Subspecialization in geriatrics through board examination and the accreditation of geriatric fellowships signaled recognition of the geriatric imperative in mental health education. However, mercantilism in the health care workplace, reduced support for graduate medical education, and a decline in residency applicants dictate that general rather than geriatric psychiatrists will provide the majority of specialty mental health services to older adults in the twenty-first century (+1). Yet curricular requirements for experience with late-life mental illness remain minimal in general psychiatry training. In 1995 the Council on Aging of the American Psychiatric Association appointed a work group to review the geriatric curriculum in general residency training. Several issues motivated the review.

First, projections of the prevalence of mental disorders among future cohorts of older adults are controversial (+2). However, the increasing number of older Americans means more seniors will attempt suicide, develop dementia, have emotional disorders, or experience excess disability when mental illness complicates a physical condition. And as prejudicial attitudes about psychiatric treatment decline, there will also be an increase in the number of seniors seeking psychotherapy for problems of daily life that transcend diagnostic categories. These increases demand a response in services, research, and professional education. Second, despite a number of model curricula for geriatric fellowship training, those dealing with general residency years 1—4 are more than 10 years old. Third, significant advances in clinical research and service delivery concepts that apply to older patients have not achieved wide currency in health services planning or education. Fourth, compared with geriatric recommendations for internal medicine and family practice, the American Council on Graduate Medical Education (+3) program requirements for psychiatric residency training are minimal (+4). They exclude rotation on a geriatric unit from the 9-month minimum inpatient experience. A supervised geriatric experience covering a "variety of disorders" and in a "variety of settings, including long-term care," is required but left vague.

We examined material from the United States, United Kingdom, Canada, and Australia, including publications in primary care, general, and geriatric psychiatry. We summarize accomplishments, suggest further advances, anticipate problems with implementation, and propose solutions.

In 1950 the Group for the Advancement of Psychiatry drew attention to the neglect of mentally ill American elders (+5), as Sir Martin Roth (+6) rejected the therapeutic nihilism associated with psychogeriatrics in the United Kingdom (+7). The first National Institute of Mental Health (NIMH)-supported training program in geriatric psychiatry was established at Duke University in 1965, followed in 1975 by the Center for Studies of Mental Health of the Aging that included both clinical and research training (+8). The 1976 Conference on Training Programs in Mental Health and Aging (+9) found a shortage of clinicians trained to care for older adults. Publications by Kahn et al. (+10), Butler (+11), and Busse and Pfeiffer (+12), as well as the births of the Boston Society for Gerontologic Psychiatry (1960), the American Association for Geriatric Psychiatry (1978), and the Council on Aging of the American Psychiatric Association (1979) , set the stage for a formal curriculum.

In 1982, NIMH funded the American Psychiatric Association to formulate a detailed curriculum for medical students and psychiatric residents (+13). Several authors both before and after 1982 (+5,+14—+26; Shamoian, unpublished) described psychogeriatric rotations or curricula for students and trainees in psychiatry. Taken as a whole, the recommendations are comprehensive, with a flexible approach to implementation that recognizes program diversity.

However, by 1989 only 36%, of psychiatric residency training programs required a geriatric rotation and only 21% offered geriatric electives (+27). By 1993 half of a representative sample of the 140 accredited programs in the United States provided a "substantial" geriatric experience, defined by a required inpatient, outpatient, or elective experience in which at least 50% of residents participated (+28). Programs with geriatric exposure were more frequently those with geriatric divisions or fellowships. Of the programs with substantial geriatric offerings, two-thirds identified inpatient or consultation-liaison services as geriatric psychiatry sites and half reported a geriatric psychiatry outpatient clinic. Only a third of the programs cited nursing home, a quarter a day treatment program, and less than a fifth home-care experiences. This pattern was not unique to psychiatry. As of 1992, only one-third of internal medicine residency training programs offered nursing home rotations (+29). Similarly, medical schools in the United States provided little instruction in palliative or end-of-life care, unlike those in Canada, the United Kingdom, and Australia. As a result, most physicians in the United States have had no training in palliative care and are ill-equipped to refer their patients to practitioners of palliative medicine (+30).

