Lisa M. Bellini, MD and David A. Asch, MD, MBA

Department of Medicine, University of Pennsylvania School of Medicine and Veterans Affairs Medical Center, Philadelphia, Pennsylvania

Abstract: Expanding primary care and ambulatory experiences in internal medicine training programs is limited by insufficient resources devoted to their development and implementation, heavy inpatient service demands and loyalty to the traditional inpatient based training model. Overcoming these barriers is a challenge likely to create new approaches to ambulatory education. The Pilot Education and Ambulatory Care (PACE) program at the Sepulveda VA is one such initiative that represents a multidisciplinary approach to primary care services, improving quality and access to patient care services, as well as the quality of primary care education. We describe the development of a similar model for resident training and primary care service, also funded through the Veterans Affairs (VA), but implemented with fewer resources. The program redirects faculty and administrative resources to a new clinical environment centered around primary care. We believe this educational model has broad implications for the future of medical training because it is achievable and thus exportable to other programs.

The economic and social pressures driving health care reform have generated concerns about the excess of some medical subspecialists and the relative shortage of primary care physicians. The Accreditation Council of Graduate Medical Education (ACGME) has recognized these concerns and mandated that residents spend 25% of their time each year in ambulatory settings.¹ Those responsible for graduate medical education are now searching for ways to increase ambulatory and primary care experiences in internal medicine training programs. Tradition, inertia, lack of sufficient resources and often heavy inpatient service demands on residents limit the changes that can be implemented in these residency programs. As a result, program directors face great challenges in trying to reshape their programs to meet both the health care needs of society and the educational needs of their trainees.

The disparity in trainee resources between inpatient and ambulatory care was most apparent at Veterans hospitals with traditionally limited support of outpatient services, particularly primary care services. The Department of Veterans Affairs launched two major initiatives to address this issue. The first initiative was in October of 1989 when the Veterans Health Administration Western Region and associated medical schools developed a set of recommendations to shift the balance of medical education from the traditional inpatient model to one that focused on ambulatory and primary care.² From these recommendations, the Pilot Ambulatory Care and Education (PACE) program at the Sepulveda VA was established. Accompanied by specific interventions directed at continuity, prevention, mental health care and humanistic care, the PACE program increased primary care capacity and workload, as well as significantly improved the process and outcomes of ambulatory care.^3-5 The resources required to develop, implement and evaluate PACE were tremendous and partially funded by Veterans Health Administration Western Region.

The second initiative, referred to as the Primary Medical Education (PRIME) program, was launched by the Office of Academic Affairs in the fall of 1993. The PRIME program sought to support the development of new multidisciplinary programs in primary care education through the funding of trainee stipends in a variety of health related professions, including residents training in internal medicine, family practice, podiatry and psychiatry, as well as nurse practitioner, physician assistant, social work and pharmacy trainees. Unlike the funding for PACE, PRIME grants did not provide support for the development, implementation or evaluation of the proposed programs. The purpose of this paper is to describe the development and implementation of the primary care program we developed through PRIME. We hope that by sharing our successes and frustrations we can help others develop similar programs.

The Philadelphia PRIME Program

The Philadelphia PRIME Program used trainee stipends (14 medical residents, six psychiatry residents, six nurse practitioners, a pharmacist, and two social workers) to develop a primary care program in two categorical internal medicine residency programs that shared the Philadelphia Veterans Affairs Medical Center (PVAMC). The PVAMC is a 260 bed acute care hospital that had primary academic affiliations with both the University of Pennsylvania and the Medical College of Pennsylvania (now Allegheny University Hospital, East Falls).

The program incorporated three educational assumptions. The first assumption was that if a subspecialty-oriented, inpatient-based training program was to reshape itself into a primary care program, it had to embrace major changes in the organization and setting of medical training.^6-7 We focused our programmatic resources on a group of trainees interested in primary care and created a model that essentially fused inpatient and outpatient care. The second assumption was that the provision of primary care requires a different base of knowledge and set of skills than does inpatient, subspecialty care.^8-10 In addition to providing the fund of knowledge developed through traditional training, the curriculum focused on developing the interpersonal skills necessary to work effectively with an assortment of health care professionals as well as the development of practice and management styles around the patient, rather than around an established inpatient or outpatient setting. The third assumption was that in order to support a primary care training program at an institution heavily represented by subspecialists, the program would need to support the retraining of established clinicians and administrators who would need to practice in a new clinical environment centered around primary care.^7,11-12

The program consisted of two firms or interdisciplinary health care teams that had both inpatient and outpatient responsibilities. The members of each firm included 18 internal medicine house officers from each affiliate school (six from each post graduate year level), two psychiatry trainees (one PGY-1 and one PGY-3), two nurse practitioner trainees, two social work trainees, two podiatry trainees and two pharmacy trainees. Faculty preceptors from each service were added to each firm. In addition, a shift in the clerical support from within the PVAMC enabled the dedication of three clerical personnel to each firm.

