Secondary Logo

Journal Logo


Efforts to Graduate More Primary Care Physicians and Physicians Who Will Practice in Rural Areas: Examining Outcomes From the University of Minnesota–Duluth and the Rural Physician Associate Program

Zink, Therese, MD; Center, Bruce, PhD; Finstad, Deborah; Boulger, James G., PhD; Repesh, Lillian A., PhD; Westra, Ruth, DO; Christensen, Raymond, MD; Brooks, Kathleen Dwyer, MD

Author Information
doi: 10.1097/ACM.0b013e3181d2b537
  • Free


Graduating adequate numbers of primary care physicians who will practice in rural areas (rural physicians) is an ongoing challenge. The State of Minnesota responded to this challenge during the late 1960s by creating two programs within the University of Minnesota (UMN) Medical School: the Rural Physician Associate Program (RPAP) and the medical school program in Duluth. These two programs have a similar goal: to increase the number of well-trained rural physicians who are practicing family medicine. Their strategies for attaining these goals are complementary.

The medical school on the Duluth campus (UMN–Duluth) was established in 1969 with a modest appropriation from the Minnesota legislature to UMN. The doors opened in 1972 with a class of 24 medical students; the average class size currently is 60 students. Students spend their first two years in Duluth and have their own curriculum.

The institutional mission is sharply focused on increasing the number of well-trained physicians entering family medicine who will practice in rural settings. In the early 1990s, the mission expanded to recruit Native American students. Admissions procedures incorporate literature-based predictors of specialty choice and practice site: rural origin and interest in family medicine at the time of matriculation.1 Applications and interviews elicit pertinent information about an applicant's residential background, his or her desire to practice in rural or Native American communities, his or her articulation of what it is like to grow up in a rural community, and his or her knowledge of and perceptions about the advantages and disadvantages of living in a rural community. Requests for letters of recommendation specifically ask referees to comment on how likely the student is to choose rural family practice.2 Preference is given to students who are from rural communities and who express a desire to specialize in family medicine and to practice in a rural or Native American community.

Since 1972, the Family Medicine Preceptorship Program has been central to the curriculum. During the first year of medical school, each student is assigned to a practicing family physician in Duluth and surrounding areas. Students meet with their preceptors 10 times during the course of the academic year. No other activities are scheduled during these times. Sessions are varied across days of the week and in both mornings and afternoons, to expose the student to the variety of activities of the preceptor (e.g., hospital rounds, clinic practice, nursing home rounds). These sessions begin during the eighth week of medical school and continue every three weeks throughout the first year. At the end of the first year and again for three sessions during the second year, students live with a rural preceptor and his or her family for three consecutive days and are exposed to the physician's everyday working environment and lifestyle. These communities are located in Minnesota and western Wisconsin, and many are contiguous to or located in medically underserved areas. UMN–Duluth medical school faculty meet with students three times a year to discuss the students' experiences.3,4 In 1990, this program received the National Rural Health Association's National Outstanding Rural Health Program Award as “a statewide or regional program that promotes or facilitates the development of rural health care systems.”

Minnesota's second initiative was the Rural Physicians Associate Program (RPAP), which UMN established in 1971 with a class of 23 third-year medical students from the Twin Cities (Minneapolis and St. Paul) campus (UMN–TC). Over the years, the class size has ranged from 19 to 47; the average is 33 students per class.5 Students spend nine months in a rural community under the mentorship of a primary care preceptor, usually a family physician. RPAP developed its training model in the early 1970s.5,6 Students experience the full scope of primary care and become part of the small-town community. They follow patients over time, acquire hands-on experience in a variety of procedures, and complete several required specialty rotations.7 In recent years, online learning modules and class discussions fostering connectivity and learning across sites were incorporated.

Both the Twin Cities and Duluth medical schools educate their students for the first two years with faculty and curriculum specific to each location. Students from both campuses can apply to the RPAP program for nine months of their third year. Duluth and Twin Cities students who do not participate in RPAP complete the majority of their required and elective rotations in the Twin Cities metropolitan area. Duluth offers select third- and fourth-year clerkships.

Thus, there are four groups of students: (1) UMN–Duluth students who participate in RPAP (UMN–Duluth/RPAP), (2) UMN–TC students who participate in RPAP (UMN–TC/RPAP), (3) UMN–Duluth students who do not participate in RPAP (UMN–Duluth/non-RPAP), and (4) UMN–TC students who do not participate in RPAP (UMN–TC/non-RPAP). This last group may be considered a natural control because nothing in their recruitment to medical school or training predisposes them to rural practice or primary care. The purpose of this study was to compare the outcomes of choosing family medicine or primary care and rural practice among these four groups.


