Although interest in the training of ethnic and gender issues in psychiatry has recently flourished, medicine, since antiquity, has attempted to explain perceptions of cultural differences in behavior from a variety of theoretical perspectives. Hippocrates sought explanations for ethnic variations in traits such as fearfulness based upon local climate (1), while Galen believed that non-Greek barbarians were prone to emotional dyscontrol in contrast to the more "rational" Greeks (2). Since barbarians were socially removed, as it were, from Greek men, it is not unexpected that the prevailing psychological models would all portray the "other" as unfamiliar or uncontrollable. The field of psychiatry has evolved in part from such cultural roots. Therefore, it has been necessary and fruitful for psychiatrists to investigate models for treating patients of backgrounds demographically distinct and different from themselves (3,4). For example, there are issues of culture-bound syndromes, such as "nervios" (nerves) or "atagues de nervios" (attacks of nerves) among Puerto Ricans, that are culturally meaningful expressions of distress that need to be understood in treating enculturated patients whose background differs from that of the psychiatrist (5). Language differences are a frequent source of miscommunications and misunderstandings between Hispanic patients and their non-Hispanic clinicians (6,7). Similarly, the overrepresentation of African American adults involuntarily committed to public mental health institutions raises questions about whether many mental health professionals' commitment decisions are based on misunderstandings of relevant standards of behavioral functioning (8). As another example, in metropolitan state mental hospitals in Chicago, being African American was found to be predictive of being diagnosed with schizophrenia, even after diagnostic criteria were statistically controlled for (9), again suggesting that ethnic misunderstandings by mental health professionals may lead to misdiagnosis. Also, there may be ethnic differences in the severity and comorbidity of psychiatric problems (10), adding to the challenges faced by psychiatrists in diagnosis and treatment of patients from different ethnic backgrounds.
Teaching of models for understanding patients of ethnically different backgrounds is now being included in many psychiatry residency programs, although findings from cultural psychiatry have not been fully integrated into the practice of psychiatry (11), and basic research on such topics as misdiagnosis in minority populations is still needed (12,13). Therefore, this teaching occurs in the face of inadequate, but not inconsiderable, knowledge about ethnic issues relevant to psychiatry.
Similarly, training residents about gender issues relevant to diagnosis and treatment in psychiatry is also occurring despite a need for more research in this area. Findings similar to those on ethnic minorities have been found for women in psychiatry, such as overrepresentation of women as psychiatric inpatients (14) and disproportionate diagnosis among women, compared with men, in specific diagnostic categories, such as dissociative identity disorder, mood disorders, and eating disorders (15,16), as well as gender bias in psychiatric prescription practices (17). There is scant published information addressing the training of psychiatrists on gender issues. Shapiro (18) has recommended that residents training in child and adolescent psychiatry be invited to participate in research on issues such as the effects of core gender identity vs. sex assignment to improve the ability of the residents to conduct rigorous and applicable research. Others have recommended that psychiatric training should be strengthened as early as in medical school with an emphasis on gender and ethnic minority issues (19).
Given the need to train residents in both ethnic and gender issues in psychiatry, it is important to begin to appraise residents' adequacy of training in these areas. We describe 1) the development of a measure to assess appraisals of residents' adequacy of training in ethnic and gender issues in psychiatry and 2) the results obtained by using this measure in a national survey of residency training directors.
There has been a tremendous growth in the ethnic minority population in the United States, reaching 24.4% in the 1990 census (20). Thus, the U.S. population has great ethnic diversity, as well as being nearly evenly split by gender. However, training in psychiatry on ethnic and gender issues is in its infancy. In the past, psychiatry as a field has been accused of ethnic and racial discrimination, both in patient diagnosis and treatment (21), some of which unfortunately still exists today. Increased attention is being paid to issues of diagnosis, etiology, and treatment with patients from diverse ethnic backgrounds (22—28), as well as increased consideration of these issues in terms of gender differences (15,16,29,30) or consideration of gender and ethnicity together (31—32). However, further understanding is needed of the impact of ethnicity on psychiatric diagnosis and treatment (33). As ethnic diversity increases in the United States, we will have more minority medical students who may choose psychiatry as a career (34) and psychiatrists will see more patients from ethnic backgrounds different from their own (35,36).
Little research has been done to assess the status and needs of psychiatry training in the areas of ethnicity and gender. This study had four main objectives: 1) to develop a questionnaire measure of residents' adequacy of training on ethnic and gender issues and their experiences in working with patients of minority ethnic backgrounds, 2) to use this questionnaire to assess training directors' perceptions of their residents' adequacy of training, 3) to examine training program factors associated with appraised adequacy of training and experiences, and 4) to determine the desirability of specific types of training materials and programs for teaching ethnic and gender issues.
