The United States is facing a continuing shortage of primary care physicians.1,2 Responses to the Graduation Questionnaire of the Association of American Medical Colleges by physicians from 1997 to 2006 show there was an overall decrease in those who chose general internal medicine from 15.7% to 6.7%, from 10.2% to 6.6% in general pediatrics, and from 17.6% to 6.9% in family practice.3 On the other hand, there was an overall increase in the proportion of physicians who chose internal medicine subspecialties (6.8% to 11.4%) and pediatrics subspecialties (2.2% to 4.4%).3 Recent trends reveal a major drop in graduating medical students specializing in general internal medicine over the past 15 years. In 1998, approximately 55% of residents specialized in general internal medicine, whereas in 2005, this proportion dropped to 20%.4,5
This ever-growing shortage of primary care physicians is magnified in rural communities. In 2005, there were 55 primary care physicians for every 100,000 people in rural areas compared with an estimated need for 95 per 100,000.6,7 Whereas approximately 19% of the U.S. population lives in rural America, only about 11% of physicians practice in rural locations.8 The shortage of rural family practitioners can be attributed to various factors, some of which include an aging rural population, a retiring medical workforce, and fewer medical students interested in practicing in rural areas.9
Research conducted by the Health Resources and Services Administration Rural Health Research Centers found that effective recruitment strategies for primary care include focusing on students with rural backgrounds, exposing students to rural areas and issues during medical school, and offering financial incentives to practice and remain in rural areas. It also concluded that older and nontraditional medical students are more likely to practice in rural areas.10 Obstacles that discourage practicing as a rural primary care provider include lower salaries than in urban areas, cultural isolation in rural areas, lower-quality schools and housing options than more metropolitan areas, and a lack of spousal job opportunities.5,11
Rabinowitz and colleagues12 have systematically reviewed the outcomes of comprehensive medical school programs designed to increase the rural physician supply, and developed a model to estimate the impact of their widespread replication. The investigators found six studies that summarized the rural outcomes of medical school programs addressing rural physician supply: the Rural Physician Associate Program (RPAP) of the University of Minnesota; the University of Minnesota Medical School, Duluth; the Upper Peninsula Program of Michigan State University; the Physician Shortage Area Program (PSAP) of Jefferson Medical College, Thomas Jefferson University; the Rural Medical Education (RMED) Program, State University of New York; and the RMED Program of the University of Illinois College of Medicine (COM) at Rockford. All of these comprehensive medical school rural programs have produced a multifold increase in the rural physician supply, and the authors concluded that widespread replication of these models could improve access to health care in thousands of rural communities.
The University of Illinois RMED Program is part of the National Center for Rural Health Professions, located on the Rockford campus of the University of Illinois. The RMED mission is “a medical education program designed to recruit students from rural areas, who will upon completion of residency training, return to rural Illinois as primary care physicians.” Candidates considering the RMED Program participate in a dual application process, applying to both the University of Illinois COM and the RMED Program. Potential students must follow the American Medical College Application Service requirements as well as complete an RMED Program application.13 Rural track students take the regular COM curriculum as well as the supplemental, required RMED curriculum (for details, see Glasser and colleagues14). The RMED curriculum occurs during all four years of medical school in addition to the course work required by the COM. From the first to the fourth year of medical school, the themes in curriculum progress: community and population health; health care delivery in rural communities; understanding and developing a community-oriented primary care (COPC) project; and immersion in a rural community and primary care practice. Monthly RMED sessions or activities occur during the first three years of medical school. Sessions are led by a faculty member or community member with expertise on the particular topic of the session. After each session, students submit a written synopsis of the session in which they respond to specific questions regarding interpretation and application of the material presented. These written assignments are done individually in the first year and as group assignments in the second year. In the first and second years, field trips and special events (e.g., “Southern Exposure,” the “No Harm on the Farm” tour) provide students with the opportunity to directly observe and interact with rural health care providers and rural communities. At the end of the third year, each RMED student will have completed his or her institutional review board (IRB) certification, developed a COPC project abstract, and visited the preceptorship site community.
In the fourth year, RMED students spend 16 weeks precepting with a rural primary care physician located in Illinois. Primary care physician preceptors are screened and credentialed as preceptors and are responsible for evaluating the RMED student’s clinical activities. During the 16-week preceptorship, students devote approximately 70% of their time to clinical aspects of rural primary care in both the office and hospital settings. Students are also encouraged to spend, on average, one half-day per week working with other health care providers in the community. Students are responsible for logging details for at least 250 patient encounters during the 16-week preceptorship block. During the 16 weeks, students also complete a COPC project. RMED faculty in Rockford guide the implementation of the COPC project. The final project is presented as a poster during Research Day at the Rockford campus.
