The shortage of physicians in rural areas represents one of the most serious and enduring problems in the U.S. health care system.1–3 This shortage is especially significant because the one in five Americans who live in rural areas have greater medical need, being sicker (e.g., more likely to have diabetes, more likely to die from cardiovascular disease) and older than those living in nonrural areas.1–4
This rural physician shortage has existed for most of the past century and is especially critical regarding primary care physicians: Rural areas have less than 60% of the per capita supply of generalists compared with urban areas.5–7 This disparity between rural and urban physician supply has not changed in recent years,7 and current trends suggest that it will only worsen: Only 9% of current physicians practice in rural areas; rural physicians are more likely to retire over the next decade than their nonrural peers; and fewer than 3% of recent medical students plan to practice in rural areas or small towns.8,9 In addition, rural areas will disproportionately bear the impact of the decreasing numbers of primary care physicians, especially of family physicians, who represent the largest source of rural health care providers.10,11 Future expansion of health insurance to a larger number of people will also likely exacerbate the current rural physician shortage, as recently occurred in Massachusetts.12,13
Over the past few decades, a number of programs have been implemented in order to address the rural physician shortage.14,15 Many of these, including the National Health Service Corps and loan repayment programs, have had some degree of short-term success.16,17 In addition, community health centers and rural training tracks in family medicine residency programs have had an important impact on access to care in rural areas, although both of these have also had significant problems with recruiting adequate numbers of family physicians.18,19 Many have proposed using more nurse practitioners (NPs) and physician assistants (PAs) as a solution to the rural provider shortage; however, the scope of the problem is greater than that which can be entirely solved by increasing the number of these nonphysician clinicians even though NPs and PAs serve a very important role in rural areas.3,20
The most successful outcomes regarding the rural physician shortage have been comprehensive medical school rural programs (i.e., programs with both  a defined cohort of students and  either a focused rural admissions process and rural curriculum or an extended rural clinical curriculum). These programs have had a successful track record, and despite different definitions of rural, approximately one-half of their combined graduates have practiced in a rural area. These programs, including those based in both public and private institutions and those located in different regions of the United States, have graduated more than 1,600 physicians over three decades.21
The Physician Shortage Area Program (PSAP) of Jefferson Medical College (JMC) of Thomas Jefferson University is one of these successful medical school programs that has increased the supply and retention of rural family physicians.22,23 The PSAP was initiated in 1974 to address the shortage of physicians in rural Pennsylvania. As previously described,22–25 the PSAP recruits and selectively admits medical school applicants who have grown up in a rural area and who are committed to practicing family medicine in a rural area, especially in Pennsylvania. During medical school, PSAP students have family physician faculty advisors, take their required third-year family medicine clerkship in a rural location, take their senior outpatient subinternship in family medicine (usually at a rural preceptorship), and receive a small amount of additional financial aid (almost entirely in the form of repayable loans). On completion of medical school, PSAP graduates are expected to complete a family medicine residency and practice family medicine in a rural location of their choosing, although no formal mechanism exists to ensure compliance.
Because of the existence of the Jefferson Longitudinal Study of Medical Education (JLS), which is a unique longitudinal tracking database of JMC graduates beginning with the graduating classes of 1968,26 the PSAP is the most studied of the comprehensive medical school rural programs. Outcome studies have shown that the 1978-1991 PSAP graduates were more than 8 times as likely to practice rural family medicine than their non-PSAP JMC peers, and 19 times as likely to practice family medicine in rural Pennsylvania than all other non-PSAP physicians who graduated from all Pennsylvania medical schools.22–24 The PSAP is also the only medical school rural program with published long-term retention outcomes: The 11- to 16-year retention rate for family physicians in the same rural area is 68%—and another 11% are practicing family medicine in a different rural area.27
Despite the prior published success of the PSAP and other medical school rural programs in producing rural physicians, it is important to determine whether these programs have continued to be successful in more recent years, during which major changes have taken place in the U.S. health care system. Because the most recent published outcomes for PSAP graduates were from the class of 1991,22 we focused this study on the geographic and specialty outcomes of more recent PSAP graduates (1992-2002). Further, because the statewide distribution of physicians is also of key importance for access to care, we determined the distribution of these more recent PSAP graduates in rural Pennsylvania.