We hypothesized that adequacy of training on ethnic and gender issues would be positively associated with having residency coursework on these topics, as this is the rationale for providing such programs. Furthermore, we predicted that residency training directors who were female or an ethnic minority would be more likely to appraise the residents more negatively in their adequacy of training on ethnic and gender issues, because we suspected that persons who are female or an ethnic minority may be more critical about these issues because of the salience of these issues to their lives personally. It has been observed that that women working in psychiatry may be more likely to experience gender-specific problems attributable to entering a predominantly male profession (37), which we thought should help to sensitize psychiatry training directors to the need to ensure adequate training of their residents about gender issues (38). We thought a parallel process would also occur for psychiatry training directors who were an ethnic minority. We also wanted to explore the possibility that differences in training directors' overall perceptions of their residents' adequacy of training in working with ethnic and gender issues might differ partly by age, possibly because of changes in societal attitudes about ethnic and gender issues that might have helped to shape the views of training directors as to what is appropriate, although we did not have a directional hypothesis about this relationship.
There were two additional variables hypothesized to be positively associated with adequacy of training in working with ethnic issues. First, we predicted that this adequacy of training would be positively associated with the proportion of residents in each training program who are ethnic minorities, particularly given the changing ethnic composition of psychiatric residents (33). Second, we hypothesized that this adequacy of training would be positively related to the amount of contact that residents in the program have with patients of diverse ethnic backgrounds, as has been proposed in the literature on psychiatric education (35,36,39).
The sample was drawn from the Graduate Medical Education Directory 1994—95 (40), which describes all accredited psychiatry residency training programs in the United States. On October 14, 1994, a questionnaire designed to assess training needs in ethnicity and gender was mailed to all 200 programs listed; however, representatives of four institutions notified us that they no longer had a residency training program in psychiatry or only offered a fourth-year research elective with no regular training program. Follow-up mailings were sent on November 17 and December 14 to residency training directors who did not respond to the first mailing. Of the remaining 196 programs, questionnaires were completed by 142 respondents, a 72.4% return rate. Informed consent was ensured by using wording in the cover letter accompanying the questionnaire that was approved by the Stanford University Panel on Human Subjects.
We developed a 43-item questionnaire that was used in this study. This measure assessed 1) training program directors' appraisals of their residents' adequacy of training in working with ethnic and gender-related issues in psychiatry, 2) their residents' experience in working with patients of diverse ethnic backgrounds and the likelihood that the mix of the ethnic backgrounds of their patients would change in the next 10 years, 3) their residents' ethnic backgrounds, 4) the usefulness of specific kinds of teaching materials, 5) their views toward requiring residency coursework in ethnic and gender issues in psychiatry, and 6) personal background characteristics of the respondent. These specific measures are described more fully next.
Adequacy of Training in Working With Ethnic and Gender Issues in Psychiatry.
Thirteen items assessed training program directors' appraisals of their residents in working with patients in ethnic and gender issues. The Adequacy of Training on Ethnic Issues Scale, which assessed appraisals of residents' adequacy of training in managing ethnic issues, included eight items (e.g., "Establish rapport with patients of diverse ethnic groups"). The Adequacy of Training on Gender Issues Scale, which assessed appraisals of residents' ability to work with patients on gender-related issues, included five items (e.g., "Understand gender stereotypes in therapist—patient interaction"). Responses to items in both scales were recorded on a 1—3 scale, with "1" indicating not having been trained and "3" indicating having been adequately trained. Scales were scored by summing the values of the items. Internal consistency was satisfactory for each of these scales: for the Adequacy of Training on Ethnic Issues Scale, Cronbach's alpha=0.88; for the Adequacy of Training on Gender Issues Scale, Cronbach's alpha=0.79.
Experience Working With Ethnic Minority Patients.
Eight items assessed experience with patients of specific types of ethnic backgrounds: African American and Caribbean American (Black), Asian-American or Asian (e.g., Chinese, Japanese, Korean, Southeast Asian), "European American or European (White), Latino (e.g., Mexican American, Puerto Rican, Cuban, Central or South American), Native American (American Indian), Pacific Islander (e.g., Filipino, Native Hawaiian, or Samoan), Middle Eastern American or national of Middle East country, and immigrants and refugees. Responses were recorded on a 1—3 scale, with "1"="no or very little experience," "2"="some experience," and "3"="very much experience." The Experience Working With Ethnic Minority Patients score was computed by summing the scores across the six items assessing experience with patients of African American and Caribbean American, Asian-American or Asian, Latino, Pacific Islander, Middle Eastern, or Native American (Indian) backgrounds.
