Evaluating Rural Medical Education Programs

Robert C. Bowman, M.D.

Introduction

For decades various "cures" for the shortage of rural physicians have been proposed. Many have been funded with the expectation of a quick remedy. All have been frustrated by the complexity of the issues involved. The demand for rural physicians continues, with more and more of the smallest towns facing the loss of all medical services. This paper arose from a legislative effort to document the expenditures of the state regarding health education. This evaluation continues, but the lack of experience and studies in the areas are a real problem. This paper can help you assess your state and identify areas of strength and weakness.

Challenges to Rural Medical Educators

Those who would study any intervention impacting on trainees during medical school face a long wait before the results are known. Students must graduate, do years of residency training, and enter practice before even the initial results are known. In the case of rural medical education, it is also important to see whether graduates stay in rural practice. As a consequence, researchers would need funding over many years. They would have to have very patient institutions, promotions might be delayed, grant officers would change, and political administrations would be out of office. Years after students had graduated, the first results might identify factors that may no longer have relevance, given the changes in medical education and medicine during the interval. This poses a dilemma. Interventions in rural medical education are often proposed on a "best guess" approach, based on the latest research as interpreted by faculty who have experience in rural health.

Since there are no schools that prepare rural faculty or researchers, rural faculty are often hybrids with backgrounds in other areas. Researcher-educators who understand rural practice, who can work effectively in medical institutions, and who are dedicated enough to wait for results over a number of years are most rare. Many loved the control and autonomy of rural practice and find it hard to work within an institutional framework. Others love to do procedures and clinical work and immerse themselves in these efforts.

Medical schools, in the twentieth century, once quick to adopt the scientific method and on the cutting edge in the advancement of science, have done little to advance the state of the art in rural health. Some of the most effective rural faculty and programs have depended on outside funding from the state or foundations.

Surprisingly little is known about rural medical education. It is difficult to access information in this area. An understanding of basic rural health is beyond the scope of today’s advanced search engines, crossing across the disciplines of medicine, economics, geography, and sociology. Rural medical education is even more limited. The few established studies involve single institutions and few variables.

Optimal studies would include multiple locations in a randomized prospective trial over many years, including special attention to areas such as selection bias, the background of the candidates, the curriculum, the rural nature of the state, and the precise details of the intervention. One of the most common interventions, the rural preceptorship, illustrates many of the problems inherent in rural studies. Often the interventions are elective, not required, so only interested trainees take the preceptorship (Pathman). Few preceptor studies involve much detail regarding the curriculum, preceptor, location, or other variables. Some preceptor studies note a great influence on a rural decision, others disagree (Grimes). Some limit the influence to certain sub-groups of trainees such as those with an urban background (Bruce). Others credit the preceptor for the impact, not the preceptorship (Mercer). Finally some feel that the influence of only a few weeks of training is small compared to many other factors over many years of training. So even this, perhaps the most-studied area of rural medical education, is poorly understood.

The Solution: A Continuous Series of Rural Programs

A new school is arising in medical education that takes a more comprehensive view of career influences. This school asserts that each level of preparation and training is dependent on previous efforts. Successful interventions would be continuous, concurrent, and coordinated. Such an effort would require significant commitment in the form of infrastructure, particularly political power for health educators and time and resources for faculty to create and maintain interventions.

Rationale: The support for this appoach involves work on multiple levels throughout medical training. Data in some related areas is also important, such as admissions data regarding minorities in medicine. Although seemingly different, there are similarities in both preparation and career choice. Minorities and rural background students often come from more limited educational backgrounds, and are more likely to go to underserved locations after the completion of training. Both have successful pre-professional programs that have demonstrated increases in minority or rural background admissions despite national trends to the contrary (PEPP, UTMB).

At the opposite end of the training ladder, recruitment and retention studies support multiple levels of interventions. Duluth-RPAP, the Jefferson PSAP program, the curriculum at Mercer and most of the newer medical schools are just a few examples.

Many rural medical education researchers agree that it is important to target trainees at each year of training to continue to keep them in the rural pipeline. If there are gaps in the segments of the pipeline or between the segments, then potential rural physicians will leak out and be lost. The segments include pre-professional, admissions, early experiences, clinical experiences, recruitment, and retention. An evaluation should take into account the effectiveness of the various components as well as the interdependence of the various programs.