Both RPAP and UMN–Duluth maintain data on their students. Each program collects the following information: demographics (e.g., age, gender, specialty choice, and first practice location), the type of community where the student was raised (i.e., rural or metropolitan), and his or her training, specialty, and fellowship (if any). RPAP collects information on all practice locations. The RPAP database is updated by surveying program graduates by mail every three years. Graduates were last surveyed during the summer of 2007. Returned surveys and newsletters are used to track individuals whose addresses have changed. Several sources were used to obtain similar information on the UMN–TC/non-RPAP students and to update current practice locations for UMN–Duluth students: (1) a database compiled by Dr. Robert McCollister, a former associate dean at UMN–TC Medical School, (2) the UMN Alumni Association, and (3) a master file purchased from the American Medical Association. Web searches were conducted to identify specialty choice and current practice location for graduates whose data were incomplete. The institutional review board of UMN determined that this study was exempt from review.

Descriptive statistics were used to describe demographics, the type of community in which the student was raised (i.e., rural or metropolitan), specialty choice, residency, fellowship, and first and current practice locations for classes graduating from 1990 through 2004. Communities were identified as rural or urban (here called “metropolitan”) by using the Office of Management and Budget's8 definitions of metropolitan and nonmetropolitan populations. Communities that were not listed were considered metropolitan if they were within 50 miles of an urban center; otherwise, they were designated as rural. Although metropolitan statistical areas (MSAs) are imperfect delineators, they are commonly used, and they allow comparisons across schools and states.9 Site designations were updated if their classification changed from rural to metropolitan over the periods examined; very few changed status.

Because our outcome measures were binary (yes/no) in nature, logistic regressions were run for rural practice location and specialty choice, after control for time. There are three independent variables: RPAP (yes or no), medical school location (Twin Cities or Duluth), and childhood community (metropolitan or rural). We were concerned about interactions between our independent variables, and so we explicitly placed terms for two- and three-way interactions in the logistic regression.


The entire sample included 3,365 students: 215 UMN–Duluth/RPAP, 276 UMN–TC/RPAP, 427 UMN–Duluth/non-RPAP, and 2,447 UMN–TC/non-RPAP graduates. Demographics for the four groups were similar, except that UMN–Duluth/non-RPAP graduates were significantly older than were graduates from the other groups. Both UMN–Duluth and RPAP had significantly more students who were raised in rural communities (Table 1). Eighty-five percent of the UMN–Duluth students grew up in Minnesota, compared with 79% of the UMN–TC students. Eighty-seven percent of UMN–Duluth and UMN–TC students are from the five-state area (Minnesota, Wisconsin, Iowa, North Dakota, and South Dakota).

Table 1
Table 1:
Demographics of Students (N = 3,365) at Either the Duluth or Twin Cities (TC) Campuses of the University of Minnesota (UMN) Medical School Who Did or Did Not Participate in the Rural Physician Associate Program (RPAP) Between 1990 and 2004

Student groups chose general primary care (internal medicine, pediatrics, internal medicine/pediatrics, or family medicine) and family medicine specialties in the following order from most frequently to least frequently: UMN–Duluth/RPAP, UMN–TC/RPAP, UMN–Duluth/non-RPAP, and UMN–TC/non-RPAP (Table 2). Students in those four groups chose primary care at a rate of 86%, 73%, 57%, and 36%, respectively (Figure 1). Internists or pediatricians who completed a specialty fellowship are included in the specialist group. Students who chose obstetrics–gynecology (n = 172) are also considered specialists, and the rates of that choice were 5.1% (n = 11) among UMN–Duluth/RPAP, 5.4% (n = 23) among UMN–Duluth/non-RPAP, 3.3% (n = 9) among UMN–TC/RPAP, and 5.4% (n = 129) among UMN–TC/non-RPAP students. Fifty-nine percent of all of these graduates practice in Minnesota, and 68% practice in the five-state area.

Table 2
Table 2:
Specialty Choice for Four Groups of Students (N = 3,340) in the University of Minnesota (UMN) Medical School Program and the Rural Physician Associate Program (RPAP), 1990–2004
Figure 1
Figure 1:
The percentage of graduates of the University of Minnesota (UMN) Medical School who chose to enter primary care, 1990–2004. The students attended medical school for two years at either the Duluth (UMN–Duluth) or Twin Cities (Minneapolis and St. Paul; UMN–TC) campus; in the third year, they either participated in the Rural Physician Associate Program (RPAP), a nine-month, longitudinal rural immersion program, or did not participate (non-RPAP).

