Homeless individuals generally have significant healthcare needs (1–3), yet often have great difficulty in obtaining health care, as evidenced by their high utilization of emergency departments (4). Previous studies have suggested that doctors’ lack of specific training in caring for the homeless population and negative attitudes toward homeless persons can create barriers to accessing healthcare for the homeless patient (5). Little formal instruction on the causes of homelessness or the care of individuals experiencing homelessness takes place in medical student education in the United States. A search of the Association of American Medical Colleges Curriculum Management & Information Tool revealed that only 21% of medical schools have formal instruction about the homeless (6). A survey of psychiatric residency programs found that 51% offered clinical experiences that specifically serviced homeless individuals, but only 11% of these were mandatory (7).
As for attitudes toward homeless persons, a 2003 study (8) found that medical students hold more negative attitudes toward homeless people at the end of their medical school curriculum than at the beginning. It was hypothesized that at least some of the change appeared to be due to “negative” contact experiences with homeless patients during the clinical years and with doctors who viewed homeless people as less worthy of medical care. A recent study characterizes medical student attitudes toward poor people as becoming more negative, cynical, and conservative as training progresses (9). These authors point out that when these beliefs combine with individualistic attributions toward poverty, they have the potential to influence the care provided. In comparison to medical students, residents’ attitudes change toward homeless persons may indicate a different trend. Buchanan and colleagues (10) showed that residents’ attitudes could become more favorable during their training. In their study, primary-care residents who completed a 2-week rotation in homeless healthcare had higher beliefs in societal causes of homelessness and more comfort with homeless people (10).
No research to-date has examined the attitudes of students’ pre- and post-specific clerkship experiences with faculty and residents who teach them to examine the effects that the hidden curriculum may have in the development of student attitudes toward the patient populations (11–13). Evaluations at our institution by third-year medical clerks after an ambulatory experience in a clinic serving homeless individuals prompted the current investigation of the attitudes of medical students toward individuals experiencing homelessness. Although students’ comments about the rotation were uniformly positive, they revealed that students had significant knowledge deficits in the causes of homelessness before the rotation. This seemed most prominent in the broader socio-economic causes of homelessness, as over 90% of the students indicated that substance abuse and mental illness are major reasons for homelessness, whereas fewer than 50% felt that unaffordable housing was a major factor (14). Since fewer than 20% of our medical students rotate at a homeless clinic, we wondered about the broader attitudes of medical students toward this patient population.
The purpose of the current study was to explore changes in medical students’ attitudes toward homeless persons during their psychiatry and emergency medicine clerkships. Simultaneously, this study explored current attitudes toward homeless persons held by Psychiatry and Emergency Medicine residents and faculty in an attempt to uncover the “hidden curriculum” in medical education, in which values are communicated from teacher to student outside of the formal instruction. Psychiatry and Emergency Medicine clerks were selected as the sample population because we believed that these two rotations would provide the most exposure to patients experiencing homelessness. Sampling these two rotations also allowed for the possibility that items relating to mental illness and physical health care might be differentially affected by rotations, which emphasized these areas.
Participants were 145 students (79 third-year students, 66 fourth-year students) and 72 residents and faculty (31 Psychiatry residents and faculty; 41 Emergency Medicine residents and faculty) at one United States community-based medical school. Response rates were 98% for third-year students and 100% for fourth-year students. For faculty and residents, response rates were 46% for Psychiatry and 72% for Emergency Medicine.