The Impact of Rural Training, Early and Often

Despite the obvious common sense of training in more rural locations, only recently has there been sufficient documentation of the impact of rural training. Fowkes did a multisite study of physician assistant and nurse practitioner programs and noted the association between rural training and eventual rural location. Studies by Bowman demonstrated this for family practice residency programs. Mercer did studies on the preceptor, rather than the preceptorship, and noted a dramatic impact of the preceptor. This impact may be enhanced by the fact that rural family physicians have Mercer students during the transition between basic sciences and clinical medicine. Essentially this is their first exposure to clinical medicine.

Other programs note the important impact of "voting early and often" including Jefferson’s Physician Shortage Area Program in Pennsylvania (rural admissions, rural track, rural contacts yearly - see Rabinowitz article), East Carolina (medical students spend a weekend with rural physicians prior to medical school), East Tennessee (entry orientation and interdisciplinary training at a rural site over several years), University of New Mexico (3 months of rural interdisciplinary training at the end of the first year), and WWAMI which integrates multiple types of yearly training throughout the curriculum.

The Decision for Rural Practice: Complex and Changing Rapidly

The decision for rural practice can best be seen as a serious of influences over many years. Fully 30% of senior medical students with interest in rural practice knew this long before medical school as compared with their classmates who knew their career choice only 15% of the time. Better admission screening could potentially identify likely candidates at an early stage. The decision for rural practice does not end with medical training, however. Rural doctors in practice decide on a regular basis whether to continue in rural practice. Personal, family, practice, and career interests may change.

We also know that the decisions change over the years of training. Likely explanations for this trend include the influences of spouse and family which begin during this same time period, the influences of urban-based training, and the exposure to other medical careers. Another possibility is that students know that they must say that they are interested in rural practice to increase their potential for admission. Once in students would then be free to pursue their real interests. Surveys of these students could note a falsely elevated interest in rural practice, followed by a rapid decline in the next years (similar to what is actually found). An indicator of this is found in the unwillingness of matriculating students to sign up with the primary care loan program. Students who truly desired a rural practice would know that primary care was by far their most likely choice. Students are also influenced by their peers and the national outlook on specialty careers. Academic centers have few faculty, programs, or resources to support students with rural interest. Rural student interest groups can support students in times of great change.

Another problem may be the constant turmoil of academic and urban family medicine, where most students do their rotations. Students who take family medicine training in rural locations may see a purer, more consistent form of family medicine, one that is more isolated from the constant turmoil of academic medical centers.

Personal + environment The various influences on a rural career choice include personal characteristics of the trainee and the effects of training (Bowman). Other factors may have more of a global influence. Workforce demand can be one of those factors. In recent years medical students have dramatically shifted to more primary care specialties. Various organizations claim to have had influence, but most agree that the career choices of medical students are responding to the major changes in the practice environment. Despite these changes there is little effort to prioritize the selection of students who might withstand the considerable urban influences of medical training to consider rural practice. With medical school applications in decline, there is a great opportunity to increase the proportion of students from rural backgrounds or rural practice interest, if pre-professional programs are in place. Without some intervention, this opportunity is likely to be lost. In the past, declines in total medical school applicants have also meant lower rural background applications.

Admissions of those with rural backgrounds is important. This has been the basis of most successful rural interventions. Studies demonstrate that rural physicians are composed of equal proportions of rural and urban background students. Little is know about the rural students who did not choose rural or the urban ones who did. Admissions committees also face even higher priorities. They must choose candidates that are likely to do well on board exams. Smaller and newer schools may face great challenges in this area. Some studies (Harris) have shown that the higher the admission test scores and grade point, the less likely it is that the student will choose rural practice. This may also be a factor in declining applications and acceptances from students with rural backgrounds across the nation. Only a few states have programs to reverse this trend in declining rural background students. Kentucky’s Professional Education and Placement program is one of the oldest (began in 1971). Like RPAP in Minnesota, this program continues to receive legislative funds. Kentucky has broadened its applicant pool despite some lean years at the national level in the past.