More recently, McCartney's 1997 survey of 74 training programs found the percentage requiring a geriatric rotation had increased to 55% but that only 40% listed a nursing home as an elective site (McCartney JR: personal communication, April 8, 1997). Required nursing home rotations were listed in only 28% of the programs. Thus, progress toward a mature geriatric curriculum for general psychiatric trainees remains incomplete despite the advent of added qualifications in geriatrics, the greater number of older adults in need of psychiatric care, and the accreditation of some 45 fellowship training programs in geriatric psychiatry.

The United Kingdom Health Services started to promote collaboration among psychiatrists, geriatricians, and social service agencies in the 1960s. The United Kingdom in 1981, Canada in 1987, and Australia in 1995 introduced guidelines for geriatric training in general psychiatry programs. Since 1981, United Kingdom psychiatric trainees have been required to spend "a significant proportion of time" in psychogeriatric service (+31). By 1985 the number of psychiatrists choosing to specialize in the care of older adults increased from "a handful" to over 250. Subsequently, the Working Party of the Royal College of Physicians and the Royal College of Psychiatrists outlined their training and service delivery objectives for the elderly (+32). This interest was mirrored in Canada. In 1981, only 4% of general psychiatry residents had exposure to a geriatric psychiatry service (+33). By 1986 this figure had increased to just 11% (+34). Then in 1987, the Royal College of Physicians and Surgeons included guidelines for geriatric training in general residency programs. These were appended with a more detailed interpretation in 1988 (+35). In brief, the guidelines outline a 3-month full-time or 6-month half-time experience and didactics in geriatrics in more than one clinical setting.

The Section in Psychiatry of Old Age of the Royal Australian and New Zealand College of Psychiatrists was established in 1988. After surveying all psychiatry trainees and program directors in 1992, guidelines were developed in 1995. All trainees are to spend 3, but preferably 6, months working full-time in a specialist psychiatry service for the elderly. Such training occurs after the completion of the postgraduate year (PGY)-1. Throughout Canada, the United Kingdom, and Australia there were obstacles to implementation. Flexibility has been the key response. Problems with core content, rotation length, site exposure, and supervisor expertise have been addressed in various ways. The essentials are similar in Canada, the United Kingdom, and Australia with a didactic core and clinical experience (+36—+38). The Nottingham program has established an "á la carte" system whereby an essential topic list is given to residents who are then responsible for ensuring that they obtain appropriate teaching and clinical exposure (+39).

The concern of integrating geriatric psychiatry into already "full" programs has been met by implementing several rotation lengths: from 3 months full-time (at any time from PGY-2—PGY-4) to 6 months half-time in Canada; to four 1-month periods over 4 years in some United Kingdom programs; to 3—6 months full-time in Australia. Other programs do not mandate length of time. One argument against the integration of geriatric psychiatry into general residency training has been that not all programs have access to a geriatric psychiatry inpatient unit. Neither Canada nor the United Kingdom make an inpatient setting mandatory, but they have been creative in placing residents on geriatric medicine wards and in encouraging community placements, especially long-term care facilities. Collaborations with other specialties and a multidisciplinary team approach are also emphasized.

Qualified supervisors in Canada include not only geriatric psychiatrists but also members of the Division of Geriatric Psychiatry of the Royal College that has membership criteria (+40). In Australia, supervisors are expected to be accredited by the Faculty of Psychiatry of Old Age. In the United Kingdom, supervisors include psychogeriatricians (geriatric psychiatrists), geriatricians (geriatric internists), and multidisciplinary teams (+37). Nonetheless, concerns remain about inadequate numbers of qualified teachers (+36). However the United Kingdom and Canadian experience, to date, indicates that implementation has been aided by two factors in particular: 1) an academic geriatric course organizer and 2) collaboration with department heads of other clinical disciplines. Hermann et al. (+41) surveyed the impact on Canadian practice. They reported that 9 of 16 residents who completed a geriatric psychiatry rotation developed general practices with a "substantial" elderly population. A subsequent survey (+40) of residents and training directors found that training guidelines that have been developed to consider program resources are indeed adhered to closely. Thorpe et al.'s 1993 survey showed that 13 of 16 programs in Canada had implemented a 3-month minimum geriatric rotation by 1992 (+36).

