Physician Staffing for the VA: Volume I (1991)

The Establishment of Affiliations Between Vamcs and Medical Schools Has Been Pursued As an Policy Objective of the VA Since 1946. the Original Purpose of These Affiliations Was To Ensure That the VA Health Care System Was Staffed with Competent Physicians Who Maintained Currency with Advances in Medicine Through Their Medical School Relationships. These Affiliations, Which Became Formal Agreements in 1973, Grew To Involve, in Varying Ways, 134 of the Current 172 Vamcs. the Affiliations Have Enhanced the Physician Presence in the Vamcs in Various Ways: VAEmployment (Part-Time Or Full-Time) of Medical School Faculty Appointees, Placement of Residents in Vamcs, and Access To Consulting Physicians On an Paid Or Unpaid Basis. Thus, Affiliation Arrangements Are Important in the Physician Staffing For Most Vamcs. Quite Logically, the Scope of This Study Provided For Specific Attention To Be Given To Affiliations As They Affect Physician Staffing Requirements and the Patient Care Provided To the Veterans.

From Their Beginning in 1946, the Affiliations Involved the VAin the Teaching of an New Generation of Physicians (As Well As the Training of Other Health Professionals). Because Medical School Faculties Are Principal Performers of Medical Research, Faculty Status For VAStaff Also Increases the VAInvolvement in Biomedical Research. Thus, the Affiliation Agreements Have Facilitated the Incorporation of Teaching and Research Functions As Formal Components of the Mission of the VAHealth Care System.

To Reflect the Integral Part That Affiliations Play in the Physician Staffing For Most Vamcs, the Original Planning Study, Which Generated the Scope of Work For This Study, Gave Important Attention To Medical School Affiliations. the Work Plan Developed By the Planning Study, in Noting the Importance of the VAEducation and Training Mission Within the VAHealth Care System, Stated:

The first of these analyses—examining the staffing implications of the teaching function inherent in the affiliation arrangements—has been an integral part of the overall methodology described in chapters 4 through 6. This chapter looks more broadly at the implications of the affiliation agreements for physician staffing requirements and also at issues of cost and quality of care received by veterans.

Affiliations are pervasive in the current VA health care system, and 45 years of VA policy have supported their development as an important way to obtain the physician staffing necessary for high-quality patient care. In addition, the current VA mission closely intertwines teaching and research activities with the affiliations. Thus, careful attention should be paid to the bases for retaining, strengthening, modifying, or reducing affiliations.

It is clear to the committee that a VA health care system without affiliations would be a very different system, and that the teaching and research aspects of the VA mission as legislatively authorized would be very difficult, if not impossible, to carry out without affiliations. Yet, the continued existence of unaffiliated VAMCs and the wide range of affiliations provides opportunities for comparison of effects. At the same time, these differences create a potentially unstable situation. The committee has heard that both within and outside the VA, questions have been raised about the value of affiliations and their effects on costs of VA medical care. Medical schools are not clamoring for new affiliations, and some schools have questioned the viability of current arrangements. The following discussion lays out the issues as the committee sees them and examines available evidence to provide support for the committee's conclusions and recommendations.

In summary, the questions examined in this chapter include the following:

• What are the direct effects of affiliation agreements on physician requirements? Are more or fewer VA staff physicians required to meet patient care needs because of the affiliations?

• What is the net effect of affiliations on the costs of meeting physician requirements? Is there an effect of affiliations on overall patient care costs?

• What are the benefits to patient care of affiliations?-Is there better access to state-of-the-art tertiary care?

  -Is there better access to highly qualified physicians across the whole spectrum of health services, including recruitment and retention of VA physicians?

  -What is the general effect of links to teaching and research on quality of care?

• Do the affiliations create any problems for meeting the full range of patient care needs of the veterans?-Because the affiliated institutions are mostly tertiary care facilities, do affiliations serve the primary care, rehabilitation, and chronic care (including psychiatric) needs of the population served by the VA?

  -By depending on residents for substantial amounts of care, are such desirable characteristics as continuity of care compromised?