Usefulness of Training Materials in Ethnic and Gender Issues in Psychiatry.
Six items were developed to assess respondents' perceptions of the usefulness of specific kinds of training materials. Three items assessed the usefulness of materials for teaching ethnic issues in psychiatry: "A syllabus of recommended readings," "Books/articles," and "Videotapes/other audiovisual materials." Three parallel items assessed the usefulness of these same kinds of training materials for teaching gender issues in psychiatry. Responses were recorded on a 3-point scale: "1"="if no or very little help," "2"="if some help," or "3"="if very much help."
Views About Requiring Training in Ethnic and Gender Issues in Psychiatry.
Two items assessed opinions about when a course in ethnic and gender issues in psychiatry should be offered, one item about ethnic issues, and the other about gender issues. In each item, the respondent was asked to choose one of the three following answers as a basis for offering this training: "as an elective only if a faculty member is conducting work in this field," "as an elective in every psychiatry training program," or "as a required core curriculum course."
Percentage of Residents of Minority Ethnic Background.
Ten items were used to asses the ethnic background of residents in the respondent's training program. Eight items used the identical ethnic categories as were used in the assessment of Experience in Working With Patients of Diverse Ethnic Backgrounds. In addition, the categories of "Other" and "Unknown" were included. For each ethnic category, the respondent was asked to indicate the number of residents currently in the respondent's training program. This scale was scored by summing all of the numbers of residents reported across categories and then dividing this into the sum of the numbers of residents reported across all of the ethnic backgrounds, except "European American or European (White)," yielding a percentage of the residents who were an ethnic minority.
Respondent's Background Characteristics and Program Description.
Finally, three items assessed the respondents' ethnic background, gender, and age. Two items asked whether the training program offers one or more courses that specifically teach about ethnic and gender issues in psychiatry, scored as "1" if the program did not include this training and "2" if it did. These responses yielded two variables: one labeled "Residency Coursework in Ethnic Issues," the other "Residency Coursework in Gender Issues." Also, the questionnaire asked the respondent to estimate the likelihood (from "1"="not at all likely" to "2"="somewhat likely" or "3"="very likely") that with changes in access to health care that the demographic mix (i.e., ethnic minority) of the patients seen by the residents will change in the next 10 years.
Primarily, the data were analyzed as frequencies. In examining the appraisals of residents' adequacy of training in working with ethnic and gender issues, multiple regression analysis using the simultaneous procedure was conducted on the Adequacy of Training on Ethnic Issues Scale and the Adequacy of Training on Gender Issues Scale. For the analysis of the Adequacy of Training on Ethnic Issues Scale, the independent variables were "Residency Coursework on Ethnic Issues"; "Percentage of Residents of Ethnic Minority Background"; "Experience Working with Ethnic Minority Patients"; and training director's age, ethnicity, and gender. For the analysis of the Adequacy of Training on Gender Issues Scale, the independent variables were "Residency Coursework on Gender Issues" and the training director's age, ethnicity, and gender. Throughout these analyses, when there was missing data, the results were computed on the available data; therefore, because of missing data, the sample size varied slightly across the analyses for different variables. The statistic used to describe the statistically significant relationships in these multiple regressions was the beta, the standardized estimate of the coefficient.
Description of the Respondents and Their Programs
The background characteristics (ethnicity, gender, and age) of the training directors who responded to this survey are summarized in T1. The majority described their training programs as currently providing training on issues of ethnicity (82.3%, n=117/142) and gender (71.1%, n=101/142). The mean percentages of residents of various ethnic backgrounds in their training programs were 61.2% European American or European; 12.9% Asian-Americans or Asians; 7.6% Middle Eastern Americans or nationals of the Middle East; 7.5% Latinos; 6.0% African Americans, Caribbean Americans, or Africans; 3.8% Pacific Islanders; and 1% Native Americans.
Assessment of the Adequacy of Training on Issues of Patient Ethnicity and Gender
T2 shows the training directors' appraisals of their residents' training needs on issues of patient ethnicity and gender analyzed by frequencies and percentages on the individual items included in the Adequacy of Training on Ethnic and Gender Issues Scales.