The spouse factor A big unknown with any program or intervention is the spouse. Students enter college 90% unmarried, endure 11 years of college, medical school, and residency in urban areas predominantly, and emerge 90% married. Not surprisingly they meet urban spouses, often with professional careers. Locating one professional in a rural location is difficult. Locating two with medical practices or two different professional careers may be nearly impossible. Not surprisingly the interest in rural practice starts high and declines from thousands to only a few hundred a year of the 16,000 who enter the medical pipeline each year. Programs that accept students from the more rural state schools, programs that target married students with roots in rural communities, and programs that keep students out of urban areas during the training years would therefore be more likely to socially engineer a solution to the rural crisis. WWAMI branches out the first year of training to smaller locations, Nebraska and others accept out of rural high school for special tracks in small colleges, and all schools could look more closely at the spouse situation.

The Minnesota Rural Physician Associate Program is a good example for closer study for those who would plan and implement effective rural interventions. The program involves multiple levels, it coordinates primary care training in a more pure form, and it impacts the spouse. RPAP resulted from a the frustrations of a state senator who wanted more rural physicians. The state forced change on the institution and, to its credit (or debit), the state funded the program with one million dollars a year. Even with good funding, the program faced critics at the institution as well as outside. RPAP survived and demonstrated quality education. It also boosted rural physician production in the state. RPAP alone should not receive all the credit. The Duluth program acts as a feeder to RPAP to admit rural-oriented candidates and support them in their rural decision over the first two years. Many Duluth students choose the RPAP track of 30 or so students a year. RPAP students also have to take spouses with them on location. This may be a bias factor that may opt some out of the program and attract some in, depending on the experience. In any case medical education would do well to pay more attention to successful commercial rural recruiters who concentrate much of their effort into attracting the spouse.

Funded rural faculty Another outstanding characteristic of RPAP is the fact that the program funded rural faculty on a full time basis to initiate and develop the effort. Most commonly rural medical education programs receive only a small percentage of the total effort of the faculty (STFM Group on Rural Health report). Often family practice program directors, some of the busiest of faculty, take on the job of rural program development. Not content with just a rural product, John Verby, the initial director of RPAP, did more than just a rural program. He performed some of the most extensive evaluations of any medical education intervention. A former rural physician, he had to learn how to relate to the institution and state, as well as developing the evaluation and curriculum developer skills of an educator. Verby also noted the important economic and practice support contributions of students who did long term preceptorships. RPAP students had slightly lower initial performance criteria than other medical students at the University of Minnesota, but outperformed them after RPAP in multiple areas important to clinical medicine. Verby also worked behind the scenes to bring multiple solo and small group physicians together to locate offices adjacent to rural hospitals and share call systems. This facilitated his teaching efforts, but it also made it more possible to recruit rural physicians in the state. Other institutions such as Mercer and the University of Washington have felt the impact of former rural physicians acting through admissions, curriculum, and clinical experiences. Outcomes of this rural program would also have to at least examine these other impacts, in addition to the 60% of graduates that choose towns of less than 50,000. At the state and national level, the numbers of rural background admissions and rural-interested students should be tracked. AAMC Graduation Questionaire data should also be assessed yearly.

Rural Training Interventions Are Interdependent Duluth’s efforts in the first two years of medical school to recruit rural-interested students serve RPAP well. Programs that bridge gaps between medical school and residency such as the Nebraska Accelerated Residency Training Program also bypass the match and the decision process that could result in physicians choosing urban practice or a practice outside the state. The final decision is also important. States often spend millions on education, loan repayments, and scholarships only to loose physicians to other states who recruit and support physicians in a more comprehensive fashion at the end of the training pipeline.

Keeping Up With Other States

Since most states provide incentive funds to graduates, the funds are spent in a competitive fashion, rather than as an attempt to increase the number of rural graduates. Scholarship and loan repayments funds may actually go to trainees who already planned rural careers, but if the state did not provide these funds, the students might go to other states who did have incentive programs. In some cases the students might select urban rather than rural locations without the funds to shape their careers (specialty, residency, location, spouse, etc.). Funding for State Offices of Rural Health, Area Health Education Centers, and state recruiters falls into this category also.

