Intended for healthcare professionals

Education And Debate

Qualitative research and evidence based medicine

BMJ 1998; 316 doi: https://doi.org/10.1136/bmj.316.7139.1230 (Published 18 April 1998) Cite this as: BMJ 1998;316:1230
  1. Judith Green (j.green{at}lshtm.ac.uk), lecturer in sociologya,
  2. Nicky Britten, senior lecturer in medical sociologyb
  1. a Health Services Research Unit, London School of Hygiene and Tropical Medicine, London WC1E 7HT
  2. b Department of General Practice, United Medical and Dental School of Guy's and St Thomas's Hospitals, London SE11 6SP
  1. Correspondence to: Dr Green
  • Accepted 21 November 1997

Qualitative research may seem unscientific and anecdotal to many medical scientists. However, as the critics of evidence based medicine are quick to point out, medicine itself is more than the application of scientific rules.1 Clinical experience, based on personal observation, reflection, and judgment, is also needed to translate scientific results into treatment of individual patients.2 Personal experience is often characterised as being anecdotal, ungeneralisable, and a poor basis for making scientific decisions. However, it is often a more powerful persuader than scientific publication in changing clinical practice,3-5 as illustrated by the occasional series “A patient who changed my practice” in the BMJ.6

In an attempt to widen the scope of evidence based medicine, recent workshops have included units on other subjects, including economic analysis and qualitative research.7 However, to do so is to move beyond the discipline of clinical epidemiology that underpins evidence based medicine. Qualitative research, in particular, addresses research questions that are different from those considered by clinical epidemiology. Qualitative research can investigate practitioners' and patients' attitudes, beliefs, and preferences, and the whole question of how evidence is turned into practice. The value of qualitative methods lies in their ability to pursue systematically the kinds of research questions that are not easily answerable by experimental methods.

We use the example of asthma treatment to illustrate how qualitative methods can broaden the scope of evidence based medicine. Although there is consensus over evidence based practice in the treatment of asthma,8 questions remain about general practitioners' use of clinical guidelines and patients' use of prescribed medication.9

Summary points

Qualitative methods can help bridge the gap between scientific evidence and clinical practice

Qualitative research findings provide rigorous accounts of treatment regimens in everyday contexts

This can help us understand the barriers to using evidence based medicine, and its limitations in informing decisions about treatment

Recognising the limits of evidence based medicine does not imply a rejection of research evidence but awareness that different research questions require different kinds of research

Naturalism

Various qualitative methods are used in health research, but they share some basic orientations (see box).10 The first is a commitment to naturalism, or understanding health behaviour in its everyday context. Results of drug trials may inform practitioners about the optimum effects of therapeutic agents, but even pragmatic trials are not the same as everyday experience. For instance, in today's highly competitive workplace, some people report that they need to seem “healthy” at work and do not want to be seen taking medications. Thus, appearances may be more important to some people than symptom relief.

Interpretation

A second aim of most qualitative studies is that of interpretation: investigating how patients and practitioners make sense of “objective” variables such as peak flow readings. Subjective meanings are crucial to an understanding of how treatment regimens integrate with everyday life. For instance, Adams et al found that half of the asthmatic people they interviewed did not see themselves as asthma sufferers.12 Their “bad chests” were perceived as an acute and temporary problem, better treated with drugs to relieve the symptoms than daily prophylactic medication.

Various strategies that enable asthma sufferers to continue “normal” everyday life, despite symptoms that health professionals would see as objectively problematic, have been reported. Adams et al cite avoidance of triggers such as sexual intercourse, physical exertion, or spending time outdoors in the summer.12 Although these adaptive strategies might seem extreme, patients interpret alternatives, such as using daily preventative medication, as accepting a stigmatising label of “asthmatic.” Thus, findings about patients' possible interpretations of symptoms are essential to understanding likely medication use.

Basic orientations of qualitative methods

  • Naturalism—understanding treatment regimens in an everyday context

  • Interpretation—what meaning do symptoms and treatment regimens have for patients and practitioners?

  • Process—how might these meanings change over time?

  • Interaction—how does communication between patients and practitioners impact on the meaning of medication?

  • Relativism—scientific “reality” may look different from different perspectives

Process

The third contribution of qualitative studies to the evidence base on asthma is the assumption that social life is a process, and that interventions and the changes they bring have to be accommodated within the patient's “biography.”13 Patients with asthma may have to manage medication over a whole lifetime, rather than for the limited duration of even a long drug trial, and integrate an “asthmatic” identity into their other social roles. As the study by Adams et al suggests, for many patients this will involve taking control of their medication, and limiting contact with professionals.12 Conrad's classic study of compliance with medication showed that patients with epilepsy often “tested” themselves to see if their epilepsy had resolved by taking themselves off medication completely or for short periods of time.14 People with asthma also have to integrate symptoms and their management with both the practicalities of everyday life and the psychological “self” over time.12

Interaction and relativism

Qualitative studies often take interaction as a focus of research, rather than a given factor. Katon and Kleinman viewed encounters between doctors and their patients as the bringing together of often conflicting explanatory systems about health and illness, which required negotiation to achieve good outcomes.15 The medical interview may be a meeting between very different views of reality.16 Qualitative sociologists have shown the value of a relativist approach, that focuses on these different explanatory systems.

Attitudes and adherence

One illustration of this from published reports on adherence is general attitudes to medication. Although one “reality” is that medication for asthma reduces morbidity and mortality, and can benefit users with few costs to their everyday lives, qualitative studies suggest a rather different “reality” for patients. Firstly, there is evidence that some patients have negative views about medicines, regarding these as unnatural substances that diminish the body's own ability to fight disease and cause dependence.17 Many doctors, however, make the commonsense assumption that patients are seeking medication.18 This general finding is borne out in specific studies of patients with asthma.19 These studies show that patients worry about becoming physically and psychologically dependent on bronchodilators, and have even more deep seated worries about the long term effects of inhaled corticosteroids.11 Regarding patients' views of reality as ignorant or misguided and attempting to persuade them of the value of a biomedical approach have limited value in increasing adherence, and the need to integrate patients' perspectives has been recognised recently.

Assessing qualitative research

These orientations of naturalism, interpretation, process, interaction, and relativism may be shared by commonsense accounts of medical practice, such as the personal anecdote. There are, however, important differences between anecdotes (stories told for their dramatic or other qualities, without analysis or critical evaluation) and qualitative research. Rigorously conducted qualitative research is based on explicit sampling strategies, systematic analysis of data, and a commitment to examining counter explanations. Ideally, methods should be transparent, allowing the reader to assess the validity and the extent to which results might be applicable to their own clinical practice. The generalisability of qualitative research is likely to be conceptual rather than numerical.22 The studies discussed here, for instance, provide evidence for a number of conceptual issues such as the links between denial of the diagnosis of asthma and medication use or the importance of generally negative views about medication. For the