The majority of interest and activity in rural graduate medical education appears to be in family medicine, but other specialties provide some rural exposure to residents.

1. GME funding that would encourage rural training must be simplified and linked to outpatient care. Medicare also needs to continue to be flexible regarding payment to physicians who supervise residents on rural rotations.

2. Training programs will need more flexibility from residency review committees so that they can facilitate rural training, especially regarding limitations on the length of rural training.

3. There is clearly a need for a greater understanding of factors that can influence the practice location choices of residents. Carefully planned trials including multi-state and national interventions could induce greater numbers of graduating residents to locate in rural areas. Community Driven Approach Article for comprehensive approach with proven results

4. Exceptions to the recently imposed cap on residency positions may need to be made to allow rural programs that have begun development to continue, as well as allowing for the creation of new rural programs.

5. Information about rural medical education should be distributed more effectively to medical students considering a career in rural health as well as faculty, GME programs, and institutions that are considering rural medical education interventions.

6. Rural communities and practitioners would benefit from longer rural GME experiences. Long rotations would encourage more local funding of rural programming. Community Friendly Aspects

7. Studies regarding the future of obstetrics services in rural communities are needed to anticipate potential problems with prenatal access. For an idea see Federalize Liability For Underserved

This report summarizes the issues relating to graduate medical education (residency training) of physicians for rural practice. It includes a summary of available published and unpublished literature as of Fall, 1997 and the results of our interviews with persons involved with rural residency training. The report summarizes what is known about rural residency training, including costs and outcomes, difficulties, and barriers to the creation of programs. The report emphasizes efforts in family medicine because family medicine has been far more active in this area than other specialties. Only 8 - 10% of internal medicine, pediatrics, and obstetric residents choose rural locations and these tend to be areas that are not underserved (Study of Models 1992).

The need for greater numbers of physicians to practice in rural areas is clear. For example, in 1975, rural counties in the US held 24% of the population but only 16% of its primary care physicians. By 1990, this had worsened - 23% of the US population resided in rural counties but only 14% of primary care physicians practiced there. This occurred despite a 75% increase in the number of primary care physicians in this time period, while the overall population grew only 17% [GAO, 1994 - cited in Norris and Acosta, 1997].

Rural Medical Education Involves Multiple Steps

Rural-interested students seem more aware of current issues in medical education. More rural-interested students wanted increases in the medical school content areas of primary care, health promotion and disease prevention, public health, nutrition, the role of medicine in the community, the physician-patient relationship, drug and alcohol abuse, and patient education when compared with medical students as a whole. Again, involvement is the key for them. Those hoping to increase the numbers of rural physicians through GME would do well to facilitate the interests of residents as well, and get them involved in rural communities, community medicine projects, overseas missions, and similar types of training.

Family medicine, since its inception, has placed a much greater emphasis on training in outpatient settings. Many programs have offered or required some training in rural settings. There are two primary goals of training residents in rural areas: producing greater numbers of physicians who will practice in rural areas and producing physicians who are better prepared for the demands of rural practice. These two are not completely unrelated, as providers who are better trained to deal with the clinical demands of their practices are potentially more likely to stay in their practices. There is a strong feeling among educators involved in rural physician training that training in rural areas helps to prevent medical students and residents (and their families) from developing ties to urban areas that might keep them in urban locations.

The initiation of formal training in family medicine had an impact on the graduation rate of rural physicians. From 1975 to 1985 students choosing family medicine were over three times more likely to choose rural practice as other medical graduates (AAFP Graduation Surveys, Study of Models). Rural communities finally had a medical specialty that produced graduates that could serve them.

A national study of rural family physicians noted the importance of rural experiences in training as 31.5% took a required rural rotation in residency, 48.5% took an elective rural residency month, and 59.5% took a rural clerkship in medical school (Norris).

A 1987 survey of U.S. family medicine residency programs found that 49% of the 293 responding programs offered elective rural rotations, but only 31% of programs (n=84) required them (Blondell 1989). A 1995 survey of family medicine residencies noted a major increase in rural programming with 40.2% of 353 programs with required rural experiences and many others with new forms of rural training.

During the period between the prior two studies, there has been a modest decline in family practice graduates choosing rural practice. Now family practice graduates are only twice as likely to choose rural practice as compared to other medical graduates with 22% or 580 graduates per year as compared to 680 a decade ago. The numbers choosing towns of less than 2500 have declined by 50%. Chart notes the total FP graduates rises over the years but the number choosing rural practice is not increasing

On first examination this appears to suggest that rural training has had no impact. Of course during the same time period there has been a threefold decrease in graduating medical students interested in rural practice (sequential AAMC GQ reports), major changes in rural health, and increased pressures on students to choose primary care careers. More women and minorities have entered family medicine. Also an increasing proportion of new family practice programs have located in towns of over 1 million in population. These locations rarely graduate rural physicians with only two of 33 such programs above the national average of 22% of family practice graduates choosing rural locations. (Bowman, FP Grad studies, more at Fam Med Res Prog and Grad of Rural Family Physicians) The cap on new programs may be a help in distributing programs where they are more needed.

