom (9.2%; 169/1,828), Poland (8.5%; 155/1,828), Saudi Arabia (7.2%; 132/1,828), and Ireland (6.4%; 117/1,828). Between 2003 and 2007, the top five countries from which IMGs obtained their medical degrees were Saudi Arabia (19.6%; 369/1,879), India (7.9%; 148/1,879), Kuwait (5.3%; 99/1,879), Egypt (5.2%; 98/1,879), and Pakistan (5.1%; 96/1,879). Saudi Arabia and India were the only countries that remained as top providers of IMGs to Canada during both time periods.
According to the 1990–1994 and 2003–2007 cohort data on practice location two years following completion of training, a total of 7,413 CMGs and 1,280 IMGs graduating in 1990–1994 and 7,113 CMGs and 1,206 IMGs graduating in 2003–2007 practiced in Canada (Figure 2). Among the CMGs who practiced in Canada, the proportion of CMGs who practiced in Ontario and Quebec remained relatively stable at 38% (2,818/7,413 and 2,671/7,113) and 26% (1,943/7,413 and 1,825/7,113), respectively (Figure 2). In the same time period, among IMGs who practiced in Canada, the proportion of IMGs who practiced in Ontario and Alberta increased from 42% (542/1,280) to 47% (563/1,206) and from 8% (104/1,280) to 13% (151/1,206), respectively (Figure 2).
Overview of the findings
The data demonstrate that IMGs tended to be older, more likely to be men, and more likely to pursue a career in family medicine than their CMG counterparts. The characteristics of age, gender, medical specialty, country of MD, and practice location are examined in more detail below. Although the explanations are specific to Canada, our hope is that the ideas presented in this section may be used by other countries to evaluate their own physician recruitment strategies and potentially start collecting data on IMGs by forming organizations similar to CAPER.
Characteristics of the IMGs and CMGs
The average IMG is older than the average CMG on completion of postgraduate training.
From 1989–1993 to 2003–2007, the average age of CMGs increased from 29.8 to 31.1 years, and the average age of IMGs increased from 36.1 to 37.0 years (Table 2). Although the average age for both CMGs and IMGs we studied was increasing over time, IMGs have traditionally been, and continue to be, older on average than CMGs. However, the age gap between CMGs and IMGs may be closing as CMGs graduating in 2003–2007 were older than their 1990–1993 CMG counterparts (Table 2).
There are several possibilities that may explain why IMGs during the time period of our study were older than CMGs. A recent study demonstrated that many IMGs have already completed some form of residency training overseas in their home countries and have mainly reentered residency training in Canada for licensing purposes.6 Another potential reason why IMGs are older is the competitiveness of the Canadian residency match service (CaRMS) process for IMGs. For instance, out of the 1,387 IMGs who entered the CaRMS match in 2009, only 292 IMGs were successful, leaving a total of 1,093 IMGs unmatched.7 IMGs who fail to match may be reapplying to residency year after year, increasing the average age of IMG applicants.
A qualitative study of the IMG recertification process in Canada demonstrated that many IMGs experienced training entry barriers, such as ambiguous selection criteria and lack of feedback, and were frustrated with the disproportionately small number of IMG training positions despite the physician shortage in Canada.8 However, the Canadian government may also wish to look at the political and moral implications of draining other countries of older physicians who have practiced in their home countries. Furthermore, if IMGs are older on completion of postgraduate training, they are more likely to retire earlier than CMGs, and the short-term solution of recruiting IMGs to benefit the Canadian population may not have as large an impact as the government had originally planned. Perhaps strategies to recruit younger IMGs to apply to Canadian training programs will ultimately lead to a younger workforce that can serve the Canadian population for a longer period of time. Similarly, other countries may consider recruiting younger physicians from abroad to serve their own populations for longer periods of time.
The traditional male-dominance gender gap has reversed among CMGs but still exists among IMGs.