Noting the growing number of elderly patients and their disproportionate use of resources in primary care, the American Medical Association in 1994 set up work groups to address levels of geriatric education in internal medicine residency training programs. The curriculum work group suggested that residency programs in primary care develop and implement didactic and training experience in aging (+42). To develop expertise with disease prevention and management in the elderly, the work group argued for a mandatory block rotation, including participation in geriatric assessment and consultation and experiences in acute-care geriatric units and in nursing homes. Other training sites might include home-care agencies, daycare centers, respite care, hospice programs, and dementia clinics.

The work group also urged the development of a faculty who demonstrate excellence in geriatric care. The teachers may include a multidisciplinary team, with psychiatrists, neurologists, nurses, physical and occupational therapists, nutritionists, social workers, and others. The work group also suggested that the Department of Internal Medicine chairperson play a central role in the implementation of the program. Training programs in internal medicine are now required to have "formal teaching and regular, supervised clinical activities" in geriatric medicine. Assignments to geriatric "services must be offered," including nursing homes "and/or home care" (+3). Similarly the Program Requirements for Residency Education in Family Medicine also require that geriatric elements "of the curriculum must be available throughout the resident's entire program," with "didactic conferences and clinical experience" provided across a variety of settings, including nursing facilities and patients' homes. Of note, the duration and site of the experience are not mandated, leaving programs responsible for offering a curriculum of lectures and clinical exposure but not a specific rotation.

Adding to the emphasis on aging within the internal medicine residency training was the reduction in geriatric fellowship training from 2 years to 1 year, with the expressed intent to incorporate more geriatric content into generalist training years and produce more geriatric specialists in a shorter time (+43). The reduction in fellowship training was controversial. Some argue that 24 months are necessary to acquire the clinical and research skills necessary to develop future faculty (+44).

As described earlier, the existing literature provides considerable detail as to knowledge and skills the practicing general psychiatrists will need to work with older adults in the twenty-first century. We have reduced and updated the categories and presentation of that knowledge in +Table 1 and +Table 2. Obstacles to implementation and possible solutions appear in +Table 3. The tables represent a summary consensus drawn from past recommendations for mental health specialists and recent guidelines published for internal medicine and family practice (+3,+45—+48), as well as reactions to our draft document from the Council on Aging of the American Psychiatric Association, the Graduate Education Committee of the American Psychiatric Association, the Education Committee of the Association for Geriatric Psychiatry, and members of the American Association of Directors of Psychiatry Residency Training.

In the first two post-graduate years, residents should receive at least one lecture on the sociology and demography of aging societies, detailing the implications for social policy and health care financing (+Table 1). The political history of Social Security and Medicare should be included. Theories of aging and the biology of the aging brain and body, including changes in sleep and sexuality, should also be covered. Developmental milestones of late life (grandparenting, relocation, bereavement, retirement) and the differentiation of aging from age-related illness and lack of exercise are included to emphasize healthy, successful, and productive aging. A positive attitude toward older adults should be advanced by teaching cultural competence on ageism, poverty, religiousness, ethnicity, and the problems of older émigrés. Unlike today, when the majority of older Americans are white, by 2010 persons over the age 65 in the United States will be a far more heterogeneous population (+49). Projections based on present demographic trends indicate that ethnic minority populations will experience a disproportionate increase in morbidity due to dementia (+50). The minimal geriatric skills should include sensitivity to cultural differences; the ability to explore patient and family expectations; and awareness of alternative therapies (botanicals, rituals, medications from abroad, healers) used by ethnic elders, plus the degree of acculturation (+51).

The epidemiology of late-life mental illness, suicide, substance abuse, and elder abuse are also addressed in the first 2 years. This focus leads to instruction on utilization of community resources and social service agencies as well as legal and ethical issues, including guardianship, advanced health directives, durable power of attorney for health, research participation, and finances. These topics should be covered during outpatient rotations, while the assessment of decisional capacity may be taught during the consultation-liaison rotation. Age-related considerations in psychopharmacology and medical comorbidity are introduced early, to counter the resident's concerns over adverse effects and drug interactions in the elderly. Adapting psychotherapeutic approaches for older adults follows to ensure that aged persons and their companions, spouses, and families are seen early in the resident's experience as appropriate candidates for psychotherapy.