Building on our ongoing longitudinal studies of PSAP graduates,22–24,27 we obtained from the JLS database the current (2007) demographic and specialty information (including board certification data acquired from the American Board of Medical Specialties and self-reported specialty from the American Medical Association [AMA] Physician Masterfile) for all JMC living graduates from the classes of 1992–2002. We merged these data with 2007 addresses and specialty data from the alumni files of the Jefferson Foundation, which is updated every three months and has previously been shown to be highly accurate.23 As in our prior studies,22–24 we used office address (i.e., county), and when these were unavailable, we used home address (also county), assuming home and work county were of similar rurality.
Similar to prior studies, we defined current 2007 specialty as that in which the graduate was board certified. For those also certified in geriatrics, sports medicine, or adolescent medicine, we used their primary certified specialty. In all other instances, if the physician was certified in two or more specialties or not certified at all, then we used the primary self-reported specialty from the AMA Physician Masterfile. When that was unavailable, we used JMC alumni self-reported specialty.
The federal criteria for defining a county as rural that we used in our prior studies changed substantially in 2003, so we were unable to use the same definition of rural practice location for this study. Therefore, we considered graduates to be practicing in a rural county if their 2007 county was designated as rural on the basis of its Rural-Urban Density Typology (RUDT)28 (we considered graduates in all other counties to be practicing in an urban county). The RUDT classification identifies rural and urban counties on the basis of (1) the percent of the population that is rural or urban, (2) the population in urbanized areas, and (3) population density. We used this classification because it most accurately met the goals of our study29 (i.e., to determine the rural location of family physicians). Because no universally accepted definition of “rural” at the county level exists, we also used an alternate definition, which we had previously used.23 According to this alternate definition, rural counties are those where less than 50% of the population live in an urbanized area (i.e., a densely settled territory that contains 50,000 or more people).23
We calculated relative risks (RRs) and 95% confidence intervals (CIs) comparing outcomes for PSAP versus non-PSAP graduates. In all analyses, we considered two-sided P values of less than .05 to be of statistical significance. We used SPSS version 16 (SPSS Inc., Chicago, Illinois) to perform our statistical analysis. The institutional review board of Thomas Jefferson University approved this study.
Of 2,394 JMC graduates from the classes of 1992–2002, 104 (4.3%) were PSAP graduates. Of these, all 104 (100%) were alive in 2007, as were 2,281 (99.6%) of the 2,290 non-PSAP graduates. Practice location in the United States was available for 93.3% of living PSAP graduates (97/104) and for 87.9% of living non-PSAP graduates (2,004/2,281), and practice specialty was available for all graduates.
Overall, 43.3% of PSAP graduates (42/97) were currently practicing in rural areas of the United States compared with 15.8% of non-PSAP graduates (316/2,004) (RR = 2.7, CI 2.1−3.5, P < .001). Similarly, 61.5% of PSAP graduates (64/104) were practicing the specialty of family medicine compared with 13.1% of their non-PSAP peers (299/2,281) (RR = 4.7, CI 3.9−5.6, P < .001). PSAP graduates were almost 10 times more likely to combine the specialty of family medicine with practice in a rural area (32.0% [31/97]) than were their non-PSAP peers (3.2% [65/2,004]; RR = 9.9, CI 6.8–14.4, P < .001).
When we applied the secondary definition of rural (i.e., counties where less than 50% of the population live in an urbanized area), the pattern of results were similar; that is, the RRs of PSAP versus non-PSAP graduates were similar, although the absolute percentages of physicians practicing in rural areas were comparably lower for both groups (e.g., 21.6% of PSAP graduates [21/97] and 2.1% [43/2,004] of non-PSAP graduates were rural family physicians [RR = 10.1, CI 6.2–16.3, P < .001]).
Regarding the statewide impact of these 1992–2002 graduates, PSAP graduates were more likely to be practicing in Pennsylvania (rural or urban, any specialty) than were non-PSAP graduates (PSAP 55.7% [54/97] versus non-PSAP 29.2% [586/2,004]; RR = 1.9, CI 1.6–2.3, P < .001). PSAP graduates were also preferentially practicing in rural Pennsylvania counties (any specialty) (PSAP 24.7% [24/97] versus non-PSAP 2.0% [40/2,004]; RR = 12.4, CI 7.8–19.7, P < .001). In fact, more than 1% of PSAP graduates were practicing in 18 of the state's 48 rural counties, whereas no rural county had even 0.3% of non-PSAP graduates (Figure 1).
