Since the early 1980s, a continuous tide of homelessness has swept people with psychiatric illness into a position of special vulnerability. Macro-economic trends, including the loss of unskilled industrial employment, the rise of a service economy, and especially, a low-income housing shortage lead the causes of homelessness. Concurrent social and mental health policy shifts, notably the way deinstitutionalization was carried out, made it increasingly difficult for people with mental illness to maintain community stability, as well. In fact, about one-third of all homeless persons have a serious psychiatric disorder and up to two-thirds suffer from substance use disorders (1).
Public concern about the plight of homeless individuals who are mentally ill, combined with attempts to adapt mental health systems to the large numbers of such individuals, has impacted psychiatry’s focus on the issue. As early as 1984, the American Psychiatric Association (APA) identified homelessness among the mentally ill as a critical professional concern, publishing the first of two landmark reports on the subject (2). Since then, the profession has developed identifiable roles in the care of this clinical population, including an important academic one (3). And while the phenomenon of homelessness and mental illness has generated a considerable volume of literature, the need to educate psychiatric residents about the design and provision of services for mentally ill homeless people has received relatively limited attention (4).
The literature on psychiatric education concerning homelessness is anecdotal. The earliest training efforts during the 1980s were sponsored by local psychiatric societies or psychiatric residents themselves (5—7). Initial training was necessarily modest, driven by a social crisis, with established practitioners and residents learning together about a "new" population with little, if any, formal resident teaching. In 1994, the National Resource Center on Homelessness and Mental Illness collaborated with the APA to fund annual full-day training sessions for psychiatric residents at the APA’s Institute on Psychiatric Services in order to meet presumptive training needs (8). With the permanence of homelessness, it would be reasonable to postulate that some residency programs have attempted to address the issue within their communities through formal didactics and clinical experiences, although the extent to which this is being done has not been documented.
In discussions of the goals of contemporary psychiatric education, a number of authors (9—12) have examined the skills future psychiatrists must acquire in order to remain adaptive in the evolving mental health care environment. Beigel and Santiago (9) noted particular skills. They include the ability 1) to flexibly practice in a variety of roles—from consultant to clinical leader—within an array of nontraditional outpatient settings, 2) to be able to welcome patients as full treatment collaborators, 3) and to be expert in caring for people with severe psychiatric disorders. Educating residents in how to work with challenging patient populations, specifically homeless individuals who are mentally ill, can meet educational objectives beyond an understanding of the clinical population itself. For example, mentally ill homeless people typically present complex clinical problems, including substance use and prominent physical health issues, which are in turn compounded by the vicissitudes of severe poverty and social alienation.
To examine how academic psychiatry has approached this public health issue and begin to identify recommendations for training, an exploratory study was carried out among psychiatric residency programs in the United States. A survey questionnaire was circulated to directors of residency programs asking them to describe if and how their programs addressed training in homelessness and mental illness, including the nature of didactic and clinical experiences for residents on the topic of homelessness.
The questionnaire consisted of a 10-item multiple choice survey that requested the following information from respondents: 1) to identify whether they consider their program to be in an urban, suburban, or rural setting; 2) to indicate the size of their program (number of residents); 3) to provide the number of hours per residency year of didactic instruction in homelessness issues; 4) to provide the percentage of residents rotating in homelessness programs and whether that time is mandatory or elective; 5) to identify the types of community sites used for clinical rotations (e.g., street, shelter, drop-in center temporary housing, other); 6) to indicate whether resident supervision was done on-site or on campus; and 7) to identify how rotations are funded (e.g., community site, by the program, or other). Additionally, the survey requested programs that did not currently have curricular elements concerning homeless populations to provide reasons for this absence (e.g., no curricular time, insufficient funds, logistics too complicated, relatively low program priority, low community prevalence of homelessness, other). In all "other" categories, the questionnaire asked respondents to note specifics.
We mailed surveys for anonymous completion with self-addressed, stamped return envelopes to the members of the American Association of Directors of Psychiatric Residency Training (AADPRT) in the United States who had a title of adult residency training director or the equivalent (N=178). There was only one mailing, without incentives offered to respondents, and no more than one survey was sent to each training institution. The survey was also posted on the Columbia University Fellowship in Public Psychiatry web site and people were informed that it could be submitted through that medium, as well.