There are approximately 50 students per medical class assigned to the COM Rockford track; of these, 13 to 20 students per year participate in the RMED Program’s rural track. Within this context, we examine and report on the retention/practice outcomes related to rural location and primary care for the RMED Program of the University of Illinois COM at Rockford in comparison with non-RMED COM graduates between 1997 and 2007.
Our study was approved by the University of Illinois–Rockford IRB. We used data related to the characteristics of RMED and non-RMED students who graduated from the University of Illinois between 1997 and 2007 from a variety of sources. We did not include COM graduates from 2008 and later, to increase the likelihood that residency training would be completed and the graduate would be in active practice. Basic information on RMED graduates regarding practice location(s) since graduation and practice specialty is tracked by the RMED Program office in Rockford by a staff member who updates information on all RMED students and graduates. The size of the current town/city population where these RMED graduates are practicing is based on the 2000 U.S. Census. We used this information to describe RMED Program graduate outcomes for 11 years (from the classes of 1997 through 2007) related to medical specialty, practice location, and length of time in practice. Finally, we examined RMED graduates’ current practice location distance relative to the site of their 16-week, fourth-year preceptorship, their hometown identified at medical school application, and their residency location. We determined the distance from current practice location zip code to these other locations by using a Web-based distance calculator.15
Additionally, related to program evaluation, we compared RMED graduate specialty choice and practice location outcomes with all other University of Illinois COM graduates over the 11-year period. The University of Illinois is the largest medical school in the country, with about 300 students admitted each year. We obtained a list of the names of all graduates between 1997 and 2007 from the university registrar. In January to March 2012, using data from the national provider identification (NPI) lookup site,16 we identified the practice locations and practice specialties of 2,823 graduates who were in practice as of March 1, 2012 (see Figure 1). The study population is 92.1% of the 3,064 graduates from the classes of 1997–2007. The NPI is a U.S. government-maintained database to which each physician and hospital is required to report. After locating the physician, we used additional Web sites providing information about physicians in practice to verify current practice specialty and location.
In terms of practice location, we present size of town or city of current practice. Although this information may be of interest for possible comparative purposes, it can be deceiving. It is possible that graduates could be practicing in communities of 10,000 people, but within just a few miles of a major city. We therefore conducted an analysis using the most recently available zip code levels set by the Rural–Urban Commuting Area code 2.0 (RUCA),17 which are designed to define the level of rurality on the basis of census information. The RUCA codes range from 1 to 10 with subcategories. We considered a zip code to be rural if the RUCA code was 4 or higher and nonrural if the RUCA code was less than 4.17 The zip code of the current practice location for both RMED and non-RMED graduates was linked to the RUCA code of the zip code for each graduate with a known practice location. Finally, we determined the primary care shortage status of the practice zip code for each 1997–2007 graduate based on eligibility for Centers for Medicare and Medicaid Services (CMS) primary care Medicare bonus payment.18 We used cross-tabulation and chi-square analyses to compare the practice and specialty characteristics of RMED versus non-RMED students. Finally, we calculated odds ratios related to location and specialty choice. We analyzed data using SPSS version 19 (SPSS Inc., Chicago, Illinois).
At the time of our research, the University of Illinois RMED Program has had 211 graduates. From the 11-year period between 1997 and 2007, 160 are now in practice, 93 (58.1%) of whom are males. A total of 105 graduates (65.6%) are in the specialty of family medicine. Eleven (6.8%) are general internists or practice internal medicine/pediatrics, and 6 (3.7%) are general pediatricians. Overall, 122 graduates (76.3%) are in primary care practice; this number increases to 127 (79.4%) if obstetrics–gynecology is included in the definition of primary care. In terms of practice location, the majority of RMED graduates, or 116 (72.5%), have remained in Illinois. Among non-RMED graduates, 41.4% (1,102) are practicing in Illinois after completion of residency.
For the 160 RMED graduates in practice, 98 (61.3%) practice in towns of 20,000 people or less, with 37 (23%) practicing in towns of 5,000 or less. Using the RUCA code cutoff of 4 or higher, 90 (56.3%) of the 1997–2007 RMED graduates are currently practicing in rural areas. In terms of length of time in practice, the mean number of years is 5.3 years, with a range from 0.5 to 11.3 years. A total of 110 (68.8%) have remained in their original practice location. For these, the mean length of time in the community is 4.3 years, with a range of 0.58 to 10.6 years.