Experience Working With Ethnic Minority Patients
T3 shows the appraised experience of the residents in the current training program in working with patients of diverse ethnic backgrounds. Examination of this table shows that substantial percentages of the residents are reported to be receiving little experience in working with patients from a number of ethnic backgrounds, particularly Asian-American or Asian, Latino, Native American, Pacific Islanders, Middle Eastern, or immigrant backgrounds. Furthermore, half (51.4%, n=73/142) predicted that it was "somewhat likely" or "very likely" that with changes in access to health care the demographic mix of the patients their residents will change in the next 10 years.
Variables Related to Adequacy of Training on Ethnic Issues and Adequacy of Training on Gender Issues
We found that three variables were significantly and positively associated with scores on the Adequacy of Training on Ethnic Issues Scale: "Residency Coursework Training on Ethnic Issues" (Beta=0.35, t=4.61, P<0.0001); "Experience Working With Ethnic Minority Patients" (Beta=0.32, t=4.28, P<0.0001); and "Percentage of Residents of Minority Ethnic Background" (Beta=0.18, t=2.38, P<0.02), with these three variables accounting for a third of the respondents' differences in their appraisals of their residents on the Adequacy of Training on Ethnic Issues Scale (overall adjusted R2=0.339, F(6,121)=11.84, P<0.0001). We found two variables to be significantly associated with residency training directors' appraisals of their residents on the Adequacy of Training on Gender Issues Scale: "Residency Coursework on Gender Issues" (Beta=0.35, t=4.30, P<0.0001) and the residency director's age, with the younger residency training directors found to be more likely, compared with the older residency training directors, to appraise their residents lower on the Adequacy of Training on Gender Issues Scale (for age, Beta=0.19, t=2.28, P<0.05). These two characteristics accounted for about 15% of the differences on the appraisals of residents on the Adequacy of Training on Gender Issues Scale (overall adjusted R2 =0.155, F(4,126)=6.94, P<0.0001), with the younger residency training directors appraising their residents' adequacy of training on gender issues more negatively than did the older residency training directors. No other relationships were significantly associated with appraisals of residents on the Ethnic or Adequacy of Training on Gender Issues Scales.
Desire for Coursework and Materials on Ethnicity and Gender
The majority of residency training directors indicated that the basis for a course in ethnic issues should be as a required core curriculum course (86.8%, n=118/136), whereas 11.8% (n=16/136) indicated that it should be offered as an elective in every training program. Only 1.5% ( n=2/136) indicated that it should be offered as an elective only if a faculty member is conducting work in this field. Similarly, the majority of residency training directors indicated that the basis for a course in gender issues should be as a required core curriculum course (83.7%, n=113/135), whereas 12.6% (n=17/135) indicated that it should be offered as an elective in every training program. Only 3.7% (n=5/135) indicated that it should be offered as an elective only if a faculty member is currently working in this field. The results shown in T4 indicate that the majority of the respondents reported that it would be "very much help" to have educational materials (a syllabus, books/articles, and videotapes/other audiovisual materials) made available to their residency training program. Most of the other respondents rated each of these materials of "some help." The respondents reported that gender- and ethnicity-related educational materials would be helpful to their program.
This study developed a measure that may be useful in future research assessing the adequacy of training in working with patients of different ethnicities, minority patients, and those of the opposite gender from the care provider. This recommendation is suggested by the findings of good internal coherence in the items included in each of the Adequacy of Training on Ethnic and Gender Issues Scales. This study's findings of significant relationships of these scale scores with some of the hypothesized variables further suggest that these measures have the capacity to predict variation among programs. These findings suggest that these measures have satisfactory psychometric properties for assessing perceptions of the adequacy of residency training programs in educating residents on issues of ethnicity and gender in psychiatry. Furthermore, the high level of internal consistency found among the items in the Adequacy of Training on Ethnic Issues Scale implies that a didactic model founded on teaching basic skills in cross-cultural methods may be as important as teaching the specific factors relevant to mental health for each ethnic group studied.
This study's results suggest that through exposure to other groups at three levels of proximity: as colleagues, as patients, and as didactic exemplars, psychiatric residents can enhance their clinical education and their ability to treat those ethnically different from themselves (3). The Adequacy of Training on Ethnic Issues Scale is strongly related to the degree of contact that residents have had with ethnic minority patients. These programs varied in the amount of contact with patients of specific ethnic backgrounds reported by the training directors. Furthermore, the majority of the training directors reported that their residents have very little experience with patients who are Native Americans or Pacific Islanders, and only some or little experience in working with Asian-Americans/Asians, Latinos, Middle Eastern nationals, and immigrants and refugees. These data confirm the judgment that certain residency programs are unable to provide residents with contact with patients of particular ethnic backgrounds, perhaps because of the limited ethnic diversity of the local population (35—36,39). These results are similar to the findings of a recent survey of clinical and counseling psychologists (41), which also found that the greatest patient contact reported was with European Americans, and many reported little contact with patients from a number of other ethnic backgrounds.