Table 3 presents each group examined by the type of community in which students were raised (rural or metropolitan) and the current practice community. More than half (57%) of the students in the UMN–Duluth/RPAP group were raised in a rural environment and currently practice in a rural setting. The UMN–TC/RPAP and UMN–Duluth/non-RPAP groups have similar percentages of students who were raised in a rural environment and are currently in a rural practice (38% and 33%, respectively). Of the metropolitan-raised students who went to UMN–Duluth and/or participated in RPAP, the percentages who were recruited to rural practice were 48% of UMN–Duluth/RPAP, 25% of UMN–TC/RPAP, and 23% of UMN–Duluth/non-RPAP students. Nearly all (91%) of the students with an entirely metropolitan medical school experience chose to practice in metropolitan areas.

Table 3
Table 3:
The Type of Childhood Community and the Current Practice Settings for Students in the Four Program Combinations of the University of Minnesota (Duluth or Twin Cities [TC] campus) and the Rural Physician Associate Program (Participation or Nonparticipation), 1990–2004

Logistic regression with interactions demonstrated that the UMN–Duluth and RPAP experiences are additive for the outcome of graduating students who chose rural practice. Being raised in a rural community is also significant for predicting rural practice, although its unique effect is smaller than the unique effects of RPAP and UMN–Duluth. There is a small but significant interaction between being raised in a rural community and attending UMN–Duluth (Table 4).

Table 4
Table 4:
Logistic Regression With Interactions for the Outcome of Current Rural Practice Setting Among Students at the Duluth Campus of the University of Minnesota (UMN) Medical School Who Were or Were Not Raised in a Rural Community and Who Did or Did Not Participate in the Rural Physician Associate Program (RPAP), 1990–2004

Both the UMN–Duluth and RPAP experiences significantly affect the outcome of choosing a primary care specialty, and the effects are additive. Having grown up in a rural community has no significant effect on this choice (Table 5).

Table 5
Table 5:
Logistic Regression With Interactions for the Outcome of Primary Care Specialty Choice Among Students at the Duluth Campus of the University of Minnesota (UMN) Medical School Who Were or Were Not Raised in a Rural Community and Who Did or Did Not Participate in the Rural Physician Associate Program (RPAP), 1990–2004

More than half of all students (56%; 1,854/3,340) chose non-primary-care specialties. Eight percent (141/1,854) of these specialists currently practice in rural settings. The percentages of each group of students who are not primary care specialists and who practice rurally are 43% of UMN–Duluth/RPAP, 15% of UMN–Duluth/non-RPAP, 11% of UMN–TC/RPAP, and 8% of UMN–TC/non-RPAP students (results not shown in tables) (P < .001 for the difference between the percentage of UMN–Duluth/RPAP students and all others). Hence, the UMN–Duluth experience, more than participation in RPAP, significantly increases the chance that a specialist will choose a rural practice.

We examined the effect that an RPAP site's location close to a metropolitan area had on the graduate's choice between rural and metropolitan practice. For example, some RPAP sites are no longer actually rural but are now bedroom communities of the Twin Cities. Students who participated in RPAP in these communities were just as likely as those who participated in RPAP in rural communities to choose rural practice (37% and 41%, respectively; this difference is not significant).


The combination of the first two years of medical school in Duluth and the nine-month RPAP immersion experience yielded the highest number of rural primary care physicians. Of students who participated in both curricula, 54% chose rural practice. Both curricula successfully recruited metropolitan-raised students to rural practice; participation in either recruited 25% (109/395) of metropolitan-raised students to rural practice. Among all students, regardless of the type of community in which they were raised, involvement in either the UMN–Duluth program or RPAP led 34% (314/893) to choose rural practice. The rural practice rates of these UMN–TC/non-RPAP students (9%) were higher, but not significantly so, than the national rural practice rates (5%) reported in 2002.10 There was a slight negative interaction between being raised in a rural community and attending UMN–Duluth (odds ratio: 0.55; P = .03), which is not surprising, because UMN–Duluth recruitment targets students from rural communities. That is to say, the benefit of the Duluth program in promoting rural practice is that it has a slightly greater impact on metropolitan-raised students than on students raised in rural environments. Our results also demonstrated that the UMN–Duluth program contributed to increasing the number of rural specialists more than did RPAP.

Of the students who participated in both UMN–Duluth and RPAP, 86% chose general primary care and 77% chose family medicine. In contrast, of the students who participated in neither of those programs—that is, the UMN–TC/non-RPAP students—36% selected general primary care and 16% selected family medicine, outcomes that reflect the national trends.11

Given the current crisis in rural and primary care practice,12,13 these findings are important. Growing up in a rural community and expressing interest in primary care before entering medical school have been shown to predict rural primary care practice.14,15 Because both UMN–Duluth and RPAP have a rural primary care focus, and they select students who express such an interest, there is a selection bias in our sample. This study confirms that admitting students with an interest in rural primary care and both exposing them to these experiences during medical school (UMN–Duluth) and providing a longitudinal, rural immersion experience (RPAP) can recruit metropolitan-raised students to rural primary care practice. The UMN–Duluth experience also increased the likelihood that specialists would choose rural practice locations.