Program directors and chairs also noted increasing concerns about regulations that limit the training of medical students and residents. Trainees must have some responsibility for patient care decisions in order to learn. One of the real assets of training in rural areas is that rural physicians know their patients better and let students and residents do more with patients. This accelerates their learning. If we retard their training potential, they will have little choice other than to specialize. see Meeting the Needs of Underserved Rural and Inner City Areas with Accelerated Graduate Training Medicare requires that the billing physician see the patient to at least review the history, diagnosis, and treatment plan. Medicaid and other forms of insurance have also taken this view in many states. This is obviously more work for busy rural physicians who already take an extra hour a day to supervise trainees. They may be frustrated with the need to see the patients of residents that are near graduation and can function more independently, as well as the need to document this review on each patient medical record. The end result may be fewer sites, less experience for trainees at graduation, and less comfort with procedures, obstetrics, and rural areas. More residencies will need to require residents to obtain full licensure so that the "hands-on" learning can continue. Academic programs that adopted Community Friendly Aspects would make it better for rural communities and their physicians.

Other forces also restrict the training of students and residents and this may make a rural decision less likely These include liability worries, too many trainees in too few locations, and the need to make teaching centers financially viable. See Verby article Learning Forestry These all tend to reduce the scope of training and the time allotted for teaching. One more direct impact is that it is rare to find family practice residencies that give enough surgical or obstetrical experience to prepare residents to do the same procedures that many existing rural physicians do today. Some faculty feel that students who might have chosen family practice and chosen a rural practice, end up choosing general surgery or obstetrics, specialties with a lower potential for rural practice choice.

What is clear from the past changes is that each level of training seems to be dependent on prior efforts. Medical schools with success in rural graduation rates agree that preparation for rural practice needs to occur at all levels of training, preadmission, admission, all years of medical school, and all years of residency.

A number of factors are associated with choosing to practice in a rural location. Required rural training is one of the key factors as outlined below (Family Practice Programs have 36 months of training):

from Fam Med Res Prog and Grad of RFP

While it is possible that students already had chosen rural practice, students did not have ready access to information about the programs that would help them choose a rural-oriented program. Note also that few programs have more than 2 months of time for rural rotations, unless the program is in a rural location or has a rural training track. Given the restrictions of the RRC to the continuity requirement and the packed overall curriculum, even 2 months represents a significant commitment to rural training.

Obstetrical training also has an impact on resident practice choice:

from Fam Med Res Prog and Grad of RFP

Since relatively few programs have more than a minimum of obstetrical training, this is an area that deserves closer scrutiny. An intervention in obstetrics has the potential to increase the numbers of rural physicians.

Family practice residents chose rural practice more often when they were not supervised on rotations by residents or faculty of other specialty types (internal medicine, surgery, pediatrics, obstetrics, psychiatry, and emergency medicine):

from Fam Med Res Prog and Grad of RFP

Other types of rural training have an influence on rural location. One study notes that 76% of graduates from the 13 identified family medicine RTTs existing in 1996 have started practicing in rural areas (Rosenthal 1997).

There is suggestive evidence from Canada that training specialists in rural areas may lead to substantial numbers choosing rural practice. As with many studies in rural medical education, it is difficult to control for the selection bias of the trainees (Gray 1994).

Studies of nurse practitioner and physician assistant programs at multiple sites demonstrated that training in rural sites resulted in more rural graduates (Fowkes 1994).

Graduate Medical Education in Rural Areas

A goal of this report was to summarize the rural training activities for graduate medical education for specialties providing aspects of generalist care in rural area - family medicine, general internal medicine, general pediatrics, obstetrics and gynecology, and general surgery.

Interviews with key informants suggested that the vast majority of the graduate medical education activity taking place in rural areas to train practitioners for rural practice is taking place in family medicine.

The status of rural training in family medicine in 1995 was documented by a national study that surveyed the program directors of all family practice programs with graduates with 353 or 96% responding. The descriptive findings are summarized below:

Choices in Graduate Programs      Table of Graduate Programs

Family medicine has had a variety of training opportunities in rural areas. Most of these are rotations or ambulatory clinic experiences. Twenty-five residency programs are located in rural locations. Some programs involve RTTs: 1+2 year residencies in which the first year of training takes place at a parent program located in a sizable, generally urban/suburban location, and the last 2 years take place in a rural location.