Discussion and Conclusions
The results of this study show that more recent PSAP graduates (1992–2002) were almost 10 times more likely to be currently practicing rural family medicine than their non-PSAP JMC peers. PSAP graduates were also distributed much more widely than their peers among the rural Pennsylvania counties. The implications of these results are important to consider as the United States struggles with how to increase access to care for all Americans. The fact that only 40 of more than 2,000 non-PSAP graduates from 11 JMC classes practice in any of the 48 rural Pennsylvania counties highlights the serious overall shortage of rural physicians. Although providing health insurance to an increasing proportion of the population will likely decrease the major barrier to health care for those living in urban areas, the geographic availability of a health care provider will remain necessary for access to health care for those in rural areas.
Although this study did not specifically address the reasons for the success of the PSAP, prior research has shown that the admissions process (i.e., admitting students who both grow up in rural areas and plan to practice family medicine in rural communities) is by far the most important factor in rural physician outcomes, as well as in the success of the PSAP.11,17,30,31 Former studies have also shown that although PSAP students have similar academic admissions credentials as their peers and perform similarly in medical school and residency, most would not have been accepted to medical school without the program—a finding that reinforces the positive impact of the program on the rural workforce.24
The PSAP represents one of six comprehensive medical school rural programs that have published outcomes showing success in increasing the supply of rural physicians.21 The other five such programs are (1) the Rural Physician Associate Program (RPAP) at the University of Minnesota, (2) the University of Minnesota-Duluth, (3) the Rural Physician Program at Michigan State University College of Human Medicine, (4) the Rural Medical Education Program at the State University of New York Upstate Medical University (RMED SUNY), and (5) the Rural Medical Education Program at the University of Illinois College of Medicine at Rockford (RMED Rockford). Although each of these programs has proven successful, much of the prior published literature focused on graduates from the 1970s and 1980s, and less information regarding the impact of these programs on more recent graduates is available. Only two reports from RMED SUNY and RPAP,32,33 and the Internet site from RMED Rockford,34 focused specifically on graduates in more recent years. The results of the present study, combined with the relatively more recent successful outcomes of these other three programs, suggest that medical school rural programs continue to have a substantial impact on addressing the rural physician shortage.
Because the designation of rural that we had used in our prior studies (i.e., nonmetropolitan as determined by the Office of Management and Budget [OMB]) underwent major definitional changes in 2003, we were unable to use the same definition of rural practice location for the current study. The 2003 OMB nonmetropolitan definition no longer included any criteria of rurality itself (e.g., population density) and thus expanded its metropolitan classification to include more outlying counties that are economically linked to regional urban centers, but which are not in themselves urban.28 This new OMB definition now places the majority (51%) of rural people (defined by the Census Bureau as those who live outside of an urban area of 2,500 or more people) as being in metropolitan counties. According to this new OMB definition, even 40% of the farm population are living in metropolitan areas.35
Therefore, for this current study, we used the RUDT as a measure of rurality at the county level, more accurately identifying 85% of the rural population.28,36 To ensure that by using the RUDT we did not inadvertently miss important changes in the actual nature of rural communities over time, we determined that the Pennsylvania counties (where the majority of PSAP graduates practice) defined as rural by the RUDT (Figure 1) perfectly matched the 48 counties identified as rural by the Center for Rural Pennsylvania (a legislative state agency).37 In five of these counties defined as rural by the RUDT but classified as metropolitan by the 2003 OMB designation, the largest community comprised fewer than 6,000 people (i.e., Kittanning, population 4,700; Lehighton, 5,500; Marysville, 2,300; Matamoras, 2,300; and Tunkhannock, 1,900).
To ensure that using the RUDT did not affect the outcomes of this study, we also included a previously used alternate definition23 (i.e., counties where less than 50% of the population live in an urbanized area), and the primary outcome, the RR of PSAP versus non-PSAP graduates practicing rural family medicine, was similar (9.9 versus 10.1). Even using the 2003 OMB nonmetropolitan designation would have resulted in a similar RR of PSAP versus non-PSAP graduates who were rural family physicians (9.6). Although this study focused on family physicians practicing in any rural area, most of the PSAP graduates practicing rural family medicine (77%) were also located in counties federally designated—either wholly or partially—as primary care Health Professional Shortage Areas.
Because the major changes in the OMB definition in 2003 made it impossible for us to use the same definition of rural that we had used in our prior studies,22–24 we were unable to compare the absolute rural family medicine outcomes of our more recent 1992–2002 graduates (32% PSAP versus 3.2% non-PSAP) with those of earlier graduates from 1978 to 1991 (21% PSAP versus 2% non-PSAP).22 Nevertheless, the relative outcomes in this study (RR = 9.9) suggest that the PSAP continues to be similarly successful in addressing the rural physician supply in more recent years as it was previously (RR = 8.5).