Medical Residents

Internal medicine residents from both university affiliates rotated through the VA hospital and spent approximately one third of their training at this institution. Prior to the implementation of this program, resident experiences at the VA were in structured in a traditional format. The vast majority of patient care and education occurred on the inpatient services, while outpatient experiences were confined to a weekly continuity practice based at either the VA or at the affiliate. Although outpatient practices were intended to educate residents in the basic principles of primary care, weekly continuity practices in the traditional program accounted for at most 10% of a residents' total clinical experience.

Both affiliates preserved the traditional inpatient experience for non-PRIME residents rotating through the VA. The implementation of the PRIME program created four inpatient services - a traditional ward service and a PRIME service for each affiliate. Each inpatient PRIME service consisted of two internal medicine residents, three interns and other team members to be described later. The housestaff took call every fourth night. The inpatient responsibilities were distributed such that each PGY-1 resident had four inpatient ward months on the PRIME service, and each PGY-2 and PGY-3 resident had two inpatient ward months on the PRIME service. A feature unique to the PRIME program was that housestaff on the inpatient PRIME service spent roughly half of their time in the outpatient setting. In addition to a weekly continuity clinic that PRIME residents attended regardless of the site of their current clinical rotation, they also attended 3-4 morning clinic sessions a week while they were on the inpatient PRIME service. The morning clinic provided them the opportunity to evaluate new primary care referrals and patients discharged from the PRIME inpatient service. Supervision was provided jointly by general medicine fellows and subspecialty faculty. Patients evaluated during the morning sessions were subsequently enrolled in the residents' weekly afternoon continuity practice for ongoing care. Residents on the inpatient service could admit patients to themselves from either their morning or afternoon clinic sessions.

The increase in time spent in the ambulatory setting created a simultaneous reduction in the amount of time residents were available for inpatient activities, including direct care of inpatients, attending/teaching rounds and attendance at conferences. This tradeoff represented a challenging but common obstacle to any program considering significant increases in ambulatory activity. We had some success in meeting this challenge by restructuring the daily routine as shown in Figure 1..  A typical day on the inpatient service began with resident teaching/work rounds that lasted an hour, adequate time to see a maximum of 18 patients. Four mornings a week, all but one of the house officers then moved to the outpatient clinics. The house officer remaining on the inpatient ward provided coverage for the inpatient service. A conference block from noon to 2:00 P.M. separated morning and afternoon activities.

Figure 1
Daily Routine

A major departure from the traditional model was the shift of attending rounds to the afternoons. This shift caused a natural dissociation between the patient management and teaching components of attending rounds. Residents generated patient care plans during their morning work rounds and confirmed these plans with the attending by phone, typically by 9:30 A.M. The resident responsible for the inpatient service that morning carried out the plans resulting in discharges and requests for consultations occurring earlier in the day. The unanticipated benefits that occurred with the shift in attending rounds included more organized clinical approaches by residents' resulting from more time to think independently, and more time for medical students to gather and synthesize data in preparation for their presentations to the attending. Following attending rounds, the remainder of the day is spent on inpatient activities.

In designing the PRIME program, workload expectations were based on our experiences both at the VA and the affiliates. We derived the workload for the inpatient PRIME service by determining the average daily patient census on a medicine ward service and then adjusting that number to account for the increase in outpatient activities. The year prior to implementing this program, the average intern census was between 8 and 12 patients. Given that housestaff on the inpatient PRIME service were required to spend fifty percent of their time in the outpatient setting, we set the maximum census at 6 patients per intern, or 18 patients per PRIME team. Once their maximum census was reached, the PRIME service closed and the traditional services continued to admit patients. The census at the PVAMC was never large enough to exceed the capacity of the two traditional services. Similarly, taking into account the greater severity of illness of the VA patient population - the average patient at the PVAMC had 3.5 visits per year to the general medicine practices - as well as the enhanced administrative support provided by dedicated clerical staff, each resident was expected to accrue 75 patients in their clinic. A typical practice profile would include two new patients and four to six follow-up patients, depending on the level of training.

Faculty

The implementation of PRIME greatly increased our needs for general medicine teaching and supervision. Prior to the implementation of this model, both general medicine and subspecialty faculty served as attendings of record as well as teaching attendings for the inpatient services. With the development of PRIME, we expected general internists, and not subspecialists, to staff the inpatient PRIME services.

