• 8 • Taiwanese Journal of Psychiatry (Taipei) Vol. 31 No. 1 2017 Overview
The Promise of Cultural Epidemiology
Mitchell G. Weiss, M.D., Ph.D.1,2*
An emerging fi eld of cultural epidemiology, rooted in the illness explanatory model framework, has developed integrated quantitative and qualitative research methods to harness synergies of interdisciplinary collaboration in psychiatric epi- demiology and medical anthropology. The formulation of cultural epidemiology presented here may be understood as the epidemiology and elaboration of illness explanatory models. It has been defi ned as the study of locally valid representa- tions of illness and their distribution, motivated by aims to enhance the cultural sensitivity and quality of clinical care, health services and other mental health and global health interests. Refl ection on the academic, clinical and public health con- texts in which concepts and methods of our approach to cultural epidemiology developed, and a review of explanatory model studies of mental health and other health problems provide an opportunity to clarify origins, achievements and chal- lenges for research and anticipated contributions to cultural psychiatry, mental health and global health.
Key words: explanatory models, psychiatric epidemiology, cultural sensitivity, EMIC interview (Taiwanese Journal of Psychiatry [Taipei] 2017; 31: 8-24)
[1], the approach presented in this overview has Introduction been described as an epidemiology of explanatory models [2]. The disciplines of anthropology and epide- In the early 1970s as the fi eld of psychiatric miology are the basic sciences of cultural psychia- epidemiology was establishing itself, leaders of try, and consideration of their complementary re- psychiatry acknowledged the relevance of culture. lationship has motivated and nurtured development Norman Sartorius advised, “The methodology of of the fi eld of cultural epidemiology. Although studies of the inter-relationship between culture initially conceived as an approach to research for and mental disorders needs to be further devel- cultural psychiatry, cultural epidemiology has oped” [3]. But bridging the qualitative ethno- demonstrated broader utility in other areas of graphic interests of explanatory models, as ini- medicine and public health. Rooted in the illness tially conceived by Kleinman, and the quantitative explanatory model framework of Arthur Kleinman epidemiological accounts of the burden of mental
1 Swiss Tropical and Public Health Institute, Basel, Switzerland 2 University of Basel, Basel, Switzerland Received: February 26, 2017; accepted; February 28, 2017 1*Corresponding author. Socinstrasse 57, CH4002 Basel, Switzerland E-mail: Mitchell G. Weiss
health problems, as conceptualized in the fi eld of focus on the ways people understand illness (i.e., psychiatric epidemiology, has been challenging. illness explanatory models) and themselves (i.e., Cultural epidemiology studies locally valid cultural identity of individuals, their families, representations of illness (i.e., key features of ex- communities and other groups they belong to). planatory models) and their distribution [4, 5]. Epidemiological inquiries are primarily quantita- These representations are elaborated by variables, tive and empirical; anthropological studies are descriptions and narratives accounting for the ex- more likely to be qualitative and ethnographic, perience of illness, its meaning and associated ill- and to rely on insights and interpretation of fi eld ness behavior. Qualitative and quantitative re- experience. Both fi elds, however, acknowledge search methods facilitate comparisons and clarify the priority of fi eld research, and the term ‘shoe- the cultural basis of risk, course and outcomes of leather epidemiology’ distinguishes fi eld epidemi- practical signifi cance for clinical practice and ology from database studies, just as ethnography public health. may be distinguished from anthropological study The integration of qualitative and quantita- of intellectual history. tive methods to achieve these ends has been an Bridging the complementary approaches of important contribution towards fulfi lling the empirical epidemiology and experience-based, promise of cultural epidemiology. Both orienta- insight-oriented anthropology enables synergies tions contribute to a practical understanding of the for research that benefi t public health and clinical magnitude, nature and implications of the burden practice. That premise has motivated interdisci- of mental health problems. Clarity in that regard is plinary study and cultural epidemiology for cul- needed both to support advocacy interests that tural psychiatry. Several fundamental concepts in justify strategic priority and funding support, and the fi eld of medical anthropology have been es- to guide culturally sensitive clinical practice and sential considerations in developing the fi eld. relevant public health action. In this overview, we They include the framework of “emic and etic” explain the underlying concepts and methods of orientations for social analysis, the distinction of cultural epidemiology, and examine various in- “disease and illness,” and the conceptual frame- struments (especially EMIC interviews), research work of illness explanatory models. designs and their relationship to complementary clinical tools. Emic and etic Consideration of “emic and etic” orienta- Conceptual Underpinnings tions for social analysis has become a widely ac- cepted way of distinguishing frameworks derived Classical epidemiology, including psychiat- from professional study (etic outsider) and those ric epidemiology, typically documents rates and based on the lived experience of people and com- determinants of selected health problems. It there- munities (emic insider). Kenneth Pike, a linguistic by indicates their priority and hopefully what to anthropologist, introduced the terms in 1954 do about them. Much of the medical literature be- based on extension of complementary phonemic gins with an account of the epidemiology that mo- and phonetic orientations for linguistic analysis tivates and justifi es attention to a particular topic. [6]. A phoneme is a basic unit of meaning within a Clinical interests of medical anthropology often particular language as it is understood by native • 10 • Cultural Epidemiology