Other states track graduates who leave their state, to lure them back periodically with offers to return home. Oklahoma paid family practice residents more than other residents, and paid them even more when they chose a rural community early. This was developed to keep rural-interested students in state training programs, rather than allowing them to escape the state and the important influence of program location on eventual practice location.

Other studies at the state level should note any significant gaps in the rural approach to training, as well as any bridging methods that would tend to keep students with rural interest in the pipeline to rural practice.

Limitations in Rural Health Education Research in 1997

Any conclusions reached by an evaluation of rural health programs must take into account the lack of adequate studies in this area. Other concerns include the great variation between states. It is very difficult to separate the health care sector from other influences such as education, economics, and geography.

I have struggled for years with the difficulty of doing research in rural medical education. One study of the family practice residencies in the nation has taken nearly 4 years of work. Lack of federal and state funding has much to do with this. One researcher, David Kindig (head of COGME), took a 10 year hiatus from rural health research from the mid-1970’s to the mid-1980’s and noted that little had changed when he returned. Indeed many studies published in 1995 look much the same as the single factor or single institution studies of the mid-1970’s.

Limitations of the Time Period

In addition to massive changes in Medicare, Medicaid, managed care, insurance, and hospital finance, there have been other changes, particularly in rural communities. Over the past twenty years several of the smaller rural communities have lost hospitals and no longer have physicians. Others have stopped recruiting. Still others do not recruit consistently or effectively. This has changed the supply-demand equation and may be one of the reasons that fewer family physicians are going to the smallest rural locations.

Additional losses of hospitals would likely continue to increase this problem. Many small hospitals are joining together in networks for common goals such as sharing expertise, purchasing, and recruiting. Those left out of networks and those in networks that do not recruit well again would be at risk of losing services and physicians permanently

Priorities change over time. The entry year of 1975 chosen by the evaluation was a time when UNMC was one of the most rural medical schools in the nation and not suprisingly, it produced the most rural doctors in the nation (Madison). The creation of more medical schools in more rural states and the increased numbers of different missions of UNMC have changed the focus.

Other changes are evident. Family medicine was hardly a household word in 1975, but now residency-trained graduates are nearing retirement. Many believe that current practitioners may not stay in practice as long as the past generation. They have many more career options than the general practitioners of the past and the newer generation of family physicians faces increasing family and personal pressures to leave rural practice. States hoping to keep rural physicians longer must consider an increase in resources to deal with these issues (Teplin).

Another concern is the 5% per year attrition rate of existing rural family physicians that choose to discontinue obstetrics (state survey of rural physicians 1995). Coupled with a known deficiency of obstetric training in the state, this may mean real problems for prenatal and obstetric care in a large part of the state. Obstetrical training in the state in family physicians and obstetricians needs to be evaluated. The state may need to pressure those doing much of the obstetrical volume to be more involved in training efforts in the state. This may include academic centers as well as private hospitals and health systems.

Limitations In Doing Single State Studies

It is important to be able to compare the target state with other states that have done similar projects, as well as those who have not done much in this area. It is also important to control for state characteristics such as the rural nature of the state (Rosenblatt, Bowman), the mission of the program (Bowman), educational emphasis, and other state and institution factors. One of the best evaluations of the state would be multistate or national comparisons, controlling for multiple factors.

Limitations Based on Institution Type

The structure of a medical school has much to do with rural doctor production. Schools with public funding and less National Institutes of Health graduate more rural physicians (Rosenblatt). Newer medical schools tend to have less NIH funding, but also they were established as primary care oriented schools and these schools tend to graduate more rural physicians. If your school has increased NIH funding or if it trying to establish itself as a tertiary center, it may mean more of a challenge for those attempting to develop rural health programs.

Unique Characteristics of States with Rural Populations

Each state has a unique population distribution. For example Nebraska’s population is concentrated along the eastern border and around the Platte River. Nebraska is tied with Kansas in being second in the nation in counties of less than 10,000 that still have a doctor and a hospital (Texas is first). Nebraska has nearly 18% of its population served by these smallest health systems. A ratio exceeded only by the two Dakotas. The state has also moved from mostly rural to mostly urban with changing needs and priorities.

As a result, Nebraska has many small rural locations, compared to other states. This makes for challenging recruitment. Small systems will always be only a few years from oblivion. Smaller rural locations need multiskilled individuals to be able to keep hospitals open at all hours. Medical technicians and x-ray techs are notable examples.

