Ethnography and health careBMJ 2000; 321 doi: http://dx.doi.org/10.1136/bmj.321.7273.1400 (Published 02 December 2000) Cite this as: BMJ 2000;321:1400
- Jan Savage (), senior research fellow
- Correspondence to:
- Accepted 20 September 2000
Development of a culture of evidence based medicine depends on a body of research that draws from both qualitative and quantitative approaches.1 Recent BMJ articles have usefully questioned a stark polarity between qualitative and quantitative research and helped to demystify qualitative approaches. 2 3 4 There has been little mention of ethnography, however, and little argument for its use in health research.
I have examined some of these omissions, giving a broad indication of the nature of ethnography and arguing for its greater use within health care. I have given examples of ethnographic studies to suggest some of the issues that ethnography can help to explore, together with a brief outline of limitations of the approach.
Ethnography has been overlooked as a qualitative methodology for the in depth study of healthcare issues in the context in which they occur
An ethnographic study can utilise a range of qualitative and quantitative methods
The methods of ethnographic research raise ethical and other issues, which means that skilled supervision is essential
Ethnographers do not usually aim to produce findings that can be generalised
Ethnography can be useful in a predesign stage of research and can generate questions for research that can be followed up by other methodologies
What is ethnography?
Perhaps one of the reasons for the neglect of ethnography is that there is no standard interpretation of what it is. Ethnography is, confusingly, both a process and a product: the term can apply both to a methodology and to the written account of a particular ethnographic project. It is not, as is often implied, a pseudonym for qualitative research in general or a way of describing studies premised solely on semistructured interviews. On the contrary, an ethnographic approach usually incorporates a range of methods and can combine qualitative and quantitative data. For many, the defining feature of ethnography is the use of participant observation,5 entailing prolonged fieldwork. Box 1 provides an example of an ethnographic study that uses mixed methods, including participant observation, to explore complex clinical and organisational issues.
Box 1 : Ethnography as a mixed method approach
A team of ethnographers were invited to look at clinical decision making by staff at a mental health centre who were concerned by the possible impact of managed care on professional status and provision of service.6 Participant observation allowed analysis of activities such as staff meetings and the tracing of a client's path through the clinic's administrative process. Interviews and informal discussions with clinicians provided data on professional backgrounds, therapeutic orientation, and clinical activities. Findings suggested that clinicians were not becoming de-professionalised, so much as re-professionalised. In shifting from their stance as critics to promoters of managed care they were apparently losing sight of a moral vision of good treatment for mental health.RETURN TO TEXT
Ethnography has its earliest roots in social anthropology, which traditionally focused on small scale communities that were thought to share culturally specific beliefs and practices. The motives for much early ethnographic work and the neutrality of the white ethnographer in an era of Western imperialism are now viewed with some scepticism.7 Political change, both globally and within the academic world, has meant that the ethnographer's authority to provide the only, or most legitimate, account is no longer accepted.8
Although the issue of authority is not simply defused by a change in location, the focus for many Western ethnographers has shifted from remote communities to settings “at home,” such as corporate organisations. At the same time, phenomena such as new information technologies, new national and local identities, and the development of theoretical perspectives that reject assumptions about social coherence have challenged the traditional view that “culture” is a matter of shared beliefs and practices. Instead, recognition is given to the differences existing within social groups, with some social scientists arguing that “culture” marks a process of struggle to determine meaning on the part of individuals with unequal access to power.9 For example, an ethnography of a surgical firm focusing on infection control practices, if shaped by an “old” view of culture, might identify collective understandings of the team's practices, such as its agreed methods and rationale for creating a “sterile” field. In contrast, a “new” understanding of culture would suggest greater emphasis on the activities and explanations of different team members and in identifying who had the power to impose their particular practices on other staff.
Most ethnographers today would agree that the term ethnography can be applied to any small scale social research that is carried out in everyday settings; uses several methods; evolves in design throughout the study; and focuses on the meanings of individuals' actions and explanations, rather than their quantification.10 In addition, ethnography is viewed as contextual and reflexive: it emphasises the importance of context in understanding events and meanings and takes into account the effects of the researcher and the research strategy on findings.11 There is also wide agreement that ethnography combines the perspectives of both the researcher and the researched.11
The way in which ethnography is used, however, depends on several factors, including the philosophical stance of the researcher or the practicalities of research funding.12 There is, for example, no overall consensus among ethnographers about the epistemology, or theory of knowledge, that underpins an ethnographic account. Instead, different kinds of ethnographies rest on different ideas of what constitutes legitimate knowledge.12 Some ethnographers, for example, use an interpretive approach, drawing on experiential knowledge gained from physical participation in the field,13 knowledge that others might discount as unverifiable.
It might be argued that such an approach represents a narcissistic shift of focus from the experience of the participants in the research to that of the ethnographer,8 yet it offers one response to the crisis of representation in the social sciences. This crisis has arisen partly because of uncertainty about how to describe social reality and partly because of the challenge to traditional assumptions, referred to earlier, about whose voice has authority. Additionally, there is growing acknowledgement that the knowledge generated by an ethnographic approach is strongly shaped by the nature of the relationship between the researcher and the researched.14 This has prompted the development of new forms of ethnography, such as critical ethnography, which attempt to restructure the research process in ways that promote the views of those who are often silent or marginalised.15
Awareness of the diverse positions within ethnographic research is important for at least two reasons. Firstly, many researchers agree that the epistemological foundations of an ethnography should continue to exert a strong influence throughout the entire research process.11 Take the example of an ethnography concerned with the implications of physical intimacy in clinical encounters. This study was based on an epistemology that extended legitimacy to knowledge from all the senses, not only sight, which suggested the researcher's participation in, rather than mere observation of, clinical work, to collect experiential data.13 Ethnography is thus not a simple matter of the ad hoc mixing of several methods.