speakers. Phonetics refers to study of speech tal illness or local ideas about a medical problem based on a comprehensive collection of elemental constitutes an emic account. The ICD-10 and the sounds of speech derived from professional con- DSM-5 nosologies, on the other hand, are profes- sideration of all languages. sional catalogues that elaborate etic frameworks. Initially, the question of whether to take the emic view as a serious consideration for social Disease and illness analysis was controversial. It was dismissed by In the context of serious efforts by the latter some in disparaging terms as unprofessional, half of the 1970s to construct a robust medical an- sloppy and unsuitable for intellectual study. The thropology, Kleinman asserted that new concep- controversy was addressed in a debate organized tual models, such as that of “disease and illness,” as a symposium at the annual meeting of the were needed to bridge the disjunction of “biologi- American Anthropological Association in 1988 cal reductionist and cultural relativist approaches [7]. By 1990 when the proceedings were pub- ... endemic to anthropological and cross-cultural lished in a book, but widespread use was already studies in the health fi eld” [8]. The outsider/pro- well-documented, as Headland explained: fessional vis-à-vis insider/local formulation sug- gested by the emic-etic conceptualization found Most practicing anthropologists today use in- its way into medical anthropological thinking ‒ sights about the differing perceptions of real- tacitly, unacknowledged and not completely so at ity of different subcultural groups as a princi- this early stage ‒ by redefi ning “disease and ill- pal ‒ if not the principal ‒ conceptual tool of ness” as technical terms. “Disease” referred to the their trade. The emic/etic distinction, then, professional understanding of particular health underlies one of the basic contributions of problems, and “illness” referred to personal expe- modern anthropology to the working world rience of these problems, their meaning and im- (i.e., the ability to understand and interpret pact. The fi rst article in the fi rst issue of Culture, other cultures). Many anthropologists, in Medicine and Psychiatry in 1977 elaborated this fact, if not other social scientists, may owe “new conceptual model of disease and illness” [9]. their jobs to their ability to make the distinc- Although Pike’s and Eisenberg’s conceptual tion between emic and etic. [7 (Ch 1, reprint- models referred to similar dichotomous interests, ed online
grate “scientifi c” and “social” concepts of disease something big or complex visible or more readily and illness as a basis for a functional system of understandable. They provide accessible meta- medical research and care [9]. The two were com- phors for inaccessible objects or concepts. There plementary and attention to both was required. may be considerable overlap, of course, in the rep- The question of ensuring adequate attention resentational and predictive interests of any par- to the social and cultural features of illness in re- ticular scientifi c explanatory models. search and clinical practice motivated further de- Kleinman’s formulation of “explanatory velopment and rethinking of the conceptualiza- models of illness” is different from these scientifi c tion. Hahn reviewed the anthropological disease/ explanatory models. His book Patients and illness distinction and problems arising from in- Healers in the Context of Culture, which provided consistent usage during a period when the details an early comprehensive presentation of the ex- of their relationship were actively contested [10]. planatory model framework, is regarded as “one But current understanding of the distinction is of the most infl uential books in medical anthro- clearly consistent with the emic-etic framework pology and the social sciences of medicine” [16]. that has been adapted for practical interests of an In it, illness explanatory models are defi ned as agenda for medicine and health. This emic con- “notions about an episode of sickness and its treat- cept, illness distinguished from etic concepts of ment that are employed by all those engaged in disease, may also be understood as a point of ref- the clinical process” [1, p 105]. They differ from erence for Kleinman’s illness explanatory model, etic scientifi c explanatory models insofar as they though he did not present it in those terms. refer to the understanding of people affected by health problems rather than the understanding of Illness explanatory models health professionals called upon to treat or pre- The term “explanatory model” (without vent these problems by virtue of clinical training Kleinman’s explicit or implied reference to “ill- and professional expertise. Furthermore, his ex- ness”) is widely used in various disciplines and planatory models refer explicitly to illness epi- settings to describe or explain a set of relation- sodes rather than general illness beliefs, theories ships and to represent objects, ideas, experiences or a systematized ethnomedicine. They acknowl- and theories. In the fi eld of neuroscience and men- edge the “vagueness, multiplicity of meanings, tal health, professionals may refer to explanatory frequent changes, and lack of sharp boundaries models of psychiatry and psychiatric illness [11], between ideas and experiences are characteristic statistical models [12] and explanatory models of of lay EMs [i.e., explanatory models]” [1, p 107]. various mental health problems and clinical man- Kleinman argued for their relevance with agement issues [13-15]. The validity and value of reference to fi ve core clinical functions. They pro- such models is based on research fi ndings, profes- vided an approach to bringing anthropology into sional experience and academic study. In any the clinic, and a means of overcoming outdated fi eld, theoretical models tend to be valued most approaches to cross-cultural psychiatry that fo- for whether they accurately account for observ- cused on the exotic examples of far-away prob- able data and for their predictive capacity to relate lems [17]. Limiting the scope of the fi eld to “prim- explanatory variables and outcomes. Representa- itive concepts of disease,” as it was in the 1930s tional models are valued for how well they make [18], was no longer acceptable. He initially re- • 12 • Cultural Epidemiology
garded the illness explanatory model to train clini- chiatry” or “mental health” as an index subject cians in the principles and practice of mini-eth- heading. Abstracts of the 461 articles retrieved by nography, which they might then integrate in their the search were reviewed to weed out those deal- practice. He also considered the framework as a ing with etic scientifi c explanatory models, and tool for “ethnomedical” study, though he later be- 275 (60%) that referred to emic illness explana- came sceptical of that. In 1995 he wrote, tory models were identifi ed [2]. “Clinically, the explanatory model approach may In the 1980s, the fi rst decade of these ex- continue to be useful, but ethnography has fortu- planatory model studies concerned with mental nately moved well beyond this early formulation” health problems, 14 of 18 were published in CMP [19, p 9]. and 4 in Social Science and Medicine. The fi rst of Developing skills for clinically applied eth- these studies appeared in CMP authored by Dan nography (i.e., “mini-ethnography” in Kleinman’s Blumhagen. It distinguished “popular belief sys- terms) provided a way to engage with patients and tems” (ethnomedicine) from “expert belief sys- thereby ensure relevant consideration of culture tems” (scientifi c theory) in a study comparing ill- for effective medical care. He reaffi rmed the ap- ness explanatory models of 117 people with proach in 2006, suggesting it would help to over- “Hyper-Tension” with professional concepts of come inherent limitations of the increasingly pop- hypertension [21]. The study examined cognitive ular concept of “cultural competency” [20]. domains of hypertension, perceived causes, mech- Training for “competence,” he argued, must not anisms, outcome and