A list of all family medicine RTTs was assembled from in the Directory of Family Medicine Residencies and an informal list of all approved 1+2 rural family medicine programs. This included residencies located in rural areas (urban influence code > 2), and all other programs claiming to offer some rural experience. This information is summarized in Table 1.

Interviews of faculty and directors of both three year residencies in rural areas and RTTs highlighted common issues and provided contrasts. have some common issues and some issues that are unique to one or the other type of program. The interviews were more heavily weighted toward RTTs (n=10) than rural residencies (n=3) as RTTs are a relatively new phenomenon in family medicine and the issues for them are less well-known than for traditional, 3-year residencies.

Both types of programs require several years of time and start-up funding, which was variously supplied by state and federal grants and sponsoring hospitals. Potential loss of Title VII funding for family medicine residencies was cited by a number of persons interviewed as a barrier to establishment of new RTTs. Three-year residencies tend to build their own facilities and create their own practices, though they often hire one or more local practitioners as faculty and utilize these practices for the residency practice base. RTTs may use this model, but more frequently (in our sample) functioned in the setting of an already-existing practice, with the partners in the practice serving as faculty and one of them being the site director.

of the local director and practitioners.

If the site director became busy or lost interest, this could result in site closure. In the larger 3-year programs where the practice was part of the residency, faculty teaching time was shared between more individuals. At smaller sites, the loss of a single practitioner-faculty might threaten the viability of the program.   Why Choose a Small Rural Training Program?

Approximately 70% of RTTs have been reported to use telemedicine links (Rosenthal, McGuigan et al. 1997). This was confirmed in interviews. Most programs had telemedicine links to the parent hospital of the program (the site where residents did their first year of training and which sponsors and administers the RTT). In many cases, programs were able to take advantage of pre-existing telemedicine linkages - e.g., outreach efforts from the parent hospital to surrounding rural hospitals. In some cases, grant funds were obtained specifically for telemedicine links. Some programs felt that, once the initial hardware was obtained, ongoing support of the telemedicine link would not be a problem, while others stated that, if grant funds disappeared, they would be unable to afford the ongoing operating costs for the link.

Telemedicine links were being used primarily to bring didactic teaching sessions out to the RTT site from the parent program. In many cases, some or all of these sessions were also open to physicians other than the residents at the remote sites and continuing medical education credit could be obtained by attending these sessions. One site was hoping to obtain approval from the Residency Review Committee (RRC) to use a telemedicine link to allow a resident at an RTT to be precepted by a physician at the parent program. Some programs plan specific faculty development curriculum for RTT faculty via telemedicine links.

Other Rural Efforts in Family Medicine

In the past decade, family medicine instituted a dozen accelerated residency programs that accepted a few select fourth year medical students into university-based residency training as first year family practice residents. In Nebraska this program emphasized rural practice. Third year students interested in rural practice went through an orientation and interview process and satisfy fourth year medical school and first year family practice residency requirements in the same year. These residents also commit to a rural practice, receive stipend and loan benefits, and take a fourth year rural fellowship as part of the program.

A different program operated through the internal medicine department begins in the fourth year of medical school and allows students to go through the match to choose a Nebraska family medicine or internal medicine program. Most of the graduates of this Primary Care Track have chosen rural training tracks in family medicine residencies with again 75% - 100% of each group of 4 choosing rural practice (n = 18). Both of these programs have kept some of the best students in the state rather than loosing these students to training programs in other states. The development of such tracks has greatly enhanced the recruitment of rural training track residents, stabilized the recruitment at the main program site, and greatly increased the procedural training of the graduates in the areas of endoscopies, intubations, and caesarian sections. Nebraska Rural Programs Summary and Links

Oklahoma increased the salaries of its family practice residents, especially the ones who chose rural Oklahoma communities early, to accomplish the same goal of keeping trainees in the state. By the numbers for info on Oklahoma and also impact of rural physicians on the local economy from several sources

By accelerating the location decisions of graduates, states and institutions may be able to encourage more to choose rural practice.

Bridging techniques may also help to bring rural background students into early admission programs in medical schools with the hope that they will be more likely to enter rural practice. Residencies may bridge the gap between residency and practice by involving residents in rural communities or by involving communities more actively in the residency with recruitment fairs, lunches, dinners, precepting, and ambulatory clinics. The major impediment to long term commitments and bridging techniques is that it is impossible to predict whether the spouse will be willing to go to rural practice, especially when the interventions begin in the college years when 90% of students are single. The Continued Centralization of State Educational Resourcesand the Potential Impact on the Location of Young Professionals

"RPAP (MN) builds bridges of good will between the University and rural doctors. Many physicians recall the days when the hapless and hopeless referring LMD was dismissed with scorn at the professor's grand rounds. Times have changed! Monthly University visits have done much to dispel the doubts. When men of professional stature… come out to the rural hospital, make rounds, break bread, and discuss common problems with the local medical staff, new bonds of understanding grow quickly between small town and gown."