• The current VA mission (in addition to the care of veterans) includes research, education, and backup to the Department of Defense in time of war. If the affiliations are critical to these aspects of the VA mission, could these missions be sustained without affiliations? Would a modification of the mission be required if affiliations are weakened?

• If affiliations provide net benefit for the VA patient care mission, what should be the policy toward lack of affiliation in some institutions? Can affiliations be designed to benefit all aspects of VA care responsibilities—primary care, rehabilitation, and care of the chronically ill including psychiatric and long-term care needs?

The study's work plan called for the appointment of a panel on affiliations to assist the committee in its consideration of these issues. That panel, consisting of a majority of non-VA members and a minority of VA physicians, met five times and participated in three sets of site visits to 15 VAMCs. A mail survey of 24 affiliated and unaffiliated VAMCs was also conducted.

The panel made very important contributions to the deliberations of this committee on the issues covered in this chapter. The full report of the panel is included in Volume II, Supplementary Papers . The committee has given careful consideration to the panel's work in reaching its conclusions and has also had the

benefit of some additional analyses prepared by the study staff and the VA. However, the final conclusions are the responsibility of the committee.

Although the committee has attempted to marshall all available supportive evidence, these conclusions and recommendations rest substantially on the committee's judgment based on experience and logic, in addition to data. Much of the supportive data were suggestive rather than definitive, and data on crucial matters, such as the effects of specific aspects of the affiliations on patient care outcomes, are simply not available at this time. However, the committee clearly indicates when its conclusions are based substantially on its judgment rather than on data.

The VA was established as an independent federal agency in 1930, combining the Bureau of Pensions (formed in 1833), the National Home for Disabled Volunteer Soldiers (1866), and the U.S. Veterans Bureau (1921). On January 3, 1946, President Truman signed Public Law 79-293, establishing the VA Department of Medicine and Surgery. Shortly thereafter, on January 30, 1946, the chief medical director of the VA published the second policy memorandum on the association of veterans hospitals and medical schools (Worthen, 1987).

Following World War II, members of the medical community became alarmed at the number of returning servicemen filling VA hospital beds, particularly since civil service red tape and the bad reputation of VA medicine had caused a critical shortage of doctors within the system. The creation at that time of a separate personnel system for physicians and nurses employed by the VA helped to circumvent some of the bureaucratic delays in hiring physicians. The idea of having the ranking medical schools as medical affiliates of VA hospitals was implemented largely through the efforts of Dr. Paul Magnuson, an orthopedic surgeon from Northwestern University. Well known to many deans and university professors in medical schools across the country, Dr. Magnuson drafted a plan to have medical school deans supply the staff for the VA's hospitals, and determined that he could have the first two such affiliation agreements operational within six weeks.

At that time, the VA had 83,339 beds in 98 hospitals, many in remote locations. None of the hospitals had accredited residency programs; the 1,000-bed VA hospital in Pale Alto had only five doctors. The day after the signing of Public Law 79-293, 56 medical residents were placed at Hines General (VA) Hospital in Chicago, by Northwestern University and the University of Illinois. During the next three weeks, the University of Minnesota placed 26 residents at Fort Snelling; over the following months, Dr. Magnuson went to Boston, New

York, San Francisco, and other major cities to enlist the aid of medical school deans in establishing affiliations. As outlined in Policy Memorandum No. 2:

This memorandum further states that ''the Veterans' Administration retains full responsibility for the care of patients, including professional treatment, and the school of medicine accepts responsibility for all graduate education and training.''

By 1950, the number of VA hospitals had increased to 151 with 117,000 beds. During the 1950s and 1960s, the total number of hospitals continued to increase; many outdated facilities were replaced, and new affiliations were established. A new policy in the late 1960s directed that new VA hospitals be built on or near campuses of affiliated medical schools. The VA Medical School Assistance and Health Manpower Training Act of 1972 provided for grants to assist in the establishment of new state medical schools that would be affiliated with VAMCs. The act further provided funds for medical schools already affiliated with VA hospitals, to enable them to significantly expand their collective class size and, in several cases, their curricula.