the rôle of various stressors be reduced to a checklist approach that failed to and psychological symptoms. The analytic frame- address the fundamental questions of “What’s at work comparing popular and professional models stake?” for a patient and family. Clinical assess- was consistent with research interests in explana- ment of explanatory models in the framework of tory model studies of patients and healers. mini-ethnography should be a start, rather than an Other early explanatory model research in- end, of the clinical conversation. cluded a study by Mark Nichter showing the rele- vance of an idioms-of-distress approach to psy- Illness Explanatory Model chiatric evaluation. Attending to questions of Research on Psychiatry and gender and culture, he highlighted the prominence Mental Health Problems of somatization among fi ndings in expressing dis- tress among Havik Brahmin women in South Notwithstanding later reservations about the India, and the rôle of weak or inaccessible social explanatory model for ethnomedical study, the ap- support [22]. Atwood Gaines developed an ethno- proach became a popular and infl uential frame- graphic study of beliefs and practices of fi ve so- work for research in medical anthropology, and it called Christian psychiatrists. This contribution to remains so in the fi eld of cultural psychiatry. cultural psychiatry focused on psychiatric practice Many of the initial studies were published in and the explanatory models of practitioners, rather CMP. To clarify the scope and coverage of this than patients’ explanatory models or congruence literature, I undertook a PubMed literature search with those of their clinicians [23]. Cecil Helman, to identify indexed articles through 2015 with a general practitioner and medical anthropologist “explanatory model (or models)” and either “psy- in London, assessed the explanatory models of 42 Mitchell G. Weiss • 13 •
patients with respiratory and gastrointestinal ill- were attentive to such values of interdisciplinary nesses, and considered clinical implications. His engagement associated with efforts to advance analysis melded insights from his dual vantage cultural epidemiology, whether identifi ed with point as both a clinician and anthropologist [24]. that term explicitly, like ours, or implicitly in the Focusing on the dualistic implications of “psycho- explanatory model research of other researchers. somatic” disorders, he distinguished the priority It should also be acknowledged, however, of somatic symptoms for his patients and the ten- that our focus on illness explanatory models is not dency of clinicians to “psychologize” these the only relevant framework for explanatory mod- symptoms. el research and cultural epidemiology. Other cul- The literature over the years indicates in- tural aspects of being for which distributions and creasing research interest in explanatory models health impact are important include essential fea- of various mental health problems, including tures of the cultural identity of individuals and common mental disorders [25], somatization and groups; cultural characteristics of contexts, situa- neurasthenia [26], schizophrenia and other psy- tions and responses are also relevant consider- choses [27], suicide [28], substance abuse and ad- ations for health impact. Cultural factors that af- dictive disorders [29], epilepsy and seizure disor- fect epidemiological measurement and the validity ders [30], dementia [31], personality disorders of data, and cultural factors that affect the risk, [32] and the psychosocial impact of primary med- course and outcomes of health problems are all ical disorders [33]. Stigma has also been a long- relevant issues for cultural epidemiology. standing and cross-cutting interest of explanatory Historically, Rudolf Virchow’s anthropologi- model studies of mental health problems [34, 35], cal and epidemiological analysis of the social and and it has become an important interest of cultural political determinants of a typhoid epidemic, he epidemiology. In the context of global mental investigated in Upper Silesia in 1848, and Emile health, Patel and colleagues recently suggested Durkheim’s landmark analysis of social features, that study of explanatory models is relevant for determinants and types of suicide are considered practical consideration of “demand-side barriers” among early examples of cultural epidemiology. to mental health services in India and China [36]. They are discussed by Jim Trostle in his analysis of the origins of the links between epidemiology Conceptualizing Cultural and culture [37]. Another important landmark in Epidemiology the fi eld, is the experience and impact of the Pholela Health Centre in South Africa, established Efforts to enable collaboration across disci- in 1940, which used fi eld-based methods of ap- plines involved some measure of clarifi cation and/ plied medical anthropology and fi eld epidemiolo- or simplifi cation of premises and methods of an- gy for community-oriented primary health care thropology and epidemiology. Like all interdisci- [37]. plinary encounters, it has been a formidable chal- The Pholela experience was highlighted in a lenge to ensure respectful collaboration that is landmark collection of case studies, Health, neither simplistic nor rigidly parochial in its own Culture and Community, published by Benjamin self-regard. Many of the more successful studies Paul in 1955. Paul referred to an infl uential re- of illness explanatory models, indicated above, mark by a malariologist, Samuel Darling, who • 14 • Cultural Epidemiology
worked on the Panama Canal project: “If you wish and collaboration. In a chapter examining the to control mosquitoes ... you must learn to think question of whether conceptual underpinnings of like a mosquito.” Explaining its signifi cance, he anthropology and epidemiology are inevitably in wrote, “It applies not only to mosquito popula- confl ict or reconcilable, Hahn argues that “the un- tions one seeks to damage but also to human pop- derlying logics of anthropology and epidemiology ulations one hopes to benefi t” [38]. The case stud- have much in common and that practices devel- ies emphasized the value of attending to culture oped in each discipline are necessary comple- and community, not just professional expertise, to ments to practices in the other” [41, p 99]. achieve worthwhile aims of health and develop- ment. Like the clinical explanatory model ap- Cultural epidemiology of explanatory proach, the community case studies demonstrated models the relevance of experience and priorities of pa- Development of our approach to cultural epi- tients and families, rather than relying solely on demiology transpired during a seminal period of the expertise of doctors and other health innovation in the fi elds of cultural psychiatry, professionals. medical anthropology and psychiatric epidemiol- Despite the indicated benefi ts of linking ogy. A willingness to rethink assumptions and health science and social science, like other inter- question disciplinary boundaries was shaped by disciplinary endeavours, excitement and opti- Kleinman’s “new cross-cultural psychiatry” [42], mism have also been tempered by scepticism and the new biopsychosocial medical