We present findings related to practice characteristics of RMED graduates compared with all other University of Illinois COM graduates in Table 1. Findings related to graduates in nonrural locations are provided for comparison; however, our focus was to assess the outcomes of the RMED Program related to practice location and specialty. Data for non-RMED COM graduates provide a comparison suggesting what practice location and specialty would be among graduates without the RMED Program being offered at the University of Illinois. The outcomes for rural practice location differ markedly. Among RMED graduates in rural areas, 75.6% (68) are practicing family medicine in comparison with 37.8% (70) of rural non-RMED graduates (P < .001). The percentage of graduates practicing family medicine in a CMS-designated shortage zip code differs between RMED and non-RMED graduates: 37.8% (34) versus 13.5% (25), respectively. Choice of a primary care specialty (family medicine, internal medicine, or pediatrics) was substantially higher (P < .001) among RMED graduates (83.3%, or 75) than among non-RMED graduates (52.4%, or 97). Correspondingly, the percentage in non-primary-care specialties is higher among rural non-RMED graduates (47.6%, or 88) than rural RMED graduates (16.7%, or 15). Regardless of specialty, practice location in a CMS-designated primary care zip code was significantly higher (P < .001) among rural RMED graduates (43.3%, or 39) in comparison with rural non-RMED graduates (27.6%, or 51). Among graduates in nonrural locations, a higher percentage practicing in a CMS shortage area was also observed among RMED graduates (14.3%, or 10) than non-RMED graduates (1.5%, or 38). In general, RMED graduates were much more likely to practice a primary care specialty in a rural location (especially family medicine) and/or practice in a CMS shortage zip code than non-RMED graduates. Figure 2 indicates substantial consistency with regard to rural or nonrural location even if the practice setting/employer changes, with 52.5% (84) of the RMED graduates always practicing in a rural location, 10.6% (17) practicing at some time since graduation in a rural location, and 36.9% (59) always practicing in a nonrural location.
Practice characteristics for all locations are compared in Table 2. In terms of specialty choice, odds ratios indicate that RMED graduates were 14.38 times (CI = 10.16–20.35) more likely to choose family medicine than non-RMED graduates (65.6%  versus 11.7% , respectively). RMED graduates had about a 6.7 fold (CI = 4.60–9.71) higher likelihood of choosing a primary care practice specialty than non-RMEDs (76.3%  versus 32.4% ). Among RMED graduates, the likelihood of current practice being rural was about 17 times (CI = 12.18–24.35) higher than for non-RMED graduates (56.3%  versus 6.9% , respectively). The likelihood of current practice being in a CMS designated primary care shortage zip code was 12.77 times (CI = 8.58–18.99) higher among RMED than non-RMED graduates.
Examination of the current practice location in relation to distance from RMED graduates’ fourth-year 16-week preceptorship site and hometown (Figure 3) indicates that of the 160 RMED practicing graduates, 21.9% (35) were within 30 miles of their preceptorship site, 41.9% (67) were within 90 miles of their fourth-year medical school preceptorship site, 39.4% (63) were within 90 miles of their residency location, and 58.1% (93) were within 90 miles of their hometown. Separate analyses indicate that among the 116 RMED graduates practicing in Illinois, 62.1% (72) were within 60 miles of their hometown and 73.3% (85) were within 90 miles of their hometown.
Our findings clearly indicate that the University of Illinois RMED Program is succeeding in regard to both rural practice location and choice of a primary care specialty, as established in the program’s mission. National data from 1988 to 1997 indicate that 11% of clinically active physicians practice in rural locations. Yet, over 60% of RMED graduates practice in towns of 20,000 people or fewer, and 56% are in communities with a RUCA code of 4+, indicating a rural area.
Our findings compare favorably to other rural programs reviewed by Rabinowitz and colleagues12 as comprehensive rural programs. Outcomes from the RPAP at the University of Minnesota Medical School demonstrate that 82% of graduates have entered primary care and that over 50% are in or have practiced in a rural setting.19 Similarly, Jefferson Medical College PSAP graduates were more likely than their non-PSAP peers to practice rural family medicine: 32.0% versus 3.2%.20 For graduates of the Rockford RMED Program, 76% are in primary care practice and 56% to 61% are practicing in rural communities, depending on criteria for rurality. RMED graduates, compared with non-RMED graduates, are more likely to practice rural medicine, 56.3% versus 6.9%. RMED graduates, in addition to being 17.2 times more likely to be practicing in a rural location than non-RMED graduates, are also markedly more likely to be practicing in Illinois (72.5% and 41.4%, respectively).