The Health Professionals’ Attitudes Toward the Homeless Inventory (HPATHI) consists of 23 questions with Likert-type responses ranging from 1: Strongly Disagree to 5: Strongly Agree. The HPATHI has been shown to have an internal consistency reliability (Chronbach’s alpha) of 0.88 and a test–retest reliability coefficient of 0.69 (15)
From September 2007 through June 2008, students in their third year and students in their fourth year of medical school in a midsized, midwestern city were surveyed, using the HPATHI, regarding their attitudes toward homeless persons. Students were invited to participate in this study by a faculty member in Psychiatry or in Emergency Medicine at the beginning of their respective clerkships. At our medical school, Psychiatry is a 6-week rotation completed in the third year of medical school, and Emergency Medicine is a 1 month rotation completed in the fourth year of medical school; neither rotation has a formal curriculum on homelessness. However, both of these services provide care to a significant number of uninsured patients, many of whom are homeless. The HPATHI was administered at the beginning and end of the respective clerkships, thus sampling two different cohorts. The HPATHI was also administered to Psychiatry residents and faculty (via campus mailbox) and Emergency Medicine residents and faculty (at Grand Rounds) one time during the course of this study. Informed consent was obtained, and this study was approved through the Institutional Review Board at our university. Permission was granted to use the HPATHI for purposes of the present study (personal communication; D. Rochon, June 25, 2007).
Items on the HPATHI assess attitudes toward homeless persons by measuring an individuals “level of interest and confidence in their ability to deliver healthcare services to the homeless population” (15). For the psychiatry clerkship, t-tests for matched pairs (p <0.05; Bonferroni correction: p <0.002) to determine whether changes in attitudes toward homeless persons occurring in the Psychiatry and Emergency Medicine clerkships showed statistically significant differences for two items. Effect sizes were typical or medium. After completing their Psychiatry clerkship, students were more likely to agree with the statements “Most homeless people are mentally ill;”(t[78] = −4.235; p <0.001; δ=0.485) and “I am comfortable being a primary-care provider for a homeless person with a major mental illness;” (t[78] = −6.009; p <0.001; δ=0.673). For the Emergency Medicine students, no statistically significant differences were noted after their clerkships.
Current attitudes toward homeless persons held by Psychiatry and Emergency Medicine residents and faculty were also investigated. An analysis of variance used to determine whether attitude differences exist between the specialties (p <0.05; Bonferroni correction: p <0.002) revealed statistically significant differences on 7 out of the 23 survey questions. Effect sizes ranged from larger-than-typical to much-larger-than-typical. Means and standard deviations (SD) for statistically significant items are reported in Table 1. As compared with Emergency Medicine faculty and residents, Psychiatry faculty and residents indicated more agreement with the following statements:
“More healthcare dollars should be directed toward serving the poor and homeless.” (F [71]=23.97; p <0.001; δ=1.18).
“Doctors should address the physical and social problems of the homeless.” (F [71]=11.28; p=0.001; δ=0.809).
“I am interested in working with the underserved.” (F [71]=9.98; p=0.002; δ=0.752).
“I enjoy addressing psychosocial issues with patients.” (F [71]=129.58; p <0.001; δ=2.80).
“I enjoy learning about the lives of my homeless patients.” (F [71]=62.77; p <0.001; δ=1.82).
“I believe social justice is an important part of health care”. (F [71]=25.66; p <0.001; δ=1.23).
As compared with Psychiatry faculty and residents, Emergency Medicine faculty and residents indicated more agreement with the statement: “Homeless people choose to be homeless.” (F[71]=10.59; p=0.002; δ=0.785).
Although students’ pre-rotation scores demonstrated a fairly good understanding of the problems leading to homelessness, few attitudes about homelessness were changed during these rotations. Only those students on the Psychiatry clerkship showed any movement in their beliefs. Given that Psychiatry rotations are completed in the third year of medical school and Emergency Medicine in the fourth at our medical school, it is possible that the lack of change noted over the course of the Emergency Medicine rotation arose from this difference in the amount of clinical experience the two groups had acquired. That is, the more clinical exposure students have had, the more solidified their beliefs about patients and healthcare have become and the less malleable to change. In this regard, our findings are consistent with other studies, which have described hardening of medical student attitudes as their education progresses (16), with attitudes toward people experiencing homelessness becoming more negative. Also, the Emergency Medicine clerks were more likely to see homeless patients on single occasions, whereas the Psychiatry clerks more often had opportunities to work with homeless patients over a longer period of time. Perhaps the Psychiatry clerks were able to get to know their patients better and thus grew more comfortable in providing care to the homeless-persons population.