Three separate forces drove us to use internal medicine subspecialists as preceptors in the PRIME morning clinic sessions. First, the increase in the amount of supervision required by the outpatient emphasis of the PRIME program could no longer be handled by the available general internists alone. Second, as the shift of resident education and patient care moved to the outpatient setting, internal medicine subspecialty attendings risked being separated from resident training. Third, there was an increasing recognition that the primary care skills of many subspecialty attendings had declined and their presence in the outpatient resident practices would provide opportunities for regaining those skills.

Many subspecialists initially resented and resisted the idea of "retraining". Some of their concerns included a lack of confidence about providing primary care and supervising residents in that endeavor, as well as the amount of time they would need to devote to such an activity. The former concern was addressed by pairing subspecialty attendings with general internal medicine fellows during the morning clinic sessions. Perhaps because of their socialization and interests, the fellows provided needed support in this setting. In addition, it seems as if the attendings were more comfortable learning from fellows than from general medicine faculty. The latter concern was addressed by the Departments of Medicine, which supported the program.

The combination of subspecialty attendings, general internal medicine fellows and internal medicine residents resulted in additional benefits. Residents and fellows had more opportunities to learn subspecialty approaches to common problems arising in primary care settings. The presence of different subspecialists in the morning clinics enabled the residents to reduce their total number of formal subspecialty consults by saving their subspecialty management issues for the appropriate subspecialty attending. These changes helped residents develop confidence in managing these problems independently in primary care settings. Similarly, several of the subspecialty attendings admitted new respect for their generalist colleagues.

Despite these successes, the subspecialty faculty unanimously felt that, on balance, their teaching should be confined to their specialty and not include general internal medicine. Furthermore, most relayed that they did not have the time or the interest to relearn general internal medicine. Thus subspecialty faculty were replaced with general medicine faculty, with the continued presence of general medicine fellows. Changes in professional interests or philosophies, or market forces, may allow us to try this model again in the future.

In addition to the medicine trainees, other members of the interdisciplinary health care teams included psychiatrists, nursing, clerical personnel, social workers, and pharmacists. Psychiatry

Two psychiatry interns participated in the inpatient component, and a senior psychiatry resident supported the weekly afternoon continuity practices. An important benefit of the collaboration and proximity of psychiatry faculty and trainees to internal medicine trainees in the outpatient setting was the natural facility with which patients with psychiatric issues were evaluated and managed. This collaboration taught the internal medicine residents the appropriate evaluation, management, and follow-up of outpatient psychiatric problems in a supervised setting, and enabled the psychiatry residents continued exposure to internal medicine.

Nursing

PRIME residents on rotation at the affiliates returned weekly to their afternoon continuity practices unless they were on exempt rotations such as the intensive care unit or the emergency department. Even so, they were not available for patients who needed more frequent or interim care. Each firm had a nurse practitioner who received referrals for patients requiring frequent visits at shorter intervals (for example insulin dose adjustments), in addition to scheduled or unscheduled visits when the resident was unavailable. When necessary, the nurse practitioner communicated with the off-site resident and they jointly developed a management plan.

Clerical Support

We felt it necessary to reproduce the smaller and more agile structure of an office-based practice so that providing continuity was seen as a goal for clerical as well as clinical staff. A total of six clerical personnel were assigned to PRIME with each assuming responsibility for several medicine resident and faculty practices. Over time, patients developed a relationship with their clerical person, just as they did with their nurses or physicians.

Social Services

Each Firm had a dedicated social worker to help clinicians recognize and manage social problems such as general issues of disposition, transportation difficulty, unemployment, unacceptable living situations, and homelessness. On admission to the PRIME service patients were evaluated by the social worker for discharge needs, preventing any socially related delay in discharge. In addition, the social worker made rounds with the team two days a week to keep abreast of discharge planning. The social workers evaluated the ongoing social needs of the patient after discharge during outpatient appointments, typically on the same day that post discharge follow-up occurred. In addition, they participated in the weekly continuity practice sessions to assist housestaff in the early identification of social issues that interfered with the patient's ability to comply with treatment plans.

Clinical Pharmacy

The clinical pharmacist had a pivotal role because they saw all patients who presented for medication refills, reviewed the medication profiles of each residents' patients prior to the continuity clinic sessions to identify potential adverse drug reactions and polypharmacy, and provided patient education. The presence of the pharmacist enabled the walk-in clinic nurse practitioners to focus on those patients needing urgent clinical care. This intervention alone resulted in substantial reduction in pharmacy costs, through the identification of potential adverse drug reactions and duplicate prescriptions within medication classes, which offset the salary of this position.