Short Term Rural Training

Rural rotations can be formally arranged and required by the training program, or arranged as an elective by the program or resident. Elective rotations often are used to check out a potential practice site. More formal arrangements offer practice sites a consistent stream of residents, allowing them to add a nurse or other support. Both rotations support practitioners with after hours coverage. Ideally programs would have 2 - 3 month rural rotations at regular sites, with another month to check out a potential practice site. Since communities benefit from a dependable source of workforce, they are more likely to support the salary and benefits and travel costs of residents. Rotation of the sites can also add a dimension of stabilizing shortage area and 2 - 3 physician rural practices, at high risk of overloading current practitioners. The CORE Program   Community Friendly Aspects

The Society for General Internal Medicine published a directory of primary care internal medicine residencies in 1992 (Society of General Internal Medicine 1992). There were a only a few residencies located in rural counties (urban influence code > 2). Most of these were residencies for large, multispecialty groups headquartered in rural areas (e.g., Guthrie, Sayre, and Geisinger) and all highlighted the size of their referral areas and the advanced technology available in their hospitals. None mentioned preparation for rural practice as a goal. Some programs do offer elective or rotational exposure to less urban environments (though some of these are still in metropolitan counties, such as the Delano, CA rotation of the Bakersfield, CA general surgery residency and the Toppenish rotations of the University of Washington internal medicine and pediatrics residents).

The University of Washington WAMI (Washington, Alaska, Montana, Idaho) program was one of the pioneers in providing rural experiences to residents, and currently internal medicine sends up to 20 residents per year on 1-2 month rotations to 5 different sites, UW Pediatrics also uses several sites. There is a very sparse literature documenting rural training experiences in internal medicine (Crandall 1978; Parenti 1995), pediatrics (Kairys 1985), and general surgery (Asher 1984). There is some suggestion but no clear proof that these may lead to an increased likelihood of starting in rural practice (Asher, Martin et al. 1984; Kairys and Newell 1985).

There is no record that GME funds have supported any rural training other than 3 year residency programs and rural training tracks. The barriers appear to be to great to pursue, even though funding for some 5000 months of residency training is involved. Ambulatory clinics, rural rotations, fellowships, and other rural programs must obtain funding from local hospitals, the state, or practice revenues. Discussion of funding therefore mainly concerns the longer forms of rural programming.

Both 3-year residency and RTT types of programs required substantial start-up funds and several years of effort. Start-up funds were obtained from a variety of sources: the federal government through Title VII grants, state government grants, funding from sponsoring hospitals, and private grant funds. Starting an RTT is clearly much more like starting a new residency than adding a couple of residents to an existing program.  Choices in Graduate Programs

Once an RTT is running, some felt that costs per resident were roughly the same as for residents in a traditional program, and estimates of the annual cost per resident around 1991 in four programs ranged from $61,000 to $100,000 (Rosenthal 1992). For 1996-7, Family Medicine Spokane estimated their annual cost for RTT residents to be $73,274, about half the annual cost of a traditional resident (R. Maudlin, Pharm.D., personal communication, 4/97). Others, however, felt that there were additional costs to running an RTT.

Additional costs included telemedicine links, the cost for faculty from the parent program to travel to the RTT several times a month to supervise residents, and the cost for RTT residents to travel back to the parent program for conferences and rotations that could not be supplied at the rural site. RTT residents returned to the parent program anywhere from weekly to once every 3 months, and programs paid these travel expenses. Sometimes this included an apartment or other place to live. The costs of faculty development at the RTT were significantly higher than at the parent programs. Local RTT program directors may need to attend two or three national meetings a year to obtain the training to be able to manage the program well (Residency Assistance Program and Program Director’s annual meeting) in addition to attending the student-resident annual meeting to recruit students interested in rural training.

There is a great variety of RTT program types. The costs vary by the size of the program, the distance costs, the commitment of the local practitioner-faculty, the willingness of the rural community to bear some of the burdens, and the educational resources available at the local community. Some costs are subject to interpretation. Some felt that residents were financially neutral for a practice (Maudlin 1997), with decreased productivity by faculty roughly offset by the productivity of the residents. Others felt that residents were a cost to the sponsoring practice.

It is important to note that the first RTT, the Spokane program, was granted much more flexibility by the Residency Review Committee and its rural sites are much smaller and more dependent on local faculty than many RTTs. The increased accreditation requirements for subsequent RTTs may be part of these costs.

It is clear in RTTs as well as rural rotations that many rural areas, particularly those that are more in need of physicians, seem to be much more committed to training family practice residents and therefore demand less funding. Teaching by primary care and specialty faculty is often provided free of charge in these locations.