A fundamental premise of the health reform legislation of 2010, the Patient Protection and Affordable Care Act, is that increased access to quality primary care services is essential to improving the health care system in the United States.21,22 Our findings indicate that an effective method for increasing the proportion of physicians who choose primary care is to select students who are interested in and committed to primary care practice and provide them a structured education experience that encourages the choice of primary care.
Past literature clearly indicates a special need related to shortages of primary care physicians in rural areas.5,7,8,23–25 Results of our analysis indicate that a program focused on providing rural primary care physicians is able to produce practice outcomes that can help resolve workforce shortages. These positive outcomes, especially related to rural primary care and practice in shortage areas, do not occur in the absence of a program targeting the education of rural primary care practitioners. Although a considerable proportion of non-RMED graduates do choose primary care (32.4%, 864/2,663), overall, RMED graduates are about 17 times more likely to practice in a rural location, 6.7 times more likely to practice primary care, and 12.8 times more likely to practice in underserved areas designated as primary care shortage areas. The likelihood of primary care practice is substantially increased among rural RMED graduates in comparison with rural non-RMED graduates (4.7 times higher among RMED graduates).
Our study presents evidence of the value of a medical workforce program based on a “grow your own” approach as demonstrated by the finding that 58.1% (93/160) of RMED graduates are within 90 miles of their hometown and 41.9% (67/160) are within 90 miles of their fourth-year 16-week preceptorship location. Among RMED graduates, the proximity of practice location to residency location was lower, with only 39.4% (63/160) being within 90 miles of their residency location. Similar patterns of choosing a practice location close to site of residency training have also been discussed related to increasing the supply of primary care physicians in an area. A national survey examining graduate medical education and physician practice location demonstrated that 51% of physicians are practicing in the state in which they obtained their graduate medical education. Generalist physicians are more likely to stay in their state of graduation compared with specialist physicians.26 Particularly with regard to rural training track residency programs, this trend has been used to facilitate an increased supply of physicians in rural areas. The University of Missouri Rural Track Pipeline Program looked at graduates from 1997 through 2006 and showed that 57% of these physicians practiced in a rural area. Of these students, 42.6% chose to practice in Missouri, whereas 14.8% practiced outside of Missouri.27
We would suggest several possible reasons for the pattern of return to practice relatively close to the RMED hometown and/or fourth-year 16-week preceptorship site on the basis of our years of experience training rural medical students. First, some RMED students grew up in 1 of the 25 RMED preceptorship towns and have the intent of returning to practice in their hometown or the nearby area. An extended time period back in their hometown working as a health professional may further strengthen their intent to practice in their hometown because relationships are established that endure throughout residency and lead to practice options being offered at the end of residency training. Second, some students like a particular area of the state because of the climate, surroundings, or rural culture as well as proximity to significant others or family, leading them to choose a preceptorship site in the area, and ultimately a practice location in the same region. The importance of “familiarity factors” is supported in findings reported by Hancock and colleagues.28 Third, some students are seeking loan repayments or other financial incentives during medical school or residency, and one of the first hospitals they may contact is a hospital near their hometown or perhaps the site of their fourth-year RMED preceptorship that resulted in obligations to return to the area. Fourth, some students may not have intended to return to the preceptorship location, but find they are very compatible with the physicians in the practice, which leads to development of an option to work in the practice after completion of residency. The 16-week experience allows adequate time for both preceptor and student to assess whether there is mutual interest in exploring practice opportunities in that location after residency.
One limitation of this study is that we focus on one program in one state. The data are only for an 11-year time span. It is possible that the RMED Program has not increased the percentage of University of Illinois medical students practicing primary care in rural locations from prior time periods. The difference between the percentages of RMED and non-RMED graduates practicing in rural locations does, however, suggest a discernible impact during the 11-year study period. Practice location of graduates prior to 1997 (years before the first RMED graduates) was not assessed by this study.
Review of characteristics of Rockford campus graduates’ characteristics preceding this study, from 1986 to 1996, indicated that 13.2% were from a rural hometown with a RUCA code of 4 or higher. Yet, during the time period of our study (graduates from 1997 to 2007), 32.2% of Rockford campus graduates were from a rural hometown, representing about a 20% increase in the percentage of rural background medical students compared with the prior 11 years. Additional review of campus administrative data comparing the two 11-year time periods indicated that 47.1% of 1986–1996 graduates compared with 57.3% of 1997–2007 graduates chose a primary care residency, which supports the impact of the RMED Program on specialty choice outcomes. The 10% increase in primary care residency choice on the campus occurred with RMED students composing 31% of the graduates from 1997 to 2007. Given the major shortages in rural health care availability, it is important to examine and seek to understand the outcomes of focused medical education programs like RMED. It is important to document that rural, community-based programs can result in highly favorable outcomes.