The faculty and administration at UMN–Duluth have not rested on their laurels. They continue to create programs to expose students to rural primary care experiences. In 2002, a 20-hour course, Introduction to Rural Family Medicine, which includes lectures, discussions on rural health issues, and a one-day experience in a rural community, was added to the first-year curriculum. The Summer Internship in Medicine was introduced in 2003. This two- to eight-week elective, which is offered between the first and second years of medical school to both Twin Cities and Duluth students, allows students to participate in a cross-professional experience such as pharmacy, home care, public health nursing, law enforcement, dentistry, chiropractic, or the hospital. Moreover, in 2005, the Early Admission Rural Scholars Program was started. In this program, up to five UMN–Duluth undergraduates in science or engineering who have high motivation and potential for practicing in rural or Native American communities are accepted into medical school after their junior year of college. The effect of these additional programs will be evaluated in the future.

There are limitations to this study. The Office of Management and Budget's rural designations (i.e., MSAs) are not perfect delineators, as other authors have shown.9 Both the UMN–Duluth program and RPAP have their own rural and metropolitan designations, based on community size and self-report, respectively. Hence, data on the programs' Web sites differ slightly from our findings. Current practice location was not available from our multiple data sources for 75 students (1.6% of the sample) and could not be found in Internet searches. Missing data were evenly distributed across the four groups and across all years. Both programs have collected data since their inception. We compiled the data for UMN–TC/non-RPAP students by combining several databases and online searches. We analyzed data only for graduates from 1990 to 2004, but it may be appropriate to evaluate this cohort, because a national shift away from family medicine occurred in the early 1990s.11 RPAP and the UMN–Duluth program appear to have protected the UMN Medical School from seeing its graduates follow the same dramatic national decline.5


Both UMN–Duluth and RPAP are achieving their missions, and these two programs are complementary. The Minnesota initiatives to increase the numbers of rural and primary care physicians have been successful. Efforts elsewhere around the United States to address the rural primary care crisis should build on these successful efforts.16


We are grateful to former Associate Dean Robert McCollister for carefully collecting University of Minnesota–Twin Cities medical school data for many years.



Other disclosures:


Ethical approval:

The University of Minnesota institutional review board judged this study to be exempt.


1 Phillips RL, Dodoo MS, Petterson S, et al. Specialty and Geographic Distribution of the Physician Workforce: What Influences Medical Student & Resident Choices? Washington, DC: Robert Graham Center; 2009.
2 University of Minnesota Duluth Campus. About our school. Available at: Accessed December 17, 2009.
3 Boulger JG. Family medicine education and rural health: A response to present and future needs. J Rural Health. 1991;7:105–115.
4 Boulger JG. Family practice in the predoctoral curriculum: A model for success. J Fam Pract. 1980;10:453–458.
5 Halaas GW, Zink TM, 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;24:356–363.
6 Verby JE, Newell JP, Andresen SA, Swentko WM. Changing the medical school curriculum to improve patient access to primary care. JAMA. 1991;266:110–113.
7 Zink T, Halaas GW, Finstad D, Brooks KD. The rural physician associate program: The value of immersion learning for third year medical students. J Rural Health. 2008;24:364–370.
8 Office of Management and Budget. Update of Statistical Area Definitions and Guidance on Their Uses. Available at:–01.pdf. Accessed December 17, 2009.
9 Hart LG, Larson EH, Lishner DM. Rural definitions for health policy and research. Am J Public Health. 2005;95:1149–1155.
10 Rabinowitz HK, Paynter NP. The rural vs urban practice decision. JAMA. 2002;287:113.
11 Association of American Medical Colleges. Graduation Questionnaire (GQ). Available at: Accessed December 28, 2009.
12 American College of Physicians. College warns of looming collapse of nation's primary care. Available at: Accessed December 28, 2009.
13 Kahn NB, Barnes ND. Family medicine faculty recruitment crisis of the '90s. Am Fam Physician. 1996;53:1514–1517.
14 Rabinowitz HK, Diamond JJ, Markham FW. Critical factors for designing programs to increase the supply and retention of rural primary care physicians. JAMA. 2001;286:1041–1048.
15 Brooks RG, Walsh M, Mardon RE, Lewis M, Clawson A. The roles of nature and nurture in the recruitment and retention of primary care physicians in rural areas: A review of the literature. Acad Med. 2002;77:790–798.
16 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. 2007;83:235–243.
© 2010 Association of American Medical Colleges