New Residency Review Committee rules now restrict faculty from supervising residents and seeing patients at the same time. In traditional programs this is less of a problem, because there are several residents at the program and each faculty can supervise 3 or 4 residents. In many RTT programs there may be only one resident to supervise. RRC rules do allow faculty to supervise one resident and have a 50% clinic workload, but do not allow any clinic time for faculty with two residents to supervise. This again is a rigid policy that inhibits smaller programs. RTT residents in rural areas also tend to be third year residents that are near graduation (Third years have more clinic time, there are no first year residents at most RTT sites, and third year residents have more clinic time than second years.). Third year residents need little supervision. There is some evidence that RTT programs need a mix of experienced and resident practitioners at the rural clinic site to build and maintain a good patient base. Restrictions on the use of supervising practitioner-faculty can increase costs and decrease the patient base, particularly for the smaller programs. This puts RTT programs in financial difficulty and may make it more likely for programs to be in conflict with other regulations, such as clinic patient volume.

In the future, it may be more difficult to get faculty to teach residents. Rural programs depend on volunteer community faculty to provide teaching for specialty care, as there are not other residency programs around that they can take advantage of for teaching their residents. Some programs expressed concern that they might not be able to continue to obtain this specialty teaching without having some funds to pay these specialists. As physicians are increasingly employed, decisions such as whether to teach add this third party, one that often values clinic productivity rather than the future quality of medicine in the nation.

Changes in Medicare will also impact on existing rural rotations. No resident will be able to go out to rural practices for the first six months of training. Most residents will have to be fully licensed to go on rural rotations so that the practices can bill for their services. If residents are not licensed, rural physicians will have to see every patient and confirm the history, diagnosis, and plan of treatment. This will limit the scope of practice for trainees and make it less likely that rural physicians will agree to train residents.

The most commonly-voiced complaints regarding funding related to Medicare graduate medical education (GME). This information was not easily obtained and was often confusing. GME funding includes both direct medical education (DME) and indirect medical education (IME) payments. Some RTTs were getting neither DME nor IME funds, reporting that their rural hospitals lacked the expertise to apply for these funds or the ability to pay a consultant to help them. This was much more common with IME than DME funding. Some interviewees commented that it didn’t seem worth applying for IME funds because the IME payments would be too small to be worth the effort. Two reasons were cited for this: first, the rural hospitals they were using were so small that few Medicare patients were treated there, making the Medicare base quite small, and second, much of the resident’s training time was spent outside of the hospital. These were felt to be prejudicial against small, rural hospitals and primary care training programs. The costs of applying for IME funds are similar whatever the size of the hospital and whatever the proportion of inpatient to outpatient training, so that the fixed costs of applying are a much greater burden on small than large hospitals and outpatient-focused than inpatient-focused residencies.

Interviewers frequently reported conflicts with regional Medicare intermediaries about defining a program’s Medicare base or eligibility. For example, one program fought to have the costs of building a clinic for resident education and recruitment costs of faculty (a challenge for smaller rural sites) included as educational expenses and part of their base, but their carrier was denied payment. Another program reported that, after years of effort, they reached agreement with their state’s Medicare carrier. Unfortunately the carrier changed and the new carrier refused to recognize the agreement with the former carrier.

The fact that Medicare GME payments all go to the hospital, not the program, was an issue for RTTs. Unlike traditional, hospital-based residencies, RTTs are based primarily in a rural practice.

The recent uncertainty over what changes would be made to Medicare GME payments was not cited by most programs as an issue. Of course, since many of them were receiving only DME funding or no Medicare GME funds, this is not surprising. A representative of one program that had just folded stated that, while several issues contributed to the program’s demise, uncertainty over GME funding was one major factor. Another program director strongly felt that continuing stable levels of reimbursement for rural training sites was essential for their continued stability. While Title VII funds support primary care residency program development, the tens of millions of dollars pale in comparison to the billions provided through Medicare GME payments.

Analogous to the issue of IME payments to RTTs were the findings of a preliminary evaluation of the Rural Health Medical Education Demonstration Project of the Bureau of Health Professions (Ricketts 1994), a program designed to allow IME payments for residencies providing rural experiences for their residents. It found that a number of initial applicants withdrew from the program and participating residencies complained about the lack of payment for start-up, overhead, and rural hospital costs. They also noted that the residency, not the hospital, should receive the funding since they were doing the teaching, the clinic, and most of the administrative costs.

Some rural sites also have nurse practitioners or physician assistants and Medicare rules specify that faculty cannot supervise these practitioners at the same time that they supervise residents. Clearly this requirement is an impediment to all training programs and a particular burden for smaller programs. It is also an insult to nurse practitioners and physician assistants who often need little supervision. Medicare should allow each faculty physician to supervise a total of up to 4 residents or other practitioners.