The cumulative effect of these actions is that 134 of the 172 VAMCs have some form of affiliation agreement with 102 of the 127 U.S. medical schools. These agreements represent a wide range with regard to scope and intensity of affiliation. Several of the facilities included among the "affiliated" group indicate only undergraduate medical student training rather than the presence of residents; a number of others list several staff physicians who have faculty appointments at the affiliated medical school, but apparently train neither students nor residents within the VA hospital itself. At the other end of the continuum are the large urban tertiary care VAMCs, which typically train 100 to 150 residents in many different specialties, as well as large numbers of medical students in clerkships, and trainees in many other health professions. Currently, there are approximately 80 VAMCs (mostly in RAM Groups 3 and 5, as defined in chapter 4) with substantial affiliations training residents in a number of different specialties.

Affiliations encourage the more extensive involvement of VA physicians in research and teaching activities. As shown in Table 9.1, the percentage of physician time devoted to direct patient care activities is typically higher in nonaffiliated VAMCs (RAM Groups 2, 4, and 6). Tables 9.2 and 9.3 indicate that much of the rest of the physicians' time is devoted to education and research.

The time that VA physicians spend in teaching rather than direct patient care is offset by the presence of residents and other physicians in training. In the PF models described in chapter 4, when both the staff physicians and the residents are represented in the model, the estimated coefficients indicate that (all else equal) residents can substitute for staff physicians in achieving any given output level—that is, fewer staff physicians are required to deliver the same volume of care when residents are present.

In addition, the affiliation agreements enable the VAMC to supplement its physician staffing with faculty from the medical school who provide services under the classifications of "consulting & attendings" (C&As) and "without-compensation" (WOC). C&A physicians earn a fiat fee of $40 or $75 per consultation, depending on seniority and academic rank, regardless of the duration of their visit to the VAMC. WOC physicians also provide patient care and teaching/supervision services to VAMCs for which they are not compensated at all. Data on the time spent and services contributed to VAMCs in this manner are not kept systematically, either at the facilities or in any centralized data base. In order to estimate the magnitude of physician effort that these arrangements represent to the VA system, the affiliations panel conducted a series of four site visits in November 1989 and a mail survey of 24 VAMCs in February 1990 to help clarify the contribution of C&A and WOC physicians.

Four large, highly affiliated facilities were visited. Discussions with chiefs of staff, directors, and service chiefs at these four hospitals suggested that a careful accounting of time actually contributed by C&A and WOC physicians yields a small number of physician FTEE per month (approximately 2.0 to 3.0 FTEE in the VAMCs visited). Although many physicians whose names appear on the lists for C&As and WOCs (and there are frequently a relatively large number of individuals) may be involved in such activities at a particular VAMC, many may come to the VA only once or twice a month, for an hour or two at a time, and much of their time may be spent in teaching or supervising residents. Therefore, the translation of C&A and WOC time into FTEE results in relatively small numbers. The largest and most highly affiliated facility of those visited showed a total of 3.0 C&A and WOC FTEE per month for the entire VAMC.

(More details of the site visit findings are included in the affiliations panel's report in Volume II, Supplementary Papers.)

In the committee's view, the contribution of these C&A and WOC physicians is much greater than the 2.0 or 3.0 FTEE per month that their collective hours suggest. These physicians often represent highly skilled subspecialists to which the VAMC has access largely through this arrangement. (In the absence of affiliations, this subspecialty care would typically be obtained through contracts with physicians in the community, often at rates much higher than the VA salary scale.) Other aspects of this significance are discussed elsewhere in this chapter.

As shown in Table 9.3, VA physicians in affiliated VAMCs spend some time in research activities, especially in the larger, more highly affiliated institutions. Most of the research is separately budgeted, either through VA research funds or from external sources. Although physician time spent in research is often not accounted for in the VAMC's personnel budget, it is explicitly estimated in the facility's Cost Distribution Report. The committee believes that other benefits to VA patient care accrue from this opportunity for VA physicians to engage in research, as discussed in subsequent sections of this chapter.