In this regard, our study contributes to findings about longer-term retention of physicians in rural areas. Although programs such as the National Health Service Corps are instrumental in helping meet needs in rural and underserved areas, the long-term effects may not be great.29,30 Although there is no “gold standard” for comparison, we are encouraged that 68% of RMED graduates are still in their original practice location, that 63.1% have practiced in rural areas, and that the mean number of years in a community is over 5—and is likely to get higher as the program “ages.” Primary care in rural and underserved communities is an important goal. Continuity of care in these communities is even more essential for effective health care delivery and healthy rural communities.
Acknowledgments: The authors appreciate the assistance of National Center for Rural Health Professions staff Vicki Weidenbacher-Hoper, MSW, assistant director; and Dana Evans, MS, health education specialist, in reviewing final revisions of the article and responses to reviewers. Assistance of Richard Mulnix, MS, in the campus research office related to the literature review, is also appreciated.
1. Whitcomb ME, Cohen JJ. The future of primary care medicine. N Engl J Med. 2004;351:710–712
3. Jeffe DB, Whelan AJ, Andriole DA. Primary care specialty choices of United States medical graduates, 1997–2006. Acad Med. 2010;85:947–958
6. Steinwald AB. Primary Care Professionals: Recent Supply Trends, Projections, and Valuation of Services. Testimony Before the Committee on Health, Education, Labor, and Pensions, U.S. Senate. February 12, 2008 Washington, DC U.S. Government Accountability Office http://www.gao.gov/new.items/d08472t.pdf
. Accessed August 9, 2013
7. Fordyce MA. Physician Supply and Distribution in Rural Areas of the United States 2005. Final Report #116. 2007 Seattle, Wash WWAMI Rural Health Research Center, University of Washington http://www.ruralhealthresearch.org/projects/144/
. Accessed August 9, 2013
11. 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
12. Rabinowitz HK, Diamond JJ, Markham FW, Wortman JR. Medical school programs to increase the rural physician supply: A systematic review and projected impact of widespread replication. Acad Med. 2008;83:235–243
13. Glasser M, Stearns MA, Stearns JA, Londo RA. Screening applicants for a rural medical education program. Acad Med. 2000;75:773
14. Glasser M, Hunsaker M, Sweet K, MacDowell M, Meurer M. A comprehensive medical education program response to rural primary care needs. Acad Med. 2008;83:952–961
19. Halaas GW, Zink T, Finstad D, Bolin K, Center B. Recruitment and retention of rural physicians: Outcomes from the rural physician associate program of Minnesota. J Rural Health. 2008;24:345–352
20. Rabinowitz HK, Diamond JJ, Markham FW, Santana AJ. Increasing the supply of rural family physicians: Recent outcomes from Jefferson Medical College’s Physician Shortage Area Program (PSAP). Acad Med. 2011;86:264–269
21. Rittenhouse DR, Shortell SM, Fisher ES. Primary care and accountable care—two essential elements of delivery-system reform. N Engl J Med. 2009;361:2301–2303
22. Abrams M, Nuzum R, Mika S, Lawlor G. Realizing health reform’s potential: How the Affordable Care Act will strengthen primary care and benefit patients, providers, and payers. 2011 Commonwealth Fund Web site. January
23. Rosenblatt RA. Commentary: Do medical schools have a responsibility to train physicians to meet the needs of the public? The case of persistent rural physician shortages. Acad Med. 2010;85:572–574
26. Seifer SD, Vranizan K, Grumbach K. Graduate medical education and physician practice location. Implications for physician workforce policy. JAMA. 1995;274:685–691
27. Quinn KJ, Kane KY, Stevermer JJ, et al. Influencing residency choice and practice location through a longitudinal rural pipeline program. Acad Med. 2011;86:1397–1406
28. Hancock C, Steinbach A, Nesbitt TS, Adler SR, Auerswald CL. Why doctors choose small towns: A developmental model of rural physician recruitment and retention. Soc Sci Med. 2009;69:1368–1376
29. Pathman DE, Konrad TR, Ricketts TC 3rd. The comparative retention of National Health Service Corps and other rural physicians. Results of a 9-year follow-up study. JAMA. 1992;268:1552–1558
© 2013 by the Association of American Medical Colleges
30. Bärnighausen T, Bloom DE. Financial incentives for return of service in underserved areas: A systematic review. BMC Health Serv Res. 2009;9:86