Residencies tend to be more dependent on clinical income from seeing Medicaid patients than private practices. One program director commented on how low Medicaid reimbursement was relative to other payers and how frequent denials of payment by their state Medicaid carrier were.

Since restrictive funding is a reality, programs have had to explore alternative sources of funding to continue to train in rural areas. Rural hospitals often did contribute some funding for the residents at RTT and rural rotation sites, though the programs generally did not know how this compared to the Medicare GME funding (for those hospitals receiving any GME funds). This is similar to findings of another study of primary care residency programs (Center for Health Policy Studies 1996).

Residents were felt to provide a significant service by expanding the pool of physicians available to cover rural hospital emergency rooms and helping out with increased clinic workloads, especially when the community has lost a physician. One rural hospital was felt to have been saved by the additional support provided by the new residency program. Again the two month RRC restriction keeps the residency from providing as much benefit to communities as is possible. Longer rural experiences are a better investment for rural communities, practitioners, and hospitals. There is more continuity with patients, less need for orientation of new residents, and more after hours support for local rural practitioners. Last year, residents provided over 5000 months of services to rural communities on required rotations alone. This figure does not include elective rural rotations, ambulatory or rural satellite clinics, and residents who are paid to "moonlight" in rural hospitals.

Policies that reward training in rural areas, simplified funding, and allowed residents to spend 3 months in rural areas could double the amount of rural services provided by residents. If these rotations and rural training tracks were used to stabilize rural health systems that were most in need of workforce, there could be benefits on the retention of existing rural physicians. If these policies had been in place sooner, perhaps some of the 800 family practice faculty who were once rural physicians (out of 2500 total M.D. and D.O. faculty) could stay in the communities and teach.

Public-private networking has resulted in some unique proposals to increase the amount of rural training. Rural faculty have long noted that many residents find out about rural practice through rural moonlighting opportunities. Lately residents have passed up rural opportunities to stay in high-paying urban urgent care centers. Rural faculty have encouraged networks of rural hospitals and physicians to pay residents a regular additional salary with the promise that the residents will serve network hospital for two to four weeks in their second or third year. This brings back the potential for recruitment, gets residents out of the cities, broadens their training, and helps residents develop the confidence to be out on their own, a prerequisite for those considering rural practice.

For general internal medicine, general pediatrics, obstetrics and gynecology, and general surgery, there are a number of RRC requirements that preclude having a residency located in a small, rural hospital, though it seems that the surgery RRC was willing to be flexible about some of these requirements. Some of the major requirements that present barriers to locating a residency for one of these specialties in a rural hospital are summarized in Table 2.

For internal medicine, for example, the sponsoring hospital must sponsor at least 1 other program and be a significant training site for residents from still another program - difficult to impossible for a small, rural hospital. Pediatrics training requires a pediatrics ICU in the sponsoring hospital and 4-6 months of level II or III neonatal intensive care unit experience, despite the fact that many pediatricians do not utilize such settings in their practices. There is also a requirement for 4 months of subspecialty training that would be difficult to obtain in a rural area. For obstetrics and gynecology, the barriers are not so obvious, as explicit definitions of adequate numbers of procedures are absent, but to have at least 2 residents/year in each of the 4 years of training would require a hospital with substantial obstetrical volume, beyond what most small rural hospitals could manage. For general surgery, the requirements for procedural volume would preclude having a small, rural hospital as the sponsoring or integrated institution, though having the final year at the University program apparently satisfied the RRC about the South Alabama proposed rural general surgery program.

Given the low level of activity related to rural training in specialties other than family medicine, there is relatively little to report on in terms of financial barriers. However, one program in general surgery was designed by the University of South Alabama specifically to prepare general surgeons for rural practice and was approved by the RRC for surgery and the American Board of Surgery, along with being endorsed by the American College of Surgeons. In this program, residents would have spent their third and fourth years in several community hospitals, including some rural hospitals of 100-200 beds, and then returned to the parent program at the University for their final or chief resident year. However, the parent hospital has a small Medicare base (and a large proportion of patients with Medicaid or no insurance) and could not afford to support the residents at the community hospitals due to their limited GME payments. None of the community hospitals had either the expertise or the resources to apply for Medicare GME funding while the residents would have been there, and because of this lack of funding the program never got underway (J. Raymond Fletcher, M.D., University of South Alabama College of Medicine, Department of Surgery, personal communication, 10/24/97).