Secondly, these diverse epistemological stances raise questions about the evaluation of ethnographic research and the appropriateness of criteria such as relevance and validity. These questions are particularly important for the broader acceptance and funding of this methodology in healthcare research, but provision of set criteria for the assessment of ethnographic research is notoriously difficult. While Hammersley makes some helpful proposals in this respect (box 2), it is doubtful that every ethnographer would accept all his suggestions or give them equal emphasis.10 Perhaps the best way of examining this complex issue here is by reference to more detailed discussions of the evaluation of qualitative research. 4 16
Box 2 : Possible criteria for assessing ethnographic research10
The consistency of claims compared with empirical data
The credibility of the account to readers and those studied
The extent to which findings have relevance to those in similar settings
The extent to which the influence of the research design and strategy on findings is considered (the reflexivity of the account), and the existence of an audit trail
The various perspectives encapsulated by the term ethnography can be bewildering, but the versatility of this approach is also one of its strengths, not least in the study of healthcare issues.
Ethnography and health care
Ethnography can be applied to healthcare issues in numerous ways. It has been seen as a way of accessing beliefs and practices, allowing these to be viewed in the context in which they occur and thereby aiding understanding of behaviour surrounding health and illness. 11 17 It is therefore particularly valuable as patients' views on the experience of illness or delivery of service are becoming recognised as central to a modernised NHS. Ethnography can show, for example, how the effectiveness of therapeutic interventions can be influenced by patients' cultural practices18 and how ethnocentric assumptions on the part of professionals can impede effective health promotion (box 3).19
Box 3 : Ethnography and the delivery of health care
A study of clinics serving low income, predominantly African-American women in mid-west America found that experiential knowledge held by clinic attenders was overlooked by clinic professionals, who were primarily of European-American descent.19 This was particularly important for those women with lactose intolerance, which in the United States is far more common among African-Americans. Data on interactions in the clinic showed that there were often barriers that rendered differences between staff and clinic attenders invisible or invalid and prevented the consideration of alternative dietary approaches.RETURN TO TEXT
In addition, ethnography is particularly useful in understanding the organisation of health care.17 For example, communication and information management within the NHS have been described as chaotic.20 Understanding why this is the case and how it can be improved is seen to demand methods that go beyond questionnaires and surveys. Through the nature and range of methods it can adopt, ethnography can provide a nuanced understanding of an organisation and allow comparison between what people say and what they do. It can, for instance, help to identify the ways that an organisation's formal structure (its rules and decision making hierarchies) are influenced by an informal system created by individuals or groups within the organisation or indicate how professional knowledge is locally produced in particular settings (box 4).21
Box 4 : Ethnography in the study of professional groups
Atkinson used an ethnographic approach to study the clinical reasoning in a group of haematologists through observing activities such as grand rounds and clinical lectures.21 He showed how the expert knowledge of these physicians emerged as a local and joint production through clinical talk that was simultaneously characterised by confidence, dogmatism, and uncertainty. From this, Atkinson raises important issues about the use of algorithms and decision making models within medicine and whether these acknowledge the complexities of practical work and clinical reasoning.RETURN TO TEXT
Like all approaches to research, however, ethnography has its limitations. These are amply spelt out elsewhere,10 but some examples that are particularly pertinent to healthcare research are worth raising here.
Funding bodies for research in health services are often not receptive to ethnography on the basis that, as a qualitative methodology, it does not lead to generalisable findings. Some researchers dispute this argument, claiming that qualitative research requires its own criteria for generalisability.22 Others, however, do not consider generalisation to be the purpose of qualitative research and point instead to the in depth understanding that ethnography can achieve and the way it can identify groundbreaking questions or hypotheses that can be further explored through other methodologies.16
Other problems are those associated with observation of participants. This method provides rich data but takes considerable time and sustained supervision to re-cast what might be familiar and apparently irrelevant as strange and interesting. The labour intensive nature of fieldwork also means that it is relatively costly. Some healthcare researchers deal with these problems by carrying out focused ethnographies in which fieldwork is shortened by entering the field with established research questions and less emphasis on participant observation. Finally, participant observation raises challenging ethical questions and practicalities with regard to informed consent that may be heightened by the lack of power seen in certain groups such as patients or junior staff. Informed consent therefore needs to be carefully considered, negotiated, and regularly reconfirmed with study participants.23
Ethnography is a complex and contested activity drawing on a range of epistemological positions and methods and often demanding different modes of evaluation from other methods more commonly used in healthcare research. As a detailed way of witnessing human events in the context in which they occur, ethnography can help healthcare professionals to solve problems beyond the reach of many research approaches, particularly in the understanding of patients' and clinicians' worlds.
Contributors: The need for a paper on ethnography and health care was initially identified by members on the Ethnography and Health Care Group (email@example.com). Helen Allan, Alison Crombie, Kathryn Ehrich, Daniel Kelly, and Susie Pearce contributed to the conceptualisation and early revisions of the paper.
Competing interests The academic department in which I am based enters the Research Assessment Exercise. Publication of this paper may therefore benefit my employing organisation, although the benefits would be marginal and indirect.