Results were tabulated in Statistical Package for the Social Sciences (SPSS), and data were analyzed by means of chi square, independent samples t test, analysis of variance, or linear regression, as appropriate.
Of the 178 programs, 106 (60%) responded. Of these 106 responses, 64 (60%) had didactic or clinical offerings specifically concerning homelessness, and 42 (40%) did not. Fifty-four (51%) offered clinical rotations in programs that specifically serve homeless people.
Program Location and Size
t1 shows the location (urban, suburban, or rural) and size of the 106 residency programs responding. Urban programs were significantly larger (average size=30.2 residents) than rural ones (average size=18.1) but not when compared with suburban programs (average size=22.9) (ANOVA DF=2, F=4.5; p=0.02). Program location did not predict clinical or didactic activity on the topic of homelessness. However, programs with such offerings were significantly larger (independent samples t test, df=104, p=0.02).
Mandatory or Elective Training
Among the 64 programs that did have mandatory or elective didactic or clinical training in homelessness, the 50 urban programs were the largest, with an average number of 32.6 residents per program (ANOVA F=2.6, df=2, p=0.05). Fifty six of the 64 programs with any training had at least 1 hour of didactics through residency (mean=5.0 hours over 4 years), and 54 offered clinical rotations.
t2 shows data among the 54 programs that offered clinical rotations with homeless persons and had at least one resident rotating. There were two rural programs with clinical rotations, and both had fewer than 20% of their residents rotating. Similarly, suburban programs averaged below 20%, and urban programs averaged between 21% and 40%. Although 40% of programs endorsed the statement that residents rotated in mandatory clinical rotations, apparently not all residents participated in these experiences, as most programs actually had fewer than 20% of residents in mandatory rotations working with homeless individuals.
Location of Training and Supervision
Most residents rotated in shelters, drop-in centers, or on street outreach teams. Among all programs, supervision occurred at the clinical site at a rate of 81.5%, as opposed to on campus (54%). Only 16% (N=9) of programs with rotations used exclusively on-campus supervision.
Relatively few (30%, N=16) rotations were funded by the clinical site as opposed to an alternative source, and there was a trend toward residency-based funding (56%, n=30, χ2=3.63, df=1, p=0.06). Nineteen percent (N=10) of respondents indicated an "other" form of funding. In eight of these cases, the source was through a local, state, or federal government program, while one was through a volunteer program.
Reasons for Lack of Training on Homelessness
A total of 42 residency programs offered neither didactics nor clinical experiences, and t3 shows frequencies of the reasons this occurred. The most frequently cited reason was "low program priority" (50%, N=21). A regression analysis revealed that only "low community prevalence" of homelessness was significantly associated with "low program priority" (beta=0.550, t=3.49, p=0.001). Of the 38 responses concerning low community prevalence, all four rural programs endorsed this reason, as did five of seven suburban programs, while eight of 28 urban residency programs also noted this.
Finally, in addition to noting "other" in t3, respondents were invited to offer additional comments. Among the 11 respondents who provided commentary and had no offerings in their programs, five stated that residents typically care for homeless persons in other settings such as in emergency rooms, acute inpatient services, or state hospitals. Similar statements were made by three of the 56 programs only offering didactics.
This survey shows that American psychiatry residency programs have attempted to respond to homelessness, in that 60% of programs that participated reported didactic and/or clinical training rotations specifically dedicated to this clinical population. Those programs that had training usually offered didactics as well as clinical rotations, and some were exceptionally comprehensive. However, only six (11%) programs that had clinical experiences rotated all residents. Some respondents spontaneously commented that residents encounter homeless patients in a variety of routine clinical environments such as emergency rooms, though not in dedicated rotations. Although working with homeless patients within a general clinical experience has value, we believe that trainees absorb more when they learn from focused teaching about any population, minimizing the possibility that they will miss important clinical nuance in the "white noise" of generic services or in classes; for instance, covering general cultural competence. Positive experiences focusing on a single population help residents dispel negative attitudes (13), and through focused care of homeless people, residents acquire generalizable skills in clinical engagement and in sophisticated diagnosis and treatment. Furthermore, because most homelessness programs are community-based and usually not medical in tradition, residents learn about systems of care and how to negotiate multidisciplinary rehabilitation, thus encouraging them to work with homeless populations or pursue other career opportunities in the public sector. In one urban program, senior residents rotate part-time for 6 months at one of a number of sites, where they provide treatment, participate in therapy groups, conduct in-service education, and perform outreach. Residents articulate that they have unique value at these sites, and initial outcomes indicate an increase in the number of residents entering public psychiatry fellowships (4).