In the third and fourth post-graduate years, the focus is on complex clinical problems and the scientific basis of practice. Comorbidity of physical illness and mental illness in late life is introduced, with the emphasis placed on functional outcomes and the psychiatrist's contribution to prevention and reduction of disability. Diagnostic and therapeutic approaches to maximizing autonomy and well-being would include nutrition and exercise. Related to these topics are the atypical presentation of illness in late life and the geriatric syndromes of cognitive impairment; frailty; gait disorder and falls; sensory impairments; immobility (arthritis, osteoporosis); pain; incontinence; and polypharmacy. Elements of palliative care and preserving dignity and meaning through the end of life are also covered in the latter half of the training, when a greater degree of confidence has been achieved. Finally, to prepare the resident for future progress, methodological advances in the behavioral and laboratory sciences, health services research, genetics, structural and functional brain imaging, neuropathology, neuropsychology, and psychopharmacology are explored.

To gain a direct view of healthy, productive aging, residents in their first year should be exposed to "successfully aged" seniors through attendance at service organizations, elderhostel programs, and older physicians' offices (+Table 2). Supervised assessment of older psychiatric inpatients and continuing care of geriatric outpatients, including those with dementia, should occur early in training to allow the resident adequate time to observe the course of illness and the changing nature of required interventions. Consultation-liaison experience in the general hospital should also bridge to nursing homes and other extended-care facilities (+52).

In the latter half of training when skills and identity begin to consolidate, the residents are introduced to home care and participation in multidisciplinary team meetings and rotations with geriatricians in primary care or specialty settings as well as "subspecialists" (geriatric neurology, neuropathology, neuroradiology, neuropsychology, physiatry). Palliative care and end-of-life therapeutics in hospice, home care, and hospital settings are also reserved for the final years of training. Third- and fourth-year residents should also have opportunities for participation in geriatric education of medical students, research, and other scholarly pursuits. Not every program will be able to include residents in research with older adults. Nonetheless, the need to incorporate advances from the behavioral and neurosciences as well as health services research is critical. The ongoing emergence of new medications for psychosis, depression, anxiety, and most recently for cognitive impairment is a clear example.

Training directors face a number of obstacles to expanding the geriatric curriculum in their programs (+Table 3). First, existing programs are perceived to be overloaded with "essential" (mandatory) criteria. Any new initiative either takes time from existing rotations or from electives. However, curricular overload may be a problem of perception rather than practice. With 12% of the present U.S. population being 65 or older and projected to reach 20% in the coming decades, a 3-month full-time or equivalent experience would represent but 7% of the 48 months (or 12% of PGYs 2—4) of required training in general psychiatry. Already mandated is a 2-month rotation in consultation-liaison psychiatry. In many programs, a significant proportion of the consultations are for geriatric patients. Allowing the geriatric experience to be overlapping, rather than exclusive of existing experiences in consultation, inpatient, or outpatient sites, would give many programs credit for geriatric education already provided. The low rate of geriatric exposure reported by existing programs may represent an underreporting of available geriatric experience within the clinical sites and current rotations.

Daylong seminars demonstrating psychiatric practice in the nursing home, home care, palliative care, and management of dementia caregiver burden may be a ready solution for smaller programs whose faculty or locale make their trainees' access to geriatrics problematic. Also, opportunities for trainees to participate in conferences, symposia, and national meetings of the specialty interest societies (see Appendix 1) might be provided.

A second obstacle is the perception that adding more geriatrics will duplicate the geriatric psychiatry (PGY-5) fellowship. Here we suggest that the guidelines focus on exposure and opportunities for minimal competence rather than expertise. A number of topics lend themselves to being taught from the geriatric perspective. For example, older adults are more susceptible to adverse drug reactions and multiple drug interactions. These vulnerabilities are an opportunity to present pharmacodynamic and pharmacokinetic principles that apply to younger adults but are more frequently encountered with the elderly. Similarly, physiologic changes in sleep and sexuality, the clinical import of ethnicity and acculturation, and epidemiology can be taught from the geriatric vantage as a way of covering basic material. Thus, considerable geriatrics content can be introduced into existing modules without displacing either the teachers or the topics.