While the University of Washington Internal Medicine program has a relatively large program offering rural electives to residents, funding is felt to be the biggest problem limiting their ability to place residents in rural rotations. Funding to pay for transportation of residents to and from the distant, rural sites and their housing costs while they are at these sites has limited the number of sites used and rotations offered. This

funding is becoming harder to get. Other identified needs that would strengthen and expand this program are funds for administrative support, for the development of a defined curriculum with learning objectives, and to allow for electronic links to the university and Internet so that faculty or residents could obtain clinical information, curricular materials, Medline, and other resources (Dawn DeWitt, M.D., personal communication, 10/23/97).

The RRC for family medicine has approved a number of RTTs, starting with those associated with the Spokane, WA residency about 10 years ago. Initially they allowed significant flexibility that facilitated the creation of a number of RTTs. Recently, interviewees noted that some requirements have been issued for RTTs could result in major difficulties for existing sites, as well as preventing the institution of new sites. Often there was little evidence of the need for these requirements.  1-2 Rural TT.pdf

One commonly-cited new requirement was that mentioned above for a preceptor to have at least half of their clinic schedule to be freed up to supervise one resident. This regulation did not allow preceptors to see clinic patients when 2 or more residents were seeing patients in clinic.

Another requirement will force sites to have at least 2 residents (i.e., at least 1 per year). The Spokane, WA program has had 2 RTT sites in small, remote rural locations that have been able to support 1 resident at a time for 2 years (1 new resident every other year), but cannot support 2 residents being present simultaneously. There is no evidence that such residents are not as well trained or that they have less collegiality. Indeed many residents at larger training programs feel that faculty treat them as colleagues. In rural practice, there is no barrier to this relationship.

Interviewees at both RTTs and 3-year rural family medicine residencies commented on several other problems related to the RRC. Residents should have experiences in a variety of fields. The RRC clearly prefers this to occur in dedicated blocks of time. In an isolated, rural residency, obtaining such rotations requires the residents to spend months away from the program. This is a strain on the residents, makes it difficult to attract students to the program, and threatens to make the programs fall out of other RRC regulations, such as compliance with requirements for continuity of care by residents. Several persons felt that there needed to be greater flexibility by the RRC on rotational versus longitudinal exposure to some experiences. They hoped that the RRC would be more open to experimentation and innovation.

Many feel that block rotations inhibit the ambulatory training of the family practice residents and in programs with few ambulatory clinic days, this style of training may inhibit continuity with patients more than being away from the main site for rural or other rotations. Longitudinal training centralizes training around the ambulatory patients, where 90% of future family physician time will be spent. Longitudinal training de-emphasizes dependence on hospitals and this may be important. Many rural hospitals have closed and others will likely close or convert to ambulatory centers in the next years. Family medicine training should anticipate these changes and prepare residents that will not necessarily require a hospital at their practice location.

RRCs in other specialties have anticipated future changes well. Obstetrics training now requires 4 months of ambulatory clinic, often done at family medicine residency programs. Interestingly during this four months of full time clinic, most first year obstetrics residents will see as much ambulatory care as family medicine residents do in the first year and a half of training since family medicine residents only spend a small fraction of their time in clinic with the first years often in low volume situations.

Formal medical education still does not accept the preceptorship style of training. One hundred years ago with less formal medical training and lower standards, these questions were relevant. Now even the most remote rural sites have well-trained board certified family physicians with years of medical school, residency, and post graduate training. Clearly there is a choice to have several family practice faculty train and influence one resident longitudinally in the smaller programs, or a continuation of many residents and students trained by a recurrent series of discontinuous faculty of multiple specialty types in academic centers. There are strengths and weaknesses of both methods, but no clear preference. The mix of both methods provided by the RTT model might be used to emphasize the strengths of both.

The difficulties of recruiting, retaining, and training faculty for rural residencies were commonly brought up during our interviews. As discussed above, RTTs tend to be based in existing practices, with the members of these practices becoming the primary teaching faculty for the RTT. The initial recruitment effort for such programs is therefore the process of carefully selecting the right physicians in the right practice in the right town.

Larger programs have faculty replacements available nearby. One program noted that 80% of the staff at their hospital were graduates of the program, as were many of the physicians in surrounding communities who referred their patients to the hospital.

A recurring theme in the interviews was the need for faculty development, both at RTTs and 3-year residencies. Residencies in university settings or affiliated with nearby universities may have access to local resources for faculty development. For residencies in rural areas, there are no such local resources, and faculty development efforts generally require the faculty to travel to a distant location and take more time away from their programs than might be necessary in an urban location with local resources. Funding for faculty development was another case in which the importance of Title VII funds was cited. Some feel that funding that goes primarily to large academic centers is of little benefit to those developing or teaching in rural programs.