Because the medium and smaller rural communities have the greatest need, and because family phyisicians are by far the most likely to go to these locations, progress toward more family physicians and toward less urban-based training is important. Although it is too soon to see the effect of recent changes, training in family practice has reversed from mostly in Omaha, to mostly in special rural tracks. It will be another 2-3 years before significant changes are known.

Benchmarks for Evaluation

The rural health literature has identified a few markers that could be used to measure the state’s progress toward more rural practitioners. The amount of rural training is key for all. For PA’s and NP’s the degree of independence granted by the legislature and licensing boards is critical. Rural training does result in more rural locations (Fowkes) and more independence does mean more practitioners that choose and stay in the state.

Studies in medical education should measure progress toward graduating more family physicians, especially those likely to choose rural areas. This should include an assessment of whether more family practice residency positions should be available in the state and the training environment of these programs (rural, urban, underserved). Also the duration of the graduate’s stay in rural locations should be evaluated. The special preparation for rural practice may also have some impact that could be determined with appropriate studies.

Some states are very dependent on local training. A 1995 state workforce survey noted that UNMC trains 80-90% of the rural doctors in the state in 5 of 6 specialty areas (except psychiatry). The need for rural family physicians continues to be important, but psychiatrists and general surgeons may be a more critical need in the near future. It is also unclear why physician assistant and nurse practitioner graduates are increasinly unable to find practices in Nebraska. Perhaps Nebraska practices have been slow to accept physician assistants and nurse practitioners. To some degree this forces some to locate outside the state.

There have been other difficulties in the process that would allow some states to have more rural practitioners. There is no osteopathic school in many states and there is little knowledge about osteopathic schools at the high school and college levels in these states. Osteopathic schools emphasize primary care and many espouse rural practitioners as role models. This is in stark contrast to many allopathic medical schools where superspecialties reign. The state may need more efforts to educated students, parents, teachers, and counselors at rural schools about health careers.

Clearly, the state must assess the need for multiple types of practitioners and the obstacles to their location in Nebraska, particularly general surgeons, psychiatrists, family physicians, nurse practitioners, and physician assistants.

The Impact of the Spouse on a Decision for Rural Practice: The Sequel

The location of the training should be as rural as possible, not because of greater value, but because more rural areas mean more rural spouses to meet and marry, with a greater likelihood of choosing a rural practice. Graduates of professional schools tend to meet and marry other professionals. The smaller the rural location, the more difficult it is to find a spot that meets the needs of both professionals in the family. Often one spouse has deferred higher education, and these needs are more difficult to meet in rural areas in ways that might not meet rural needs, but do meet the needs of the state in terms of health care and economic impact.

The time period from 1975 to 1995 reflects great changes in the composition of the schools, with major increases in females and minorities. Also the practitioner’s spouse (male or female) has far more say in the location decision than in previous years. The growth in demand for primary care has made it easy to find a practice in rural or urban locations, with the more difficult spouse job dictating the location in more and more situations.

Program Funds vs Internal Funds

The impact of the rural programs of the state might be enhanced or even eclipsed by the internal investments that the various departments and schools have made, or by other efforts from other sources. Some faculty may have been funded for rural positions by various departments. Others have had funding vanish although titles remain. Programs may also exist more in name than in fact. The primary impact of these programs may be on relations between rural and academic communities, but some improvement in rural location decisions might result. Hard core infrastructure positions include assistants at the chancellors office, assistant deans, directors of divisions, special project directors, full time funded coordinators, plus other staff to assist, develop, and maintain the programs. It is not enough to have start-up funds.

Federal and foundation grants fund many rural projects beyond what the state provides. These include managed care grants, recruitment and retention efforts, networking projects, telemedicine, and other efforts. UNMC also sponsors various events such as the Recruitment Fair, recruitment dinners, mailings, conferences, and numerous rural visits that may have impact on rural practitioners directly or indirectly. An evaluation study should evaluate the impact of other resources and funding on efforts that might compete with or complement the state funding.