In considering the net effect of affiliations on meeting physician requirements, the extensive use of part-time physicians by affiliated VAMCs raises several questions. Do time allocations in the VA data systems accurately reflect the time actually spent on VA functions? Are the contributions of several part-time physicians who comprise one FTEE truly equal to one full-time physician?

There are no systematic data available to address these questions. Therefore, the panel made site visits to eight VAMCs during which service and section chiefs were asked to describe the distribution of part-time and full-time physician FTEE on their services and the reliability of the data systems as indicators of levels of time and effort. These discussions revealed that, with some exceptions reported anecdotally, physician time contributions are described with reasonable accuracy in existing VA data systems. Although there may be distortions (i. e., physicians contributing more or less than their assigned hours), interviews during the site visits indicated that these probably cancel each other out. Further, the diversity contributed by the range of expertise and experience that part-time physicians bring to the VAMC compensates, on balance, for whatever loss of efficiency may result from having several part-time positions instead of one full-time position.

The essential issue in addressing these questions is whether providing the same patient care services would cost more or less in the absence of affiliation arrangements. Although it is difficult to reach definitive conclusions about this issue, on the basis of available data, the implications have been explored.

Through affiliations the VAMCs have access to an array of physicians, many of whom are highly specialized, whose services would have to be obtained by other means if the affiliation arrangements did not exist. For most specialties, VA salaries are not competitive with earnings available in private practice or medical school faculty staffs. At many tertiary care VAMCs, if consulting services were not provided by non-VA physicians through affiliations, the facilities would have to obtain these physician services at market rates—and the impact on their budgets would be substantial.

Many examples of the benefits of availability of C&As and WOCs were identified in the site visits made by the affiliations panel. The chief of surgery at one VAMC estimated that between 40 and 50 percent of the surgical procedures at his facility are supervised or performed by WOCs. It was estimated at another VAMC that 40 percent of patient care time in medicine and its subspecialties is being contributed by WOCs. For example, at that VAMC, only one dermatologist is on staff, but there are five WOC dermatologists who help cover the four clinics each week. It was also noted that consultants often provide backup coverage, but they are not paid unless an emergency brings them to the VAMC. The relatively low FTEE for C&As and WOCs, as reported earlier in this chapter, probably does not reflect the important role non-VA physicians play in providing coverage at the affiliated VAMCs.

The use of residents also provides substantial patient care services at lower rates than equivalent full-time physician services purchased on the open market or provided through VA physician staff. These savings, however, are offset in part by the costs of teaching and supervision. Since the total number of residency slots is substantial (8,400), it seems unlikely that there could be a cost-effective substitution of other physician services for all of the services now provided by residents.

A third question about costs is whether teaching hospitals are more expensive for equivalent services. It has been generally acknowledged in the private sector that teaching hospitals as a group incur substantially higher costs than do nonteaching hospitals, although the exact reasons for these differences are difficult to document. One point is clear: In the VA, as in the private sector, highly sophisticated and expensive tertiary care programs serving the

entire community and the regional area are nearly always located in affiliated teaching hospitals. Consequently, patients with the most difficult diagnoses and needing the most complicated treatments—the most expensive patients—are typically referred to and cared for in teaching hospitals. Further, the geographic environment in which most major teaching hospitals are located—large urban centers—adds to their operating costs because of increased costs for labor and supplies.

In considering the value of affiliations for the VA medical care system, the committee believes that substantial benefits accrue to patient care, including access to high-technology care, access to highly qualified physicians, and a higher quality of care, in general.

In the U.S. health care system, the provision of cutting-edge tertiary medical care is most frequently provided by teaching institutions with strong ties to medical schools—a patient care environment where the medical care is closely affiliated with education and research activities.

An inventory of clinical activities in VAMCs was conducted in 1990 by the Commission on the Future Structure of Veterans Health Care, an independent advisory group set up by the VA to recommend future strategies for its health care system. The inventory showed that affiliated VAMCs tend to have a substantially larger array of services available for their patients. Specifically, this inventory indicated that the concentration of high-tech services (the type typically associated with tertiary care) in affiliated hospitals was very high (Table 9.4). Although such services could be provided by other means, a very fundamental change in current arrangements would be required because of the pervasive nature of the affiliation arrangements. And, any proposed change would have to bear the burden of proof that the alternative approach would do the job as well.