Furthermore, the Adequacy of Training on Ethnic Issues Scale demonstrates a moderate relationship with the percentage of residents of ethnic minority background in the residency program. These results are not proof of a cause-and-effect relationship between contact with ethnic minority residents and patients during residency and the residents' levels of comfort in handling cross-cultural issues by the end of residency, but these results do suggest that such relationships may exist, and are consistent with others' recommendations (35,36,39). However, interpretation of these results is limited by our methodology. We did not examine whether residents saw ethnic minority patients who were of similar ethnic backgrounds to themselves, which would involve intra-cultural understanding, rather than seeing patients who were dissimilar in ethnic minority backgrounds, which would involve inter-cultural understanding. Furthermore, the wording on the response options for the Adequacy of Training on Ethnic and Gender Issues Scales was phrased to assess adequacy of training on the positive end of the response; however, at the lower end, there is some potential ambiguity because of the wording focusing on little formal training in the area. Future use of this survey should substitute "inadequate training" for "little or no formal training" as the option for the opposite end of the response.
This survey of perceptions of residency training directors about the experiences of residents in learning about ethnic issues provides a cross-sectional view. This research strategy cannot adequately address hypotheses about how residency training influences ethnic or adequacy of training on gender issues; adequate research on these hypotheses requires longitudinal studies of the training experience, in which residents themselves are followed over time. An alternative explanation for this study's results is that residents with a high level of comfort with and knowledge of cross-cultural issues may tend to self-select to choose a particular residency program because it offers exposure to ethnically diverse populations of both patients and colleagues. A questionnaire given to residents at the start of residency training and repeated at the end of training could help to examine patterns of change in the adequacy of training on ethnicity and gender over time in different residencies and in different resident populations.
The training directors' survey responses reflect a summation of their assessment of the adequacy of training on ethnic issues of both minority and nonminority residents. Although adequacy of training on ethnic issues was more strongly correlated with experience with ethnic minority patients than with the percentage of ethnic minority residents in the program, it is possible that resident groups with a higher proportion of minority residents appear to residency training directors to corporately embody a higher degree of ethnic adequacy of training. Studies that allow comparison of the adequacy of training on ethnic issues of minority and nonminority residents would offer a more direct assessment of the degree to which adequacy of training on ethnic issues arises through residency training experiences.
Current changes in the structure of mental health service delivery, including the transition toward managed care and the growing biological orientation of psychiatry, are tending to move psychiatrists out of the role of primary therapist and more toward roles as clinical consultants and treatment team leaders. In the future, psychotherapy services are increasingly likely to be provided by social workers, psychologists, and psychiatric nurses, with consultation or supervision of a psychiatrist. Some clinicians might argue that the changing roles of mental health service providers makes understanding of cross-cultural and gender issues less important to psychiatrists, who will offer medication management, diagnostic evaluations, and service administration, whereas it would be more important in educating those who will serve as primary therapists. However, we would expect that the education of psychiatrists about ethnic and gender issues will become more important as psychiatrists begin moving into roles as team leaders and consultants.
We need to develop models for the transmission of knowledge about the role of ethnic and gender issues in the presentation and treatment of psychiatric illness so that residents can become effective teachers, supervisors, and consultants for other mental health professionals. Significant to the pedagogical methods of teaching ethnic and gender issues in psychiatry is the importance of the dialectic between considering the patient as an individual and viewing her or him as a member of an ethnic group or gender category (3). This dual perspective is a necessary tension in developing and providing clinically relevant psychiatric courses on ethnicity and gender: it is as erroneous to overgeneralize about as it is to ignore the influences of ethnicity and gender in diagnosing and treating patients.
The majority of residency training directors surveyed believe that courses in ethnic and gender issues should be offered as required core curriculum courses, which supported the inclusion of education on these subjects that is already included in the majority of residency programs surveyed. Also, the majority of respondents would like to have access to teaching materials for such courses, suggesting the need for widespread availability of such materials. This need will hopefully be addressed by the current curricular efforts of the American Psychiatric Association's Minority and Under-Represented Component Committees. Further research is needed to examine the influence of residency training experiences in increasing residents' understanding of ethnic and gender issues and in improving their clinical proficiencies and comfort in working with patients of diverse ethnicities and minority patients.
This study was made possible by an award from the Irvine Foundation. The authors thank Kristen Cassic and our survey's respondents.