Limitations of Rural Medical Education Programs

One limitation is the lack of funding. To impact on a hundred or more students, change the curricula, change locations, and fund supervising faculty and coordinators represents a significant effort. Only Minnesota has gone to great length (RPAP $1 million per year) to fund a rural effort in an established school to help change the outcomes for the state. This effort funds a track involving some 30-40 students a year in a class of 200. Other schools brag about more dollars, but the dollars are diluted by multiple factors. For example a typical Robert Wood Johnson grant was $6 million but this was divided into 5 years by 3 schools (medicine, nursing, allied health) with multiple departments. Without a great deal of other funding and infrastructure, it represents very little impact.

Many rural efforts are voluntary on the part of the student, practitioner, and community. It is difficult to dictate major changes or restructure the experience if some support is not there. Even in a rural state such as Nebraska, communities may contribute over a million dollars a year in housing, meals, travel, and stipends - more than any grant ever proposed. If the countless hours of supervision are added in, the total would be staggering. States with a major voluntary effort may easily eclipse efforts in other states.

The limitation of many rural programs is the need to limit participation, due to lack of funding and faculty. Many rural interventions involve a small percentage of the students. This also limits the impact of the program and makes evaluation impossible as selection bias is a major concern.

Often these special programs may only attact and support those with rural interest anyway. The outcomes of such special programs may be different than just the typical first practice location in a rural town. Graduates of special rural programs may benefit from more rural-oriented training and could stay longer in the rural location, as Rabinowitz noted in his PSAP publications.

Sometimes programs have unanticipated consequences. Special admissions programs for rural background students may choose those who would have already been accepted, or else they might push aside those who would be accepted last into medical school. There is some evidence that the those accepted at the last of the admissions process were more likely to become family physicians. Studies of declining class size have shown that the family practice match declined when classes got smaller (Bauer). It is not unreasonable to think that this might occur.

Successful rural programs have faculty that work with students, with the institution, with the state, with other rural faculty doing similar programs, and with rural communities. It is a daunting task, but one that must be undertaken in order to meet the health care needs of underserved rural populations.

The following items review priorities for rural medical education programs and finally, recommendations for evaluation on an institutional or statewide basis.

An Overview

Institutional Mandate for Rural Health

The institution's mandate must be stated, understood, and applied. There should be no other major agendas for the institution. This is particularly difficult when NIH research funds eclipse all other grant funding and continue to more tempt medical schools to fund teaching with subspecialists or hospital-oriented care. This is enforced by accreditation bodies of medical education who evaluate the research and academic practice areas and can choose to penalize primary care focused schools. The teaching of rural primary care can be inhibited by the development of non-primary care residencies, the expansion of other primary care residencies beyond the resources (patients, faculty, staff) available to train physicians, or the expansion of subspecialty private practice domains. Research can pose a problem as a focus on "pure" research rather than rural or applied research can deflect the institution. A priority on academic development can take faculty and other resources away from primary care. Small institutions are far less able to support multiple missions or mandates.

Finance

New programs take money to start. Rural programs need to begin small and build up. Each project feeds into the next. Many sources of funding are possible but it takes time and faculty resources to evaluate and pursue the best ones. Right now the best sources are state and federal funds. Yet both have strings, restrictions, and the threat of recession. Constant communication with the legislatures should minimize these problems. With a true institutional mandate, medical school funds should be a source. Grants are also a possibility, but often leave programs short of people (coordinators, secretaries, faculty time) and equipment resources. Adequate resources to manage the grants often don't come until 1 - 2 years later due to delays in hiring.

Faculty

Faculty must be experienced in rural health. They should be committed, enthusiastic, and able to work together as a team. There must be enough faculty for an adequate critial mass. Rural faculty must have access to the latest information on rural and family practice programs in order to best pursue teaching resources, funding, and other necessary elements for the rural programs. Networking and travel expenses are necessary. Faculty development encompassing all the above is essential.

 

Leadership

It is necessary for information and resources to be assimilated into a vision that is do-able for the institution. This will involve change, political struggle, tact, diplomacy, and a willingness to invoke the institutional mandate when necessary. Efforts must be directed internally as well as externally.

 

Curriculum

The curricula demands more than just a series of educational programs. It involves a careful selection of faculty, experiences, and communities - all those who impact on the environment of rural medical education.

Robert C. Bowman, M.D.