The complex arrangements for providing physician services under the affiliation agreements involve faculty appointments, teaching and research

opportunities, the sharing of clinical staff with medical schools through part-time appointments, and the provision of services by C&A and WOC physicians and by residents under the supervision of teaching staff. All of these arrangements provide access to highly qualified physicians, some of whom probably would not be available to provide services in VA institutions without the affiliations.

It was not part of this committee's charge to look in detail into the recruitment and retention of VA physicians. However, the committee firmly believes that if the VA attempted to staff the currently affiliated VAMCs by some way other than through the affiliations, the recruitment and retention of equally qualified physician staff would be much more difficult. Again, the burden of proof would be on the proposer of the alternative method.

The belief that affiliations result in a higher standard of care is widely held, and this belief is shared by the committee. However, this is extremely difficult to prove on the basis of empirical data. Outcome measures, such as mortality statistics, are still in the early stages of development, not just for the VA system, but for the medical care system as a whole. In 1989, the VA did conduct a review of mortality in VAMCs using a modification of the Health Care Financing Administration (HCFA) methodology. Of the VAMCs with mortality rates significantly higher than the predicted rate (i.e., the lower limit of the 95 percent confidence limits around the observed/predicted ratio was greater than 1.0), 6 of the 11 institutions so identified were in RAM Group 6—unaffiliated psychiatric institutions. These data certainly are not definitive, but they do suggest that there may be a concentration of quality-of-care problems in the unaffiliated institutions (U.S. Department of Veterans Affairs, 1989).

Structural and process measures of quality, such as the percentage of physicians who are board certified, or the accreditation results from the Joint Commission on the Accreditation of Healthcare Organizations (JCAHO), may provide some additional inferences concerning effects on quality. (As discussed in chapter 2, the VA is currently developing improved indicators that, in the future, could be used to monitor more definitively the quality of care provided in VAMCs. When developed, they could be used to analyze the quality effects of affiliations.)

The committee also notes that the previously cited data on the inventory of clinical services defines another very basic dimension of quality. If access to a particular type of service needed for the care of an individual patient is not available at all, that constitutes prima facie evidence of a deficit in quality, in the committee's view.

The committee presumes the value of affiliations for providing access to highly specialized tertiary care. However, the health care needs of veterans include primary care, rehabilitation, and care of the chronically ill, among whom a substantial number have chronic psychiatric impairments. It can be argued that the improved linkage to tertiary care may be of little benefit to these other patient care needs.

The aging of the population will lead to a higher incidence of chronic conditions within the veteran population being served by the VA. The findings of the specialty and clinical program panels, described in the appendix to chapter 6, suggest that psychiatry, rehabilitation, and ambulatory care services may be less well served by current VA staffing patterns than are the traditional tertiary-care subspecialties. The committee notes that the VA has given special emphasis to some of these needs through establishment of special geriatric centers and the development of new approaches to long-term care. Certain rehabilitation needs have also received special emphasis within the VA, specifically spinal cord injury and the development of prostheses.

The committee believes that the VA can ensure that the pattern of services offered in affiliated institutions matches the pattern of patient care needs among the VA population by developing additional programs in ambulatory and long-term care, giving special emphasis to psychiatry and rehabilitation initiatives. Through such programs, the VA also has an opportunity to provide leadership to the medical education community that will benefit all patients.

For primary care and the care of the chronically ill, certain attributes of care, such as continuity—a component of the normative definition of primary care developed by the IOM (Institute of Medicine, 1978)—may be important for encouraging patient compliance and behavior change. A pattern of care that depends heavily on residents and part-time physicians may not be conducive to these desirable patterns of care.

Again, the committee notes that, because many VAMCs are tied so closely to tertiary care medical environments, special attention may need to continue to be given to overcoming particular problems with the pattern of care typical of the teaching affiliation environment.