model of George pessimism at the interface among both anthropol- Engel [43], the distinction of disease and illness ogists and epidemiologists [19]. Examining the and the illness explanatory model frameworks de- limits of interdisciplinary encounters, Srivastava, scribed above. Developments in the then-emerg- an anthropologist, was wary of inherent vulnera- ing fi eld of psychiatric epidemiology, exemplifi ed bilities of cross-disciplinary anthropologies, like by the international pilot study of schizophrenia medical anthropology. To better understand not [44] and the epidemiological catchment area study just the opportunities but also the pitfalls, he in the U.S.A. [45], also motivated development of called for detailed accounts of the professional ex- EMIC interviews to complement etic diagnostic periences of medical anthropologists “working in with emic cultural considerations. Our research a hospital or medical college, and trying to justfy designs, assessment methods and analytic strate- ... [a] place amidst those who might consider such gies endeavored to consider experience and to ap- positions as sheer appendages, easily dispens- ply insights and lessons from both psychiatric epi- able” [39, p 548]. Di Giacomo reported on such an demiology and medical anthropology to meet experience in an academic institution working research needs, invigorating cultural psychiatry. with cancer epidemiologists in Spain. Based on The confl uence of these seminal develop- that, she argued against the possibility of a cul- ments in both fi elds in the early 1980s stimulated tural epidemiology in “the naturalist epistemology my work as a cultural psychiatrist and post-doc- of Western institutional medicine” [40]. toral health social scientist. The interdisciplinary Other anthropologists working in clinical program in medical anthropology and cultural settings, have been more positive and optimistic psychiatry in the Department of Social Medicine about prospects for interdisciplinary integration and Health Policy at Harvard Medical School en- Mitchell G. Weiss • 15 •
couraged the agenda, as did experience develop- quantitative components of coherent datasets in a ing a course on cultural dimensions of internation- unifi ed analysis; we endeavored to move beyond al health at the Harvard School of Public Health. mixed methods to integrated methods. Acquiring The setting was conducive for harnessing clinical experience in the research group through research scholarship, epidemiological principles and health partnerships, refi ning our research strategies, social science in a joint effort to develop the closely following technological developments in EMIC framework. Inasmuch as we were working the software and emerging prospects for enhanc- outside the comfort zones of clinical, epidemio- ing mutual tolerance of qualitative and quantita- logical and anthropological research, we needed a tive methodologies all helped confront the conceptual framework to integrate experience challenges. across disciplines. To proceed with early efforts to construct in- EMIC Interviews as struments that became the EMIC framework, we Instruments for Cultural needed to think through principles of assessment, Epidemiology that is, how to frame questions, code responses, preserve narratives and manage unifi ed quantita- Development of our explanatory model in- tive and qualitative datasets. We needed to devel- terviews began in the mid1980s, producing a set op an appropriately complex but suffi ciently man- of localized instruments with a common semi- ageable approach to defi ning variables for structured interview framework that were collec- representing relevant features of explanatory tively and individually known as EMIC models (e.g., categories of distress, perceived (Explanatory Model Interview Catalogue). The causes, treatment preferences and related inter- acronym highlights their focus on “emic” illness ests), recognizing that a single explanatory model explanatory models. The reference to “catalogue” variable would be simplistic. We carefully consid- highlights plurality, thereby distinguishing the ered the limits of what could be coded and quanti- priority of locally adapting EMIC interviews from fi ed, and what should remain as a narrative com- expectations of a single standardized instrument ponent in our datasets. We also needed strategies for use in all settings, as commonly expected for for quantitative analysis (e.g., the concept of instruments used in psychiatric epidemiological prominence), qualitative analysis (e.g., thematic research and psychometric instruments for clini- deductive and inductive coding) and integrating cal assessment. these quantitative and qualitative interests. Kleinman had earlier suggested a topical Strategies for using qualitative data manage- framework for explanatory model interviewing in ment software to support thematic analysis of nar- a footnote of Patients and Healers [1, P 106]. The ratives held promise, though their capacity by the framework was also presented in a slightly modi- 1990s was rudimentary by today’s standards. fi ed form about 25 years later for his recommend- Although quantitative and qualitative methodolo- ed approach to clinical anthropology [20] Being gists were familiar with many of the respective open-ended, the questions supported an agenda quantitative and qualitative management and ana- for clinical ethnographic interviews. lytic tasks, our goal was to go a step further and The earlier guide was used in many explana- formulate strategies to relate the qualitative and tory model studies either as published or modifi ed • 16 • Cultural Epidemiology
Table 1. Adaptable generic EMIC interview framework.
Sections organized by topic Framework for questions and dataset Name of illness • Narrative and categories • Provides name or descriptive account for subsequent questions Patterns of distress • Narrative and coding for spontaneously reported categories, acknowledged in re- sponse to category probed and category identifi ed as most troubling • Prominence coding for categories based on how reported Perceived causes • Narrative and coding for spontaneously reported categories, acknowledged in re- sponse to category probed and category identifi ed as most important • Prominence coding for categories based on how reported Help seeking • Narrative and reported categories of help at home, prior help seeking outside home and current preferences • Distinguish fi rst and most important sources of help and derive prominence for analysis Stigma • Narrative and assessment of Indicators of relevant features of stigma • Each item coded for prominence based on level of affi rmation of stigma, and combined as index Ad ditional topics based on • Questions based on agenda of topical interests study-specifi c aims and • Format guided by framework for thematically coded narrative and coding catego- interests ries of response, based on strategy indicated above with attribution. It also provided a reference point cally begins with an open-ended question fol- for the agenda covered in developing EMIC inter- lowed by category-specifi c probing queries, so views for cultural epidemiological study. Unlike that unprompted and prompted responses may be the intended