The percentage of female CMGs increased from 41% (3,471/8,501) in 1989–1993 to 52% (4,016/7,734) in 2003–2007. In the same time period, the percentage of female IMGs increased from 28% (509/1,828) to 42% (791/1,879). Among recent graduates, there are now more women than men for CMGs, but the gender gap still exists among IMGs. The trend of increasing female CMGs is likely to continue as women represented nearly 58% (6,091/10,518) of first-year medical students enrolled in Canadian medical schools in 2009–2010.9 A recent review of medical school graduates in Canada, the United States, and the United Kingdom showed a steady increase during the past 50 years in female graduates, with women more likely to pursue primary care than their male counterparts.10 According to these trends, there will likely be a greater proportion of female CMGs and IMGs in the future, a trend likely to be echoed in other developed countries. In our study, women outnumbered men among both CMGs and IMGs in family medicine and pediatric specialties for 2003–2007. However, men represented the majority within the medical and surgical specialties. Several studies have demonstrated that female physicians saw fewer patients per hour than did male physicians, but had better communication skills and incorporated more preventative care into their practices.11 However, it is important to remember that if female physicians work fewer hours and see fewer patients per hour, this may potentially lead to an increased physician workforce environment. It may be an interesting study in the future, both in Canada and other countries, to evaluate physician productivity between physicians trained at home and abroad.
IMGs in Canada today have obtained their medical training from a variety of countries.
There has been a substantial shift in countries from which IMGs have obtained their medical degrees. From 1989 to 1993, 10.3% (189/1,828), 9.2% (169/1,828), and 8.5% (155/1,828) of IMGs obtained their medical degrees from India, the United Kingdom, and Poland, respectively. However, in 2003–2007, only 1.4% (27/1,879) of IMGs received their medical degrees from the United Kingdom, and 1.7% (32/1,879) received theirs from Poland. Instead, during the period of our study, Saudi Arabia became the main provider of IMGs in Canadian postgraduate training programs, representing 19.6% (369/1,879) of all IMGs in 2003–2007.
IMGs in Canada consist of two groups: foreign visa trainees and Canadian citizens/permanent residents. Permanent residents are landed immigrants in Canada who receive no financial support from their home countries. In contrast, foreign visa trainees in Canadian residency programs are sponsored by their home governments, who pay a subsidy to the Canadian government along with the salary and benefit packages for foreign trainees.12 Another benefit is that Canadian hospitals have foreign visa trainees as staff; these trainees provide important medical services to patients. After completing their training in Canada, these foreign visa trainees return to their home countries to practice, whereas permanent residents stay to practice in Canada. As a result, the foreign visa trainees have not contributed to reducing the physician shortage following their training in Canada. For instance, there has been an increase in the number of foreign visa trainees in Canada, particularly from the Middle East, as shown by Saudi Arabia's supply of IMGs increasing from 7.2% (132/1,828) to 19.6% (369/1,879) from 1989–1993 to 2003–2007 (Figure 1).
A potential solution to the problem of this physician brain drain of foreign trainees returning to their home countries after training in Canada is to increase funding for more residency positions for permanent resident IMGs to decrease the number of foreign visa trainees. Some experts have even suggested that a moratorium of two or three years be placed on foreign visa trainees.12 Though this may be too restrictive, the idea is that this would ensure that resources are used more effectively, as more positions would be made available for permanent resident IMGs, who are more likely to stay and practice in Canada. The concept of putting more resources into foreign-trained graduates who have already obtained landed immigrant or permanent resident status may also be beneficial for other countries with substantial physician shortages who are currently recruiting foreign-trained physicians.
Although a smaller percentage of CMGs are training in family medicine, there was an increase in the percentage of IMGs training in family medicine.
The percentage of CMGs who trained in family medicine declined from 54% (4,568/8,501) in 1989–1993 to 38% (2,921/7,734) in 2003–2007. In contrast, the percentage of IMGs who trained in family medicine increased from 19% (344/1,828) in 1989–1993 to 37% (699/1,879) in 2003–2007. A potential reason for the increase in IMGs pursuing family medicine is a set of criteria used by CaRMS. Before 2007, IMGs could only apply for residency positions in the “second round” of the match, after which CMGs had already matched to their programs of choice. IMGs likely have a tougher time getting residency positions in specialties in the second round compared with CMGs, so that might be one reason they take family medicine spots more often than Canadians do. Furthermore, from 2007 onward, IMGs can now apply for residency programs in Canada in the first round of the CaRMS match, either in direct competition in Quebec and Manitoba, or in separate IMG dedicated residency positions.13 This may lead to a significant change in the kinds of specialties IMGs pursue because residency positions are no longer dependent on CMGs. It may also be prudent for Canada and other countries to examine which specialties have a shortage of physicians and specifically recruit IMGs to pursue training in those specialties.