The current VA mission includes research, education, and backup to the Department of Defense in time of war. The data already referenced in this chapter indicate that the involvement of the VA physicians in research and teaching is clearly associated with affiliations, and it seems unlikely that these aspects of the mission could be sustained in the absence of affiliates.

To provide backup to the Department of Defense in time of war (see chapter 1), the VA would likely depend as well on the affiliation relationships to provide some highly specialized services, such as burn therapy and treatment of other extreme traumas.

If the affiliations are not maintained, it would be very difficult for the VA to perform these aspects of the current mission, in the committee's judgment. It can be argued that the research, education, and backup to the military provide extra benefits to the broader society, as well as helping to sustain a high-quality medical care system in the VA. Consequently, any deemphasis of affiliations that signals—correctly or not—that the VA has narrowed the scope of its mission could raise questions about the net benefit to American society of its substantial tax investment in the VA health care system.

The affiliations panel reports a growing feeling on the part of the unaffiliated hospitals that they are being treated as "second class citizens." Chiefs of staff at these VAMCs say their facilities are not sufficiently staffed, and that the VA's method of resource allocation has not adequately reflected their needs, favoring instead the already "well-endowed" affiliated institutions. Staff in unaffiliated facilities claim they are not accorded the same status as their peers in affiliated facilities.

However, there is also reported to be increasing recognition of the positive value of, and potential for, new directions in affiliations, especially on the part of unaffiliated hospitals. Possibilities for new affiliation relationships involving secondary and primary VA facilities are increasing. This is generally seen as a positive trend, which is likely to upgrade patient care in those facilities. Consortia of hospitals—involving a medical school, a tertiary VAMC, and one or more smaller, less specialized VAMCs in the same geographic area—are indicative of this trend. (One example reported to the committee involves an innovative affiliation arrangement between a major New York medical center and three VAMCs in the region.) Other models could be developed that would affiliate the VAMC with a major community hospital offering tertiary services but not strongly linked to a medical school.

There appears to be growing recognition that affiliations could have benefits for these smaller, nontertiary care facilities—benefits that include the attraction and retention of highly trained staff, as well as improved morale and increased intellectual stimulation. Further, some academic centers are interested in the primary care patient pool served by some unaffiliated VAMCs; relationships between these two types of facilities may be beneficial to both parties.

Some panel members pointed out that even in the most highly affiliated institutions, there are some specialties or program areas that are not affiliated, especially in ambulatory care. As the demand for ambulatory and chronic care services grows, attention is increasingly being focused on the need for residency education in these settings. Expanding VA—medical school affiliations to include these types of services and facilities may prove to be vital to the VA's ability to provide high-quality patient care to the veteran population.

Through its affiliations, the VA has contributed to existing knowledge and quality of patient care in long-term as well as ambulatory care services. The interrelationships and interdependencies among services were mentioned as important features of affiliation: Services may need one another in order to be accredited, as in the case of surgery and radiology.

The committee has reached a very firm conclusion that the overall effect of affiliations on the VA health care system is strongly positive. The benefits include:

• An improved ability to attract and retain well-qualified physicians and other health professionals;

• A wide spectrum of services provided by a pool of highly qualified physicians, both those on the VA staff and those whose services are made available to the VA through other relationships with the medical schools;

• Access to state-of-the-art tertiary care;

• Participation in the education of physicians, which is a mandated part of the VA mission and which cannot realistically take place currently in the absence of affiliations;

• Participation in medical and health services research, resulting in contributions to the advancement of medical knowledge and improved health services that benefit the general population as well as veterans.

Underlying all of the above is the assumption, and some inferential indications, that affiliations contribute significantly to improved quality of patient care.

In other parts of this report, the committee has urged the VA to continue the work currently being conducted by its Office of Quality Management to develop quality-of-care indicators. These indicators will be critical, not only for the full development of the physician requirements methodology, but also for a more definitive evaluation of the effect of affiliations on the quality of care. The development of structure and process measures—such as information on board eligibility and board certification, analysis of the matrix of quality measures and scores utilized by JCAHO, and further refinement of the availability of specific clinical services within all VA facilities—may all be useful interim steps. However, the full development of quality indicators will require a more sophisticated array of health outcome measures including, but going far beyond, mortality rates. In this effort, the VA should closely track the extensive effort and developmental work being done by many health services researchers and health care organizations on outcome-related quality measures.