qualitative ethnographic use of distinguished in analysis. A summary question Kleinman’s guide, the EMIC interviews were con- concludes (i.e., most troubling concern, most im- structed both to enable epidemiological coding portant perceived cause, etc.). Strategic weighting and to preserve narrative responses, either from based on whether and how a category is reported notes prepared by a researcher during the inter- enables analysis of its prominence. Thematic view and/or audio recordings for reference and/or analysis of the collection of narratives of narra- transcription. Although covering a roughly con- tives enables qualitative elaboration and links to sistent topical agenda of the framework, it was interviews of selective respondents with particular expected that the semi-structured interview categorical response codes. would be adapted for any particular study to The approach enables descriptive and com- meet the needs defi ned by setting and study aims, parative analytic study. It also provides a quantita- to thereby ensure appropriate scope, localized tive means of analysing how relevant features of formulation of questions and coding categories. explanatory models, strategically confi gured as The framework that emerged from efforts to con- variables, may explain outcomes of practical clin- struct these early EMIC interviews is outlined ical or public health signifi cance. Such outcomes (Table 1). may include timely help seeking, help-seeking Each topical section (e.g., respondent-priori- and treatment preferences, adherence to treat- tized features of distress, perceived causes) typi- ment, changes in symptomatology, self-perceived Mitchell G. Weiss • 17 •
stigma, indicators of recovery and so forth. coded text, it also enables access to coded audio Datasets from studies using these EMIC inter- for analysis even before audio recordings are views typically include variables based on fi elds transcribed. coded during the interview and available for quan- titative analysis as coded, or reconfi gured (e.g., EMIC-interview Studies for prominence) in the course of analysis. These EMIC interviews were initially devel- Integrating qualitative and quantitative oped through research partnerships in India and methods the Department of Social Medicine and Health The datasets also include narrative data, Policy (now Global Health and Social Medicine) which are typically coded thematically, fi rst de- at Harvard Medical School. The fi rst EMIC inter- ductively according to the topic of the interview view was prepared for research on leprosy and question and then based on a coding structure de- mental health at the KEM Hospital in Mumbai in termined by analytic priorities and content. Use of collaboration with DR Doongaji. The study exam- qualitative data analytic software enables refer- ined explanatory models of leprosy and associated encing narratives and/or thematically coded seg- mental health impact (depression, anxiety and so- ments to variables of the dataset as a way to select matoform disorders). It also examined the rela- records for strategically planned analysis. The in- tionship of particular perceived causes to follow- tegrated approach provides a means of elaborating up clinic attendance for treatment of leprosy [47]. quantitative fi ndings with narratives, and a way to Research at the National Institute of Mental examine quantitative distributions of variables Health and Neuro Science (NIMHANS) in that show how widespread notable views of quali- Bangalore with R. Raguram and S. Jadhav fol- tative accounts actually are in the dataset. lowed. We examined explanatory models as the The development of software for integrated basis for a clinical ethnography of psychiatric dis- qualitative and quantitative data management and orders, studies of stigma and cultural dimensions analysis (especially MAXQDA) has contributed of depression [34, 48]. The approach developed to research capacity for integrated methods. for study of stigma has been used for other mental Recent EMIC interviews are now designed for use health problems at NIMHANS, including schizo- with tablet computers, and their accuracy and ap- phrenia [49], mental health problems in other set- peal have been validated in a study comparing tings [35] and for stigmatized infectious diseases tablet and traditional paper-based versions of the [50-52]. same interview [46]. By 1996, 20 EMIC-interview studies had Audio recording the interviews on the tablet been undertaken on various topics in cultural psy- device with time stamps that are entered with each chiatry, mental health, epilepsy, suicide and delib- screen swipe indicates the point of the interview erate self-harm. A multi-country study of disor- on the recording in response to each question. ders of neurasthenia, chronic fatigue syndrome This approach enables programming question- and other conditions characterized by medically based fi rst-level thematic coding of the audio fi les. unexplained fatigue and weakness is particularly It not only facilitates quick access to topically fo- noteworthy with regard to Pacifi c Rim research in cused audio segments and their transcription into the United States (Los Angeles), Canada • 18 • Cultural Epidemiology
(Toronto), Taiwan (Kaohsiung), PR China encouraged by collaboration with the WHO (Changsha) and Hong Kong [53]. Keh-Ming Lin, Special Programme for Research and Training in who developed the consortium, compared fi nd- Tropical Diseases (TDR). Several TDR projects ings across sites [54]. The research of an Indian involved use of cultural epidemiological methods group, whose subsequent inclusion extended to document community impact and stigma of on- cross-cultural experience to South Asia, recom- chocercal skin diseases [52]; gender, stigma and mended including the collection of comparable timely help-seeking for tuberculosis [50, 58, 59]; conditions in different settings under the heading perceived causes of childhood fever affecting of neurasthenia spectrum disorders (NSDs). timely treatment to prevent malaria mortality in Paralikar and colleagues also compared explana- Ghana [60]; and stigma and services for Buruli tory models in four outpatient specialty clinics in ulcer in Ghana [61, 62]. Related studies in the Pune (psychiatry, medicine, Ayurveda and derma- Middle-East examined stigma and condom use for tology) [26, 55]. HIV/AIDS in Jordan [63]. In addition to research on various clinical In recent years, partnership with WHO’s disorders, suicide and deliberate self-harm (DSH) Initiative for Vaccine Research has supported cul- became another important mental health research tural epidemiology studies to improve vaccine interest. Adaptations of the EMIC interview fo- coverage by assessing and addressing social and cused on underlying problems and triggers of sui- cultural concepts of illness, vaccines and trust in cidal behavior. Cultural epidemiological studies health systems. With the introduction of improved of DSH have examined accounts of survivors of oral cholera vaccines (OCV) for endemic popula- DSH and family survivors of a relative’s suicide. tions, not just travellers, a study of uptake in a Research in urban Mumbai identifi ed gender-re- mass vaccination campaign using OCV in lated differences in respondents’ reasons for DSH Zanzibar showed how local experience of cholera (e.g., personality problems were more often re- and other factors affected vaccine priority [64]. ported by women; and alcohol, work and fi nancial Additional studies of anticipated OCV uptake us- problems were more often reported by men). ing comparable methods were undertaken in Research experience led to development of an ap- Western Kenya and DR Congo [65]. Cultural epi- proach to sociocultural autopsy [28, 56]. In the demiological studies of infl uenza vaccine cover- rural Sundarban region of West Bengal, such re- age in India in the pandemic of 2009 have shown search contributed to community suicide preven- that ideas about community hesitancy invoked to tion campaigns [57]. explain limited coverage in Western Europe and North America may be less infl uential in India Other global health interests than access, health system priorities for vaccina- My move to the Swiss Tropical Institute in tion and clinician prescribing practices [66]. 