Conclusions

The trainee stipends available to the PRIME program were used to create two firms based on the affiliate school of the medical residents that were integrated across the inpatient and outpatient settings. The simultaneous presence of medical resident trainees in the inpatient and outpatient settings and concentration of nurse practitioner, social work and pharmacist trainees in the outpatient setting, improved resources available to outpatients thereby avoiding unnecessary admissions and also addressed the ubiquitous problem in resident-based continuity practices of residents being unavailable when patients need frequent or interim care. Improved outpatient care translated into a 20% decrease in admissions from the general internal medicine practices and for those that were admitted, the provision of coordinated inpatient and outpatient care translated into a decrease in the average length of stay from 9 days for the traditional medicine services to 5 days for PRIME services (personal comm. T Barrett, MD).

We do not believe that the provision of trainee stipends is necessary for the successful development and implementation of this type of model. A similar program could be developed through collaboration with other disciplines already involved in training psychiatrists, nurse practitioners, pharmacists, podiatrists and social workers. Those disciplines without trainees could be similarly incorporated. The multidisciplinary team approach to patient care could be developed through the redirection of available resources - nurse practitioners, clerical support and social workers - in housestaff continuity practices. More necessary for the successful implementation of this model than trainee stipends was support for the increased supervision and teaching effort required for trainees in the outpatient setting, all of which were carved out of existing faculty and staff resources. In addition, no local support was provided for the evaluation of this model, the lack of which we consider a major weakness.

The Veterans Health Administration has supported initiatives to expand primary care and ambulatory experiences in internal medicine training programs that have been limited by heavy inpatient service demands and historical commitment to the traditional inpatient-based training model. Programs like PACE have improved quality and access to patient care services, as well as the quality of primary care education, but also required substantial financial support that may be unavailable at other institutions. With relatively few additional resources, we managed to achieve many of the goals of the PACE program. Although many elements of our program are affected by local circumstances, we believe many other elements of our program, and our experience developing it, are exportable to others with similar goals.

References

1. 1989-1990 Directory of Graduate medical education programs. Chicago, IL: American Medical Association, 1989.
2. Robbins AS, Guze PA (eds). Proceedings from the Conference on Ambulatory Care and Education, July 1988, Los Angeles, California. Acad Med. 1989;64(10 suppl).
3. Cope DW, Sherman S, Robbins A. Restructuring VA ambulatory care and medical education: the PACE model of primary care. Acad Med. 1996;71:761-771.
4. Rubenstein LV, Yano EM, Fink A, et al. Evaluation of the VA's pilot program in institutional reorganization toward primary and ambulatory care: part I: changes in process and outcomes of care. Acad Med. 1996;71:772-783.
5. Rubenstein LV, Lammers J, Yano EM, Tabbarah M, Robbins AS. Evaluation of the VA's pilot program in institutional reorganization toward primary and ambulatory care: part II: a study of organizational stress and dynamics. Acad Med. 1996;71:784-792.
6. Hayashi SA, Hayden BB, Yager J, Guze PA. Graduate education in ambulatory care. Proceedings from the Conference on Ambulatory Care and Education. July 1988, Los Angeles, California. Acad Med. 1989;64(10 suppl): S16-S21.
7. Wall EM, Saultz JW. Retraining the subspecialist for a primary care career: four possible pathways. Acad Med. 1994;69:261-266.
8. Barker LR. Curriculum for ambulatory care training in medical residency: rationale, attitudes, and generic proficiencies. J Gen Intern Med. 1990;5(1 suppl):S3-S14.
9. Lawrence RS. The goals of medical education in the ambulatory setting. J Gen Intern Med. 1988;3(2 suppl):S15-S25.
10. Robbins AS, Cope DW, Campbell L, Vivell S. Expert ratings of primary care goals and objectives. J Gen Intern Med. 1995;10:429-435.
11. Rich EC, Wilson M, Midtling J, Showstack J. Preparing generalist physicians: the organizational and policy context. J Gen Int Med. 1994;9(4 suppl):S115-S122.
12. Nelson R. Retraining physicians for primary care. Iowa Med.1995;85:175.

Dr. Asch received partial support for developing this paper as a recipient of Department of Veterans Affairs Health Services Research and Development Career Development Award.

Address questions and comments to:

Lisa M. Bellini, M.D.
Department of Medicine
Hospital of the University of Pennsylvania
3400 Spruce Street
Philadelphia, PA 19104

Telephone: (215) 662-3224
Fax:(215) 662-7919
Bellini@mail.med.upenn.edu

Bellini LM, Asch DA. The Philadelphia PRIME Program: A Model For Primary Care Education Med Educ Online [serial online] 1997;2,2. Available from: URL http://www.Med-Ed-Online.org.