Third is the problem of supervisors with appropriate geriatric credentials. Although the availability of qualified supervisors must be assured, all geriatric supervisors and lecturers need not be psychiatrists with added qualifications or fellowship training. Demonstrated competence through experience in nursing homes, home-care agencies, and attendance at continuing medical education venues and geriatric specialty meetings may be sufficient to credential the faculty. Fellowship status within the American Geriatrics Society or the Gerontological Society of America might also be used. Similarly, experience with geriatric patients through consultation-liaison rotations, case conferences with the geriatric medicine team, collaboration with neurologists who have a special interest in dementia or palliative-care consultants can be adopted without overburdening the psychiatric table of organization.

Fourth, financial support for supervisors is an obvious concern. Faculty support, beyond reimbursement for patient care or funded research, is a problem for all educational efforts. However, the availability of continuing graduate medical education (GME) monies derived from Medicare may yield added emphasis on geriatric training. The threatened reductions in GME funds may yet uncover funding streams for teaching and training protected from the need to make up salaried hours through service or research grants (+53).

A fifth obstacle is represented by the difficulties in providing exposure to geriatric patients and integrating areas such as rural care, nursing home and house calls, and cultural competence with minority elders. Training focused on sites where geriatric patients are concentrated may solve the problem of insufficient clinical material. Older persons with mental disorders may be more accessible at primary care sites, home-care agencies, or nursing homes. Naturally occurring retirement communities with senior citizens concentrated in a relatively small geographic area represent another point of access. Obviously, psychiatric practice at these sites requires liaison with other clinical services. Collaboration with other disciplines, including geriatric and family medicine and neurology, may add the geriatric venues without overextending the psychiatric program. Similarly, agencies providing social support and home-care services can demonstrate principles of geriatric care through exposure of the trainees to physical and occupational therapists, social workers, and geriatric nurse practitioners.

The danger of generalizing across ethnic elders should be prevented through case supervision and in didactic seminars. These patients will come for treatment at the request of their families or community agencies and are likely to be accompanied by a caregiver of similar background. Resources for cross-cultural training as an integral part of ongoing clinical case conferences and didactic seminars can be addressed with model curricula on competence with minority patients (+51).

A sixth obstacle is the insufficient access to laboratory and clinical scientists with up-to-date knowledge of research in late-life mental illness. On-site access to geriatrics researchers is certainly desirable but may not be critical to the educational mission of every program. Again, scientific meetings of geriatric specialty societies may present adequate exposure to research. Nonetheless, lack of geriatric psychiatry research faculty will remain a serious obstacle to the discipline's advance.

Seventh, without a clear incentive to change, programs may forestall further evolution. Ultimately, changes in the requirements for program accreditation will be essential. Yet as with internal medicine and family practice, national changes can be mandated, with the specifics left to local directors, the minimal goal being a heightened visibility of geriatrics in the general curriculum.

Finally, without resident and program evaluation, implementation is unlikely to occur. Use of patient logs to document the patient care experience, letters of affiliation with long-term care facilities and home health agencies that identify on-site supervisors, and attainment of concrete objectives at each site might be used to document the experience. Another method of demonstrating the benefits of a more visible geriatric curriculum would be improvement in percent-correct responses to geriatric questions on the Psychiatry Resident In-Training Examination and the written examination of the the American Board of Psychiatry and Neurology. Already, the American Board of Internal Medicine-certifying examination consists of about 12% specific questions on geriatric care. For psychiatry, the figure is 16%. A figure of 20% seems more appropriate in primary care and psychiatry. More extensive self-assessments material are available (+54).