1995 (renamed Swiss Tropical and Public Health Institute, Swiss TPH, in 2010) brought additional Alternative Strategies for opportunities and encouragement to expand re- Explanatory Model Research search activities in areas of infectious tropical dis- ease control. Cross-cutting interests in gender, Our EMIC interview framework for cultural stigma and community behaviour were further epidemiology has not been the only approach for Mitchell G. Weiss • 19 •
explanatory model research, and other instru- Other exclusively quantitative instruments ments and methods have also been developed for for explanatory model research include the Mental use beyond a single study. Some attended to com- Distress Explanatory Model Questionnaire plementary quantitative and qualitative interests, (MDEMQ) developed by Eisenbruch [70] and the like the EMIC, and others focused either on quan- Explanatory Model Association Task (EMAT) de- titative or qualitative methods rather than attempt- veloped by Ghane and colleagues, which was in- ing to integrate both. The Short Explanatory tended to serve as an indirect assessment [71]. Model Interview (SEMI) integrates qualitative as- Neither has been widely used, however, beyond sessment with quantitative analysis, and various initial validation studies. adaptations have been used in many studies [67]. A purely qualitative instrument for assessing In an effort to simplify coding at the time of the illness narratives has been used in a number of interview, the SEMI comprised a series of open- other studies. The McGill Illness Narrative ended questions covering an agenda similar to Interview (MINI) includes consideration of ex- EMIC interviews. Categories of perceived causes, planatory models within the scope of its 46-item priority symptoms and help-seeking experience theory-driven agenda, and it relies on “more in- are coded after the interview based on notes and tensive methods of narrative and discourse analy- transcripts. The tables reporting results from sis” [72]. Although topical interests of the MINI SEMI interviews typically refer to frequency of are well within the scope of the illness explanato- reporting without consideration of prominence ry model framework, the ideology and method- [68]. Extensive use of the SEMI has made impor- ological orientation of the MINI appear to recom- tant contributions to explanatory model research mend an alternative to epidemiological study, consistent with the framework of cultural epide- cultural or otherwise. miology, although these studies do not explicitly refer to the term. Related Clinical Tools An alternative approach that aims to simplify explanatory model research methods involves use Our approach to cultural epidemiology and of solely quantitative instruments and study de- EMIC interview designs anticipated practical val- signs. The BARTS Explanatory Model Inventory ue from translation of the research experience into – Checklist (BEMI-C) is an instrument derived enhanced capacity for culturally sensitive clinical from experience with a more comprehensive ex- practice. Such expectations were already clear to planatory model interview. Bhui and colleagues clinician-researchers working with the methods explain that clinical aims for use of the instrument [73]. For broader impact in translating benefi ts of and implicit concern about the labour of a mini- research into practice, however, more systematic ethnography justify the format of the BEMI-C, efforts were required to engage practitioners who because “clinicians do not usually have the time were not already convinced of its value. and resources to undertake a detailed and unstruc- Development of the Outline for Cultural tured exploration of EMs [explanatory models] or Formulation (OCF) for DSM-IV and the Cultural qualitative data analysis” [69, p 965]. This point, Formulation Interview (CFI) for DSM-5 has however, may blur research and clinical objec- helped to bring cultural considerations farther tives of the instrument. along towards mainstream clinical practice. • 20 • Cultural Epidemiology
EMIC interview instruments are carefully structured to facilitate assessment, coding and analysable data from many interviews. The prior- ity of the CFI, however, is primarily concerned with clinical aims of assessing a patient, fostering a treatment alliance and proceeding with a treat- ment plan for that patient. The layout of the CFI is therefore structured to facilitate productive inter- action with a patient but without clear attention to the acquisition, maintenance and analysis of a da- taset for research. The priority for a clinical inter- view is based more on what happens in the inter- view, and the value of a research interview is based more on the quality, validity and usefulness of the dataset. Figure 1. Translating research instruments into clini- An effective clinical interview that benefi ts a cal tools. PSE, Present State Examination; DIS, Diag- patient may nevertheless be a poor research inter- nostic Interview Schedule; CIDI, Composite Interna- view if it fails to provide relevant documentation. tional Diagnostic Interview; SCAN, Schedules for Similarly, a useful research interaction may lack Clinical Assessment in Neuropsychiatry; DSM, Diag- nostic and Statistical Manual for Mental Disorders; clinical benefi ts for a patient‒and such consider- ICD, International Classifi cation for Diseases. ations are notable on the agenda for ethical review of proposed research. Although careful attention The OCF was based on a framework that to the formulation of questions for clinical and re- considered the cultural identity of a patient, an search assessments are common interests, the re- emic account (illness explanatory model) of the spective rationale for each refl ects different priori- patient’s illness, the role of stressors and supports ties of purpose and intended use. in the social and cultural environment and the in- tercultural relationship between the clinician and Prospects and Promise the patient [74]. As an outline relegated to the back of the DSM-IV manual in the penultimate ap- Critical questions confronting mental health pendix, it had limited impact on mainstream train- and global health benefi t from consideration of ing and practice [75]. The DSM-5 intended not cultural epidemiology. Research strategies to ad- only to update the OCF but also to develop the dress these priorities, however, require an inter- CFI to make the OCF more accessible and appeal- disciplinary set of skills that extend traditional ing to clinicians; it was important to present the health science and social science curricula. core instrument in the body of the DSM and to Recognizing a need for training and capacity validate that instrument for feasibility, acceptabil- building, our research group developed a curricu- ity and clinical utility in a set of international fi eld lum for cultural epidemiology, and a course was trials [76, 77]. Supplementary modules further en- fi rst offered at the Swiss TPH in 2005. Workshops hanced the CFI. have also been conducted at partner institutions in Mitchell G. Weiss • 21 •