Only one Bureau of Medicine grant has targeted this specific area. This Rural Minifellowship program trained 5 faculty a year from 1990 to 1992, but it has continued as the Group on Rural Health of the Society of Teachers of Family Medicine. Materials, workshops, and newsletters initiated with this Rural Minifellowship program continue to assist faculty with rural training track, rural rotation, and rural satellite training center development at a volume of about 3 - 4 new contacts a month. Minifellowship update This "Center for Rural Faculty Development" continues because of the voluntary efforts of rural faculty and the willingness of a handful of medical schools and residency programs to allow faculty to spend more than 25% of their time on rural medical education.

Indeed most rural medical education programs have been created by former rural physicians either on a voluntary basis or by using their position as program director. Few have had fellowship or other educational training. Surveys of rural faculty revealed that only a handful have had more than 50% of their time invested in rural medical education. Physicians chose rural practice to some degree based on their love of clinical medicine and doing procedures, and many take up these challenges when teaching in training programs rather than rural research or developing new rural programs. Rural faculty development can help prevent the repetition of common mistakes, pass on critical information on funding and the selection of sites and faculty, provide the peer support needed for rural faculty who are often alone in their institutions, and facilitate the necessary funding efforts to make rural programs a reality.

Recruitment of residents was felt to be difficult by many, but not all, of those we interviewed. In general, they felt that their applicant pools were smaller than for non-rural residencies (often around 4-5 applicants for a position, versus over 60 per position at the University of Washington s family medicine residency), but that the people who applied were really interested in rural practice. In some cases, the pool of applicants applying to RTTs was noted to have little overlap with the pool applying to the parent programs. One program director remarked that being able to pay a higher resident salary than in urban areas would help with recruitment but that there were no funds to allow this. Some programs had initially filled their slots with international medical graduates, some of whom intended to stay and practice in a rural area, and some of whom intended to return to their countries of origin, but found themselves getting more U.S. medical school graduates as the programs matured. Improved match rates over time are common in both types of new programs, RTT and 3 year.

A potential problem mentioned by a number of RTT personnel was that many sites were interested in having an RTT primarily with the hope that graduates would remain there. While there is nothing inherently wrong with this, if the site needs more physicians, the danger is that, once there are enough physicians, the site might withdraw, requiring an expensive and time-consuming process to develop a new site. Most program directors felt that they had been careful enough when evaluating their potential RTT sites to avoid this, but success in filling a site that was short of physicians is one potential benefit of rural training. In some ways the nation could afford more rural training on the part of residents both to support rural physicians in practice, provide more direct rural care, and graduate more rural physicians.

It seems clear that, in family medicine, programs with a strong rural orientation, whether RTTs or 3-year residencies located in rural areas, can train physicians who are more likely to set up practice in a rural area. It is less clear for other specialties because the numbers are small. Whether these programs are primarily keeping the converted from being diverted is unknown.

A number of barriers to training residents in rural locations were identified. Not surprisingly, a large number of these were financial. The costs of the various rural programs depend on multiple factors that can make them more or less expensive. However, the way GME payments are structured and administered results in significant barriers. Residency Review Committee requirements also posed substantial barriers, particularly outside of family medicine. The recruitment of faculty and residents and the need for faculty development was a challenge for rural programs. Also, many RTT programs depended on grant funding of telemedicine that might not be able to continue without continued local or federal support.

This paper has focused on family medicine graduate medical education because that is where most of the rural graduate medical education efforts have occurred. This is not entirely by chance since, particularly for small, remote, rural areas, a family practitioner can be supported (needs 2000 population) but a pediatrician or general internist cannot (needs 5000). For larger rural areas, the population is large enough to support other providers, such as internists, pediatricians, OB-Gyns, general surgeons, and psychiatrists, and there is clearly a need for programs to train physicians for practice in such areas.

A number of recommendations for changing medical education were made by Dr. Robert Talley, dean of the University of South Dakota School of Medicine, in 1990 (Talley 1990). Among these recommendations were:

1) to develop a consensus definition of rural and subcategories of rural;

2) to educate rural communities about the purpose of residency training, the need for attending physicians to supervise residents, and opportunities to help fund rural rotations for residents;

3) the family medicine RRC should support rural rotations of up to 6 months and more broadly define what a family practice center is to allow inclusion of outreach sites;

4) RRCs should not make the number of residents at a site a factor in approving it;

5) RRCs should judge programs by the quality of their products, not process measures;

6) resident teaching should be defined by competence, not by specialty label; and

7) new opportunities for funding residency training in rural settings should be explored.

Dr. Paul Young, executive director of the American Board of Family Practice, wrote a commentary accompanying Dr. Talley s article, agreeing with many of his points, but not with his proposal to allow rural rotations of up to 6 months duration due to its impact on continuity of care, and argues that the number of residents training at a site may, in some circumstances materially contribute to the quality of the educational experience (Young 1990). Many of our recommendations below are quite similar to Dr. Talley s recommendations, as little seems to have changed since he wrote them.