The committee recommends that the VA explore strategies for developing and expanding affiliations to include facilities that currently are not affiliated. This recommendation follows logically from the previous conclusion that affiliations bring benefits to the VA medical care system. Given that conclusion, the committee believes there is no logical reason not to provide at least some of the benefits to veterans cared for in all VA facilities. Clearly these new affiliations would have to be tailored to the size of the facility and the scope of the services offered, as well as to other particular attributes of the facility. Such an expansion of affiliations would promote and encourage recruitment and retention of high-quality staff and many of the other benefits outlined in the previous conclusion.

The committee recognizes that new affiliations are not easily attained. Successful affiliations require that both the VAMC and the non-VA institution have a strong perception that the benefits are significant. Hence, to develop new affiliations, the VA should pay special attention to the attributes of successful affiliations. These include ample opportunities for research, for teaching, and for shared services, as well as an adequate support staff and a sound infrastructure at the VAMC.

Failure to consider extending affiliations could cause some to question the VA's commitment to quality of care for the entire veteran population being served.

The committee recommends that while maintaining and nurturing the current model of affiliations between VAMCs and medical schools, with its emphasis on tertiary care, the VA should work to develop innovative models of affiliation targeted specifically to the chronically ill, including those requiring psychiatric care and rehabilitation services. These innovative models would, in general, be oriented around and give emphasis to ambulatory and long-term care.

The nature of the VA patient population presents special opportunities, and needs, for the development of new models. The Geriatric Research, Education, and Clinical Centers (GRECCs) represent an example of a successful model already developed by the VA to meet the particular needs of the population being served. Other opportunities, emphasizing a broad array of health services research related to these patient care needs, could continue to make the VA health care system a resource for the benefit of the entire nation, as well as for the veteran beneficiaries. In developing these new ideas, the VA could emulate its fine record in the conduct of multi-institutional clinical trials. Equivalently, training opportunities focused in innovative ways on these particular patient care needs could make a major contribution to veterans and to the general population. A recent IOM report on financing graduate medical education in primary ambulatory care emphasized the need for strengthened environments for this particular educational purpose (Institute of Medicine, 1989). The VA is in a logical position to support its own purposes and the purposes of the broader society.

In developing these innovative affiliation approaches, the VA should explore the establishment of relationships with other medical institutions in addition to medical schools. The VA has already created the beginnings of a new model of affiliations with a recent program involving community-based health care institutions that are not primarily related to medical schools. Such arrangements would be consistent with the intent to extend the purposes of affiliations beyond providing access to acute services and state-of-the-art tertiary care. The committee believes that this extension represents an exciting opportunity for the VA that could help meet some of the staffing needs that are likely to be identified through the application of the proposed physician requirements methodology.

Commission on the Future Structure of Veterans Health Care. 1990. Inventory (data set) of VA clinical activities. Office of the Executive Director, Commission on the Future Structure of Veterans Health Care, U.S. Department of Veterans Affairs, Washington, D.C.

Institute of Medicine. 1978. A Manpower Policy for Primary Care. Washington, D.C.: National Academy of Sciences.

Institute of Medicine. 1985. Plan for a Study to Develop Methods Useful to the Veterans Administration in Estimating Its Physician Needs. Washington, D.C. Unpublished.

Institute of Medicine. 1989. Primary Care Physicians: Financing Their Graduate Medical Education in Ambulatory Settings. Washington, D.C.: National Academy Press.

U.S. Department of Veterans Affairs. 1989. Review of Mortality in VA Medical Centers. Washington, D.C.: Department of Veterans Affairs, pp. Dl-D9.

Worthen, D. 1987. The affiliation partnership between U.S. medical schools and the Veterans Administration. Alabama Journal of Medical Sciences 24:83-88.