Although the data demonstrated that fewer CMGs pursued family medicine in 2003–2007, there has been evidence published that, at least in recent years, family medicine has become more attractive for medical students, as nearly 33% (804/2,438) of graduating CMGs in 2010 chose family medicine as their first-choice discipline.14 This may be due, in part, to the increased remuneration for family physicians in Canada. For example, family physicians in Ontario who transitioned from fee-for-service payments to Family Health Teams, which are similar to the proposed patient-centered medical homes in the United States, saw an increase of approximately 40% in family physicians' average net incomes from 2004 to 2009.15 The percentage of CMGs in Ontario who entered family medicine subsequently increased from 25% to 39% in the same time period.15 Although this does not demonstrate a causative effect, it highlights the potential role of remuneration in the recruitment and retention of physicians in Canada and other countries.
The majority of CMGs and IMGs practice in Ontario.
The largest proportion of CMGs and IMGs practiced in Ontario, regardless of the time period. Specifically, there are more IMGs who practiced in Ontario from 2003 to 2007 compared with 1990 to 1994 (Figure 1). However, the fact that the CAPER data from 1989 did not include postal code data may skew the results and underestimate the true difference when comparing the practice locations of CMGs and IMGs from 1989 to 1993. Nevertheless, these data likely reflect the decision the Ontario government made in 2004 to spend $26 million to double the number of IMG residency spots in Ontario to 200, which is more than all the spots in the other Canadian provinces combined.16 As a result, more IMGs are training in Ontario residency and fellowship programs than in the other provinces and are more likely to stay in Ontario to practice. Other provinces in Canada and other countries that have underserviced areas or a substantial physician shortage may wish to consider following Ontario's lead and increase the number of IMG residency training spots in order to recruit and retain more IMGs to practice in their provinces. As IMGs have also been shown to be more likely than CMGs to practice in rural and small communities, increasing residency spots based in rural communities for IMGs could be a potential solution to the family physician shortage in rural communities.
There are several potential limitations in this study. The first involves the study population. Because this study relied on data from CAPER, data missing from the CAPER database are not reflected in our findings (e.g., the missing postal code information for 1989). This has the potential to skew our findings for the practice location data. The second potential limitation of this study was the use of descriptive statistics to compare CMGs and IMGs. The third potential limitation is the possible nongeneralizability of the study findings (but not necessarily the study method) to countries other than Canada, because many countries have different IMG criteria and recruit physicians from a variety of backgrounds. Nevertheless, this study is one of the first to use the CAPER database, one of the largest available databases on CMGs and IMGs in the world.
Future studies could examine the growing subgroup of Canadian citizens who are currently training abroad over time and whether they differ from IMGs in terms of specialty choice and location of practice, as an estimated 3,500 Canadian citizens are currently studying medicine abroad according to a recent report.17 Furthermore, the distribution of physicians by origin and discipline in underserviced areas could be examined in Canada and other countries to help guide physician recruitment and retention both in Canada and other countries around the world suffering from physician shortages.
This study demonstrates that IMGs tend to be older, more likely to be men, and more likely to pursue a career in family medicine than their CMG counterparts. Because IMGs tend to be older, they will have fewer years of productivity before retirement than will CMGs. As IMGs continue to complete postgraduate training in Canada and as the new cohort of CMGs graduate from expanded medical school positions, it will be interesting to see how quickly the physician shortage in Canada lessens and whether recent policy changes for IMG recruitment and CMG expansion could potentially overshoot the target. Therefore, it is crucial to continue to closely monitor CMG and IMG characteristics and trends to ensure that current recruitment and retention strategies are working appropriately to address the physician shortage and strike an appropriate balance.
The results of this study have helped create a generalized profile of CMGs and IMGs who have completed postgraduate training in Canada. Although the results are specific to Canada, the data could provide a template for other countries to look at their own data on the basis of the methods presented in the current study. Furthermore, readers could draw on the Canadian data presented in this study to inform policy makers in planning their own health care environments at the local, national, and territorial government levels.
The authors would like to thank Leslie Forward from the Canadian Post-M.D. Education Registry (CAPER) and Shawn Healy from the University of Toronto Postgraduate Medical Education Office for their help in compiling and analyzing the data.
The authors did not seek ethics approval because they did not use identifiable data and did not observe or survey human participants.
The abstract of an earlier version of this report was presented at the Family Medicine Forum 2010 held in Vancouver, British Columbia, October 14–16, 2010.