India, Australia, Kenya and South Africa. 4. Weiss MG: Cultural epidemiology: an introduction Sustaining these activities, however, is challeng- and overview. Anthropol Med 2001; 8: 5-29. ing, and it is hoped that recent innovations and 5. Weiss MG: Explanatory Model Interview Catalogue improved technologies may motivate commit- (EMIC): framework for comparative study of illness. ment to realize potential. Transcult Psychiatry 1997; 34: 235-63. 6. Pike KL: Etic and emic standpoints for the descrip- The logic and benefi ts of synergies from tion of behaviour (Ch 2). In: Pike KL (eds): Language linking anthropology and epidemiology as com- in Relation to a Unifi ed Theory of the Structure of plementary basic sciences for mental health and Human Behavior, 2nd Ed. The Hague, Netherlands: global health are compelling and ultimately per- N V Uitgeverij Mouton & Co., 1967: 37-72. haps inexorable. Although the approach outlined 7. Headland TN, Pike KL, Harris M: Emics and Etics: here is one among others for bringing benefi ts of The Insider/Outsider Debate. Newbury Park, medical anthropology to clinical practice and pub- California: Sage, 1990. lic health, as acknowledged in the course of this 8. Kleinman A: Culture, and illness: a question of mod- overview, it is hoped that refl ection on our experi- els. Cult Med Psychiatry 1977; 1: 229-31. ence will contribute to strategies, further capacity 9. Eisenberg L: Disease and illness: distinctions be- and achievements of this important interdisciplin- tween professional and popular ideas of sickness. ary agenda. Cult Med Psychiatry 1977; 1: 9-23. 10. Hahn RA: Rethinking “illness” and “disease”. Acknowledgements Contributions to Asian Studies 1984; 18: 1-23. 11. Kendler KS: Explanatory models for psychiatric ill- ness. Am J Psychiatry 2008; 165: 695-702. This overview is based on an invited plenary 12. Katz MH: Multivariable analysis: a primer for read- lecture at the joint meeting of the annual meeting ers of medical research. Ann Intern Med 2003; 138: of Taiwanese Society of Psychiatry and biennial 644-50. meeting of Pacifi c Rim Conference of Psychiatrists 13. Connan F, Campbell IC, Katzman M, Lightman SL, in Kaohsiung, Taiwan, on November 4, 2016. The Treasure J: A neurodevelopmental model for anorex- author thanks for travel support. He declares de- ia nervosa. Physiol Behav 2003; 79: 13-24. clares no potential confl icts of interest in writing 14. Kumar S, Ng B: Crowding and violence on psychiat- this overview. ric wards: explanatory models. Can J Psychiatry 2001; 46: 433-7. References 15. Bazzazian S, Besharat MA: An explanatory model of adjustment to type I diabetes based on attachment, 1. Kleinman A: Patients and Healers in the Context of coping, and self-regulation theories. Psychol Health Culture: An Exploration of the Borderland between Med 2012; 17: 47-58. Anthropology, Medicine, and Psychiatry. Berkeley: 16. Gaines AD: Culture, Medicine, Psychiatry and wis- University of California Press, 1980. dom: honoring Arthur Kleinman. Cult Med 2. Weiss MG: Explanatory models in psychiatry. In: Psychiatry 2016; 40: 538-69. Bhugra D, Bhui K (eds): Textbook of Cultural 17. Kirmayer LJ, Jarvis E: Cultural psychiatry: from mu- Psychiatry, 2nd Edition. Cambridge, UK: Cambridge seums of exotica to the global agora. Curr Opin University Press, 2017. Psychiatry 1998; 11: 183-9. 3. Sartorius N: Culture and the epidemiology of depres- 18. Clements FE: Primitive Concepts of Disease. sion. Psychiatr Neurol Neurochir 1973; 76: 479-87. Berkeley, California, USA: University of California • 22 • Cultural Epidemiology
Press, 1932. care in Cape Town. Transcult Psychiatry 2015; 52: 19. Kleinman A: Writing at the Margin Discourse: 659-80. Between Anthropology and Medicine. Berkeley, 31. Giebel CM, Jolley D, Zubair M, et al.: Adaptation of California, USA: University of California Press, the Barts Explanatory Model Inventory to dementia 1995. understanding in South Asian ethnic minorities. 20. Kleinman A, Benson P: Anthropology in the clinic: Aging Ment Health 2016; 20: 594-602. the problem of cultural competency and how to fi x it. 32. Alarcon RD, Leetz KL: Cultural intersections in the PLoS Med 2006; 3: e294. psychotherapy of borderline personality disorder. Am 21. Blumhagen D: Hyper-tension: a folk illness with a J Psychother 1998; 52: 176-90. medical name. Cult Med Psychiatry 1980; 4: 33. Chipimo PJ, Tuba M, Fylkesnes K: Conceptual mod- 197-224. els for mental distress among HIV-infected and unin- 22. Nichter M: Negotiation of the illness experience: fected individuals: a contribution to clinical practice Ayurvedic therapy and the psychosocial dimension and research in primary-health-care centers in of illness. Cult Med Psychiatry 1981; 5: 5-24. Zambia. BMC Health Serv Res 2011; 11: 7. 23. Gaines AD: The twice-born: ‘Christian psychiatry’ 34. Raguram R, Weiss MG, Channabasavanna SM, and Christian psychiatrists. Cult Med Psychiatry Devins GM: Stigma, depression, and somatization in 1982; 6: 305-24. South India. Am J Psychiatry 1996; 153: 1043-9. 24. Helman CG: Psyche, soma, and society: the social 35. Weiss MG, Jadhav S, Raguram R, Vounatsou P, construction of psychosomatic disorders. Cult Med Littlewood R: Psychiatric stigma across cultures: lo- Psychiatry 1985; 9: 1-26. cal validation in Bangalore and London. Anthropol 25. Jacob KS, Bhugra D, Lloyd KR, Mann AH: Common Med 2001; 8: 71-87. mental disorders, explanatory models and consulta- 36. Patel V, Xiao S, Chen H, et al.: The magnitude of and tion behaviour among Indian women living in the health system responses to the mental health treat- UK. J R Soc Med 1998; 91: 66-71. ment gap in adults in India and China. Lancet 2016. 26. Paralikar V, Agashe M, Sarmukaddam S, Deshpande 37. Trostle JA: Epidemiology and Culture. Cambridge, S, Goyal V, Weiss MG: Cultural epidemiology of UK: Cambrige University Press, 2005 neurasthenia spectrum disorders in four general hos- 38. Paul BD: Health, Culture, and Community: Case pital outpatient clinics of urban Pune, India. Transcult Studies of Public Reactions to Health Programs. Psychiatry 2011; 48: 257-83. New York: Harper K Brothers, 1955. 27. Bhikha AG, Farooq S, Chaudhry N, Husain N: A sys- 39. Srivastava VK: The future of anthropology. tematic review of explanatory models of illness for Economic and Political Weekly 1999; 34: 545-52. psychosis in developing countries. Int Rev Psychiatry 40. DiGiacomo SM: Can there be a “cultural epidemiol- 2012; 24: 450-62. ogy”? Med Anthropol Q 1999; 13: 436-57. 