A limitation of this study is that it represents results from a single medical school program, although a number of factors do provide support for its generalizability. First, previous research has shown the PSAP to be a program with outcomes comparable with those of other medical school rural programs.21 In addition, this study involves 11 classes of more than 2,100 graduates who have taken their residency training in over 343 hospitals in 44 states and who practice in 47 states. Finally, because JMC is a large, private medical school in the northeastern United States—characteristics related to lower outcomes of primary care and rural physicians38—it is likely that the PSAP could have an even greater impact on the rural workforce in other medical schools and in most other states.
As prior workforce modeling has suggested, developing or expanding comprehensive medical school rural programs (similar to the PSAP or other programs) in all U.S. medical schools could double the number of rural physicians, resulting in thousands of additional rural providers over the next decade.21,39 These programs are also among the least costly that have proven to be successful in addressing the rural physician shortage. However, given the past history of medical schools in responding to the rural physician shortage as well as the current serious financial pressures on medical schools, most schools are unlikely to decide on their own to develop these programs. Providing incentives to medical schools presents one way to encourage change, and the National Institutes of Health (NIH) Medical Scientist Training Program (MSTP) represents such an instructive model. The NIH instituted this well-funded program to address another important physician workforce need, the shortage of physician researchers. The MSTP currently provides substantial grants to more than 40 medical schools to support MD-PhD candidates, and a number of other medical schools have developed similar programs without NIH support. Similar to the PSAP, the MSTP selects students on the basis of their past experience and commitment to future careers, and provides appropriate curricula, mentorship, and support to help them achieve their goals. In existence since 1964, the outcomes of the MSTP have had a level of success similar to those of comprehensive medical school rural programs.40 The new federal program—the Rural Physician Training Grants program, which is part of the recent Affordable Care Act—represents a comparable opportunity to provide financial support for developing and expanding medical school rural programs similar to the PSAP.41 This new federal program, along with the current expansion in medical school class size (the first in decades), represents an unprecedented opportunity for the widespread replication of medical school rural programs that could have a major impact on thousands of rural communities.21,39,42
In summary, this study has shown that the PSAP continues to represent a successful model to increase the supply and distribution of rural family physicians. Recent PSAP graduates (1992–2002) are much more likely to practice rural family medicine than their non-PSAP peers and are much more widely distributed among the rural counties of Pennsylvania. Even as the United States expands the number of people who have health insurance, access to health care for rural residents will require an increased supply of rural providers, so these outcomes have important implications for access to care, especially for rural populations.
The authors wish to thank the Jefferson Center for Research in Medical Education and Health Care for access to the Jefferson Longitudinal Study of Medical Education, as well as its key staff: Joseph S. Gonnella, MD, Mohammadreza Hojat, PhD, J. Jon Veloski, MS, and Carol Rabinowitz. They would also like to acknowledge the following: the Jefferson Foundation for access to the address files of Jefferson Medical College alumni; Clara A. Callahan, MD, and Elizabeth Y. Brooks, DPM, from the Jefferson Medical College Office of Admissions for their support of the Physician Shortage Area Program (PSAP); and Paul C. Brucker, MD, president emeritus of Thomas Jefferson University, and Richard C. Wender, MD, alumni professor and chair, Department of Family and Community Medicine, for their long-standing support of the PSAP. The authors would also like to thank each of these aforementioned individuals for their review and comments on prior drafts of this report. Finally, the authors offer their most important acknowledgments to the PSAP students and graduates who provide care for those living in rural areas and small towns.
This work, as well as Dr. Rabinowitz, Dr. Diamond, Dr. Markham, and Ms. Santana, were supported in part by Health Resources and Services Administration Predoctoral Training in Primary Care Grant D56HP08346.
The institutional review board of Thomas Jefferson University approved this study.
The authors presented this study, in part, at the Association of American Medical Colleges Sixth Annual Physician Workforce Research Meeting in Alexandria, Virginia on May 6, 2010.
1 Rosenblatt RA. A view from the periphery—Health care in rural America. N Engl J Med. 2004;351:1049–1051.
2 Institute of Medicine Committee on the Future of Rural Health Care. Quality Through Collaboration: The Future of Rural Health. Washington, DC: National Academies Press; 2005.