Our data do not directly examine why few residents have clinical experiences with the homeless, but we can draw some inferences that warrant further exploration. First, although faculty instructors have gained experience with mentally ill homeless people over the past 20 years, there may be a limited corps of community psychiatrists in academia. Consistent with this hypothesis, two programs without offerings spontaneously commented that there were no faculty members adequately expert in the population. Another program that rotated residents noted limited teaching availability. It is notable that supervision, when available, usually occurred at the clinical site. The low frequency of exclusively on-campus supervision may reflect that supervisors must be well acquainted with the unique characteristics of a community-based site and the need for on-site availability of attending physician staffing. It may also reflect the limited number of community psychiatrists on campus and the limited funding available for academic psychiatrists to work at community sites. Psychiatric training guided by a public health perspective is becoming increasingly important, and there must be a better understanding of whether community psychiatric faculty is adequately recruited to meet this and other related training needs.
Programs that stated why they were unable to launch training initiatives cited "low program priority," which was frequently linked to the "lack of community prevalence" of homelessness. Rural and, to a slightly lesser extent, suburban programs cited low prevalence as a reason. Rural programs also had the least absolute percentage of residents with clinical experiences in homelessness programs. The few studies that have explored rural homelessness suggest that there exists a lower prevalence of homelessness in rural areas compared to urban areas (14). However, there is active discussion about how homelessness in the countryside is defined, thus affecting counting methodology (15). Additionally, many experts regard homelessness as more "hidden" in rural areas. Some programs have made attempts to address this issue. For example, residents will accompany an outreach team at a rural site. In one instance, an outreach team traveled to a wooded area in an off-road vehicle, where a homeless family had set up an encampment. The team brought food, helped the family prepare a meal, and provided concrete services, and the resident performed clinical assessments and gave free samples of medication.
Funding may be another barrier that reduces resident clinical experience treating the homeless population. Although insufficient funding was not frequently cited among those programs that had no training in homelessness, programs that offered rotations did so without dedicated funding. About one-quarter of the programs that rotated residents had government support. Remarkably, only a minority of community-based sites actually funded residents who perform important clinical service and receive training. This may be related to a historical separation in most localities between community programs and academic medicine. It appears that neither community-based organizations nor residency training programs have sufficiently realized the opportunities that exist in collaborating toward the goal of improving job opportunities for young psychiatrists in community-based organizations. Creating incentives to share resources may encourage more residency rotation in community-based organizations as well as improve overall care. A more collaborative incentive is direct funding from community-based organizations for desperately needed psychiatric support, in return for committed service linkages from medical centers and an easy tap into faculty for consultation and training of other staff.
In drawing conclusions about residency training in homelessness, it is important to keep in mind that this exploratory study has limitations. First, a 60% response rate yielded a relatively small sample size for subsample analyses, and the nonrespondents’ data might have changed some conclusions if many more responses had been received. Furthermore, the sample drawn from AADPRT may not be fully comprehensive, although it is likely representative of American psychiatric residencies. This was also a deliberately brief survey, sacrificing detail and accentuating its exploratory nature in order to optimize the response rate from busy residency directors. Nevertheless, we believe that this is the first survey of its kind, and it sheds an initial useful light on how training in an important public health issue is managed in today’s psychiatric residencies. In future work, we hope to develop a training curriculum with measurable "competencies," (16) which reflects the body of knowledge necessary for proper psychiatric care for homeless persons.
The authors thank Susan Deakins, M.D., and members of the APA Committee on Poverty, Homelessness, and Psychiatric Disorders (Drs. Ariel Dalfen, Joel Feiner, Leslie Horton, Rajendra Morey, Raman Patel, and Manoj Shah) for their help in developing the survey instrument and also thank Edwin Morales for his assistance in transmitting the survey and collating results.