It is not surprising that the item that most directly elicits comfort with working with people with mental illness, albeit as a primary-care provider, showed the most change during the psychiatry rotation. Perhaps on these rotations more attention is paid to eliciting information about the patient’s social history, including such issues as education and work history, entitlement income and insurance coverage, leading to a broader consideration of access to healthcare. Last, the lack of dramatic changes in student attitude over these rotations could be due to the fact that in neither rotation was a formal effort made in classroom didactic curricula or in the clinical experience to review and discuss the risk factors for people becoming homeless, the characteristics (personal and clinical) of homeless individuals, or any broader social issues of access to healthcare and housing. Psychiatric residents who provide much of the teaching, both formal and informal, to medical students on the inpatient units, are in their first and second year of training at our school, and do not have courses covering these issues until their third year. Except for the few students assigned to the Homeless Clinic for their outpatient psychiatry rotation, most students would have lacked both the concentrated experiential learning and instruction in caring for people who are homeless.
Regarding resident and faculty results, significant differences between specialties were noted. Psychiatric residents and faculty were more likely than Emergency Medicine physicians to believe that more healthcare dollars should be directed to poor and homeless patients, that doctors should address the physical and social problems of homeless persons, endorse an interest in working with underserved groups, enjoy addressing psychosocial issues and learning about the lives of homeless people, and believe that social justice is an important part of healthcare. Emergency Medicine doctors were more likely to believe that people chose to be homeless. Possibly, Emergency Room physicians do not get to know their patients as intimately as psychiatrists do; thus, they do not understand the challenges in the lives of their patients in the same way. Some of these differences may track with differences in attitudes reflected in society, such as more sympathetic attitudes being linked with being younger, female, liberal, and less wealthy (17).
Results of this study should be interpreted in light of its limitations. First, the fact that this is a single-institution study may limit the generalizability of its findings. Second, because the Emergency Medicine clerks had completed the Psychiatry rotation in the year before this study was conducted, threats to internal validity are present in that earlier participation in the Psychiatry clerkship could confound the results for the Emergency Medicine clerks. Third, it is possible that the method for data collection (via campus mailbox versus Grand Rounds) may have confounded the results for residents and faculty in the specialties of Psychiatry and Emergency Medicine. Furthermore, response rates differed between the groups, a factor that could also affect the results.
This study gives a snapshot between medical specialties and across levels of training in attitudes toward homeless persons. Medical providers’ attitudes toward people experiencing homelessness is not merely an academic issue, but affects whether patients feel welcomed and understood by their doctors (18). The introduction of more formal instruction on these topics linked with a clinical experience may enhance positive changes in students’ attitudes toward individuals experiencing homelessness.
Buchanan and colleagues described a rotation in homeless healthcare that showed improvements in attitudes and included clinical experiences, classroom lectures, nonclinical exposure to homeless servicing programs, and reflective and experiential learning activities (10). Nursing education research has produced several descriptions of attitude changes in students after a clinical experience serving homeless individuals by use of methods including focus groups, attitude surveys, and interviews (19–21).
Wear and Kuczewski’s (9) recommendations to help improve medical students’ attitudes toward poor patients include increasing the socioeconomic diversity of the physician workforce to increase the diversity of perspectives; increasing trainees’ empathetic understanding of poverty through a variety of efforts; and increasing the number of positive role-models for trainees. Another opportunity to be further explored would be pairing faculty teaching with student-developed, community-service learning activities working with homeless people, similar to those described by Brown and colleagues (22). Perhaps, linking didactic instruction with formal processing of affectively satisfying service-learning experiences, would be a novel way of enhancing attitude change and empathy without putting further demands on the medical school calendar.
Given that medical student competencies should be addressing the broader social issues of homelessness, disparities in health, the cost of healthcare, access to medical care for uninsured patients, and treating underserved populations, medical schools need to first understand the attitudes of medical students to such issues, and then develop curricula to overcome stigma. Although our study contributes to the literature and moves this discussion forward, much work remains for medical educators.