The difficulties with geriatric training are not unique to psychiatry (+48,+55—+58). Although market forces are expected to correct the oversupply of specialists (+59—+61), the number of fellowship-trained and certified geriatricians in practice and teaching is expected to decline after the year 2000 as the present cohort retires (+62,+63). As a result, there will never be enough geriatricians or geriatric psychiatrists to meet the needs of all American seniors. Demographic imperatives will press all physicians, including general psychiatrists, to work with more seniors. The multidisciplinary team approach, advocated by primary care geriatricians, further ensures that psychiatric expertise will remain in demand, provided expertise includes competence in geriatrics. Neither will advances in psychopharmacology and psychotherapy for primary care patients be fully realized without psychiatrists (+64,+65). And the legislative debate over the availability of mental health services (Mental Health Parity Act [Public Law 104—204]) is an additional sign that psychiatric care is becoming less stigmatized.

However, there seems little likelihood that the reduction in positions for specialty trainees will be reversed given concerns over the cost of training borne by Medicare's GME funding mechanisms. The Indirect Medical Education formula that prorates Medicare payments for training in teaching hospitals based on the number of trainees is also proposed for reduction (+66). Moreover, Medicare part B (physician services) guidelines for faculty supervision require that the procedure codes used for billing conform to the face-to-face interaction with the patient. This change means that procedures are effectively limited to less remunerative medication management, brief initial or follow-up consultations, or brief visits. However, the crisis precipitated by GME reductions has also resulted in calls for an "all-payer trust fund" in which managed care organizations would contribute a fair share to educational costs. Other proposals, which would not carry the trust fund's status of a federal entitlement program, include a resident voucher system in which programs would compete for trainees or an appropriation mechanism in which training funds would compete in the budget process with other social spending (+67). Nonetheless, GME monies seem likely to persist, ensuring that enough geriatric fellowship positions will remain to ensure the minimum number of geriatric educators. In any event, reimbursement for faculty will no longer be based solely on income generated from supervision (+68).

The present stresses that training programs experience will persist into the next decade (+69). Simple solutions in which any addition to the curriculum must be balanced by a deletion will be inadequate. Equally unsatisfactory is a "least common denominator" equation in which the only mandatory changes in training are those that all existing programs can conveniently achieve. A number of trends beyond demographics will force practicing psychiatrists to gain greater facility with senior patients. Thus, the question is not will training programs change, but what are the changing priorities and how might they be implemented?

Expanding the geriatric experience is a clear priority. Three months of the training program (7% of the 4 years) should be identified as geriatric, with the means of implementation left to the program directors. There is a natural stratification based on the extent of a program's resources. General psychiatry training programs with both a geriatric psychiatry and geriatric medicine fellowships (residencies) in the same institution have the riches to accomplish all we suggest. General programs with access to one or the other geriatric fellowships are mid-range, and those with neither are less advantaged. Bridges between local general and geriatric programs will require program directors or department chairs to reach out to their counterparts. Programs without access may chose to outsource the experience to either national meetings or on-campus symposia and grand rounds rather than develop geriatrics de novo. At all strata, some contact with patients from nursing homes and home-care agencies, because of the concentration of older persons in these sites, seems mandatory, but the depth of experience will vary based on resources. We urge directors to review what geriatric content is presently available within their programs, to incorporate that recognition into an identified geriatric curriculum, and to enhance what they have with an added emphasis in the lectures and case conferences. These changes will require revision of the essential program requirements for residency education. We offer these suggestions to assist training directors to manage the inevitable changes in the needs of patients and trainees alike. We echo the philosophy of Our Future Selves (+70). Improved training to care for the aged in the present means better care for all of us in the future.

This work was supported, in part, by the American Psychiatric Association's Council on Aging Work Group on the Geriatric Psychiatry Curriculum Guidelines for General Residency Training. Members included Marion Zucker Goldstein (Ex Officio Chair, Council on Aging), Mustafa Husain (Member, Council on Aging), Gary J. Kennedy (Working Group Chairperson), Rena Nora (Member, Council on Aging), Colleen J. Northcott (APA/BW Fellow), Kenneth Sakauye, and F.M. Baker (Consultants). The authors also acknowledge the communications of Drs. Ewald Busse, Charles M. Gaitz, Charles Reynolds, Lissy Jarvik, James Lomax, Paul Mohl, Eric Pfeifer, the Education Committees of the American Association for Geriatric Psychiatry and the American Psychiatric Association, and many others.