28. Parkar SR, Nagarsekar B, Weiss MG: Explaining sui- 41. Hahn RA: Sickness and Healing: an Anthropological cide in an urban slum of Mumbai, India: a sociocul- Perspective. New Haven: Yale University Press, tural autopsy. Crisis 2009; 30: 192-201. 1995. 29. Nadkarni A, Dabholkar H, McCambridge J, et al.: 42. Kleinman AM: Depression, somatization and the The explanatory models and coping strategies for al- “new cross-cultural psychiatry”. Soc Sci Med 1977; cohol use disorders: an exploratory qualitative study 11: 3-10. from India. Asian J Psychiatr 2013; 6: 521-7. 43. Engel GL: The need for a new medical model: a chal- 30. Keikelame MJ, Swartz L: ‘A thing full of stories’: lenge for biomedicine. Science 1977; 196: 129-36. traditional healers’ explanations of epilepsy and per- 44. Sartorius N, Shapiro R, Jablensky A: The spectives on collaboration with biomedical health International Pilot Study of Schizophrenia. Schizophr Mitchell G. Weiss • 23 •
Bull 1974; 1: 21-34. India. World Cultural Psychiatry Research Review 45. Regier DA, Myers JK, Kramer M: The NIMH epide- (Durham) 2015; 40-50. miologic catchment area program: historical context, 56. Parkar SR, Nagarsekar BB, Weiss MG: Explaining major objectives, and study population characteris- suicide: identifying common themes and diverse per- tics. Arch Gen Psychiatry 1984; 41: 934-41. spectives in an urban Mumbai slum. Soc Sci Med 46. Giduthuri JG, Maire N, Joseph S, et al.: Developing 2012; 75: 2037-46. and validating a tablet version of an illness explana- 57. Chowdhury AN, Banerjee S, Brahma A, Hazra A, tory model interview for a public health survey in Weiss MG: Sociocultural context of suicidal behav- Pune, India. PLoS One 2014; 9: e107374. iour in the sundarban region of India. Psychiatry J 47. Weiss MG, Doongaji DR, Siddhartha S, et al.: The 2013; 2013: 486081. Explanatory Model Interview Catalogue (EMIC). 58. Coreil J, Mayard G, Simpson KM, Lauzardo M, Zhu contribution to cross-cultural research methods from Y, Weiss M: Structural forces and the production of a study of leprosy and mental health. Br J Psychiatry TB-related stigma among Haitians in two contexts. 1992; 160: 819-30. Soc Sci Med 2010; 71: 1409-17. 48. Weiss MG, Raguram R, Channabasavanna SM: 59. Gosoniu GD, Ganapathy S, Kemp J, et al.: Gender Cultural dimensions of psychiatric diagnosis: a com- and socio-cultural determinants of delay to diagnosis parison of DSM-III-R and illness explanatory models of TB in Bangladesh, India and Malawi. Int J Tuberc in south India. Br J Psychiatry 1995; 166: 353-9. Lung Dis 2008; 12: 84-55. 49. Raguram R, Raghu TM, Vounatsou P, Weiss MG: 60. Ahorlu CK, Koram KA, Ahorlu C, de SD, Weiss Schizophrenia and the cultural epidemiology of stig- MG: Socio-cultural determinants of treatment delay ma in Bangalore, India. J Nerv Ment Dis 2004; 192: for childhood malaria in southern Ghana. Trop Med 734-44. Int Health 2006; 11: 1022-31. 50. Somma D, Thomas BE, Karim F, et al.: Gender and 61. Ackumey MM, Gyapong M, Pappoe M, Kwakye- socio-cultural determinants of TB-related stigma in Maclean C, Weiss MG: Illness meanings and experi- Bangladesh, India, Malawi and Colombia. Int J ences for pre-ulcer and ulcer conditions of Buruli Tuberc Lung Dis 2008; 12: 856-66. ulcer in the Ga-West and Ga-South Municipalities of 51. Peters RM, Dadun, van Brakel WH, et al.: The cul- Ghana. BMC Public Health 2012; 12: 264. tural validation of two scales to assess social stigma 62. Ackumey MM, Gyapong M, Pappoe M, Maclean in leprosy. PLoS Negl Trop Dis 2014; 8: e3274. CK, Weiss MG: Socio-cultural determinants of time- 52. Vlassoff C, Weiss M, Ovuga EB, et al.: Gender and ly and delayed treatment of Buruli ulcer: implica- the stigma of onchocercal skin disease in Africa. Soc tions for disease control. Infect Dis Poverty 2012; 1: Sci Med 2000; 50: 1353-68. 6. 53. Lee S, Yu H, Wing Y, et al.: Psychiatric morbidity 63. Alkaiyat A, Schaetti C, Liswi M, Weiss MG: Condom and illness experience of primary care patients with use and HIV testing among men who have sex with chronic fatigue in Hong Kong. Am J Psychiatry men in Jordan. J Int AIDS Soc 2014; 17: 18573. 2000; 157: 380-4. 64. Schaetti C, Ali SM, Chaignat CL, Khatib AM, 54. Lin K, Lin M, Zheng Y: Neurasthenia and chronic Hutubessy R, Weiss MG: Improving community fatigue syndrome: lessons from cross-cultural study. coverage of oral cholera mass vaccination cam- In: Yilmaz A, Weiss M, Riecher-Rössler A (eds): paigns: lessons learned in Zanzibar. PLoS One 2012; Cultural Psychiatry: Euro-International Perspectives 7: e41527. (Basel) 2001: 68-80. 65. Schaetti C, Sundaram N, Merten S, et al.: Comparing 55. Paralikar V, Agashe M, Weiss MG: Clinical fatigue sociocultural features of cholera in three endemic and weakness as neurasthenia spectrum disorders in African settings. BMC Med 2013; 11: 206. • 24 • Cultural Epidemiology
66. Sundaram N, Purohit V, Schaetti C, Kudale A, Joseph schedule to elicit meanings and modes of reasoning S, Weiss MG: Community awareness, use and prefer- related to illness experience. Transcult Psychiatry ence for pandemic infl uenza vaccines in Pune, India. 2006; 43: 671-91. Hum Vaccin Immunother 2015; 11: 2376-88. 73. Paralikar V, Agashe M, Weiss MG: Essentials of 67. Lloyd KR, Jacob KS, Patel V, St LL, Bhugra D, good practice: the making of a cultural psychiatrist in Mann AH: The development of the Short Explanatory urban India (Ch 5). In: Ancis JR (eds): Culturally Model Interview (SEMI) and its use among primary- Responsive Intervientions: Innovative Approaches to care attenders with common mental disorders. Working with Diverse Populations. New York: Psychol Med 1998; 28: 1231-7. Brunner-Routledge, 2004: 103-24. 68. Savarimuthu RJ, Ezhilarasu P, Charles H, Antonisamy 74. Lewis-Fernandez R, Diaz N: The cultural formula- B, Kurian S, Jacob KS: Post-partum depression in the tion: a method for assessing cultural factors affecting community: a qualitative study from rural South the clinical encounter. Psychiatr Q 2002; 73: India. Int J Soc Psychiatry 2010; 56: 94-102. 271-95. 69. Bhui K, Rudell K, Priebe S: Assessing explanatory 75. Mezzich JE, Kirmayer LJ, Kleinman A, et al.: The models for common mental disorders. J Clin place of culture in DSM-IV. J Nerv Ment Dis 1999; Psychiatry 2006; 67: 964-71. 187: 457-64. 70. Eisenbruch M: Classifi cation of natural and super- 76. Paralikar VP, Sarmukaddam SB, Patil KV, Nulkar natural causes of mental distress. development of a AD, Weiss MG: Clinical value of the cultural formu- Mental Distress Explanatory Model Questionnaire. J lation interview in Pune, India. Indian J Psychiatry Nerv Ment Dis 1990; 178: 712-9. 2015; 57: 59-67. 71. Ghane S, Kolk AM, Emmelkamp PM: Direct and in- 77. Lewis-Fernandez R, Aggarwal NK, Lam PC, et al.: direct assessment of explanatory models of illness. Feasibility, acceptability and clinical utility of the Transcult Psychiatry 2012; 49: 3-25. Cultural Formulation Interview: mixed-methods re- 72. Groleau D, Young A, Kirmayer LJ: The McGill sults from the DSM-5 international fi eld trial. Br J Illness Narrative Interview (MINI): an interview Psychiatry, in press.