3 Council on Graduate Medical Education. Tenth Report: Physician Distribution and Health Care Challenges in Rural and Inner City Areas. http://www.cogme.gov/10.pdf
. Accessed October 15, 2010.
4 Agency for Healthcare Research and Quality. Health Care Disparities in Rural Areas: Selected Findings From the 2004 National Healthcare Disparities Report. AHRQ Publication No. 05-P022. http://www.ahrq.gov/research/ruraldisp/ruraldispar.htm
. Accessed October 15, 2010.
5 Pusey WA. Medical education and medical service. I: The situation. JAMA. 1925;84:281–285.
6 U.S. Government Accountability Office. Testimony Before the Committee on Health, Education, Labor, and Pensions, U.S. Senate. Primary Care Professionals: Recent Supply Trends, Projections, and Valuation of Services. http://www.gao.gov/new.items/d08472t.pdf
. Accessed October 15, 2010.
7 U.S. Government Accountability Office. Report to the Chairman, Committee on Health, Education, Labor, and Pensions, U.S. Senate. Physician Workforce: Physician Supply Increased in Metropolitan and Nonmetropolitan Areas but Geographic Disparities Persisted. GAO-04-124. http://www.gao.gov/new.items/d04124.pdf
. Accessed October 15, 2010.
10 Colwill JM. Education for and retention in rural practice. J Rural Health. 2005;19:233–235.
11 Politzer RM, Harris DL, Gaston MH, Mullan F. Primary care physician supply and the medically underserved: A status report and recommendations. JAMA. 1991;266:104–109.
15 Wilson NW, Couper ID, De Vries E, Reid S, Fish T, Marais BJ. A critical review of interventions to redress the inequitable distribution of healthcare professionals to rural and remote areas. Rural Remote Health. 2009;9:1060. http://www.rrh.org.au/articles/subviewnew.asp?ArticleID=1060
. Accessed October 15, 2010.
16 Rosenblatt RA, Saunders G, Shreffler J, Pirani MJ, Larson EH, Hart LG. Beyond retention: National Health Service Corps participation and subsequent practice locations of a cohort of rural family physicians. J Am Board Fam Pract. 1996;9:23–30.
18 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.
19 Rosenthal TC, McGuigan MH, Anderson G. Rural residency tracks in family practice: Graduate outcomes. Fam Med. 2000;32:174–177.
20 Colwill JM, Cultice JM. The future supply of family physicians: Implications for rural America. Health Aff (Millwood). 2003;22:190–198.
22 Rabinowitz HK, Diamond JJ, Markham FW, Hazelwood CE. A program to increase the number of family physicians in rural and underserved areas: Impact after 22 years. JAMA. 1999;281:255–260.
23 Rabinowitz HK. Evaluation of a selective medical school admissions policy to increase the number of family physicians in rural and underserved areas. N Engl J Med. 1988;319:480–486.
24 Rabinowitz HK. Recruitment, retention, and follow-up of graduates of a program to increase the number of family physicians in rural and underserved areas. N Engl J Med. 1993;328:934–939.
25 Rabinowitz HK. Caring for the Country: Family Doctors in Small Rural Towns. New York, NY: Springer-Verlag; 2004.
26 Hojat M, Gonnella JS, Veloski JJ, Erdmann JB. Jefferson Medical College Longitudinal Study: A prototype for evaluation of changes. Educ Health. 1996;9:99–113.
28 Isserman AM. In the national interest: Defining rural and urban correctly in research and public policy. Int Reg Sci Rev. 2005;28:465–499.
31 Rabinowitz HK, Diamond JJ, Markham FW, Paynter NP. Critical factors for designing programs to increase the supply and retention of rural primary care physicians. JAMA. 2001;286:1041–1048.
33 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.
38 Rosenblatt RA, Whitcomb ME, Cullen TJ, Lishner DM, Hart LG. Which medical schools produce rural physicians? JAMA. 1992;268:1559–1565.
39 Hyer J, Bazemore AW, Bowman R, Zhang X, Petterson S, Phillips RL. Medical school expansion: An immediate opportunity to meet rural health care needs. Am Fam Physician. 2007;76:208.
40 National Institutes of General Medical Sciences. MSTP Study: The Careers and Professional Activities of Graduates of the NIGMS Medical Scientist Training Program. September 1998. http://publications.nigms.nih.gov/reports/mstpstudy
. Accessed October 15, 2010.
42 Mullan F. The case for more U.S. medical students. N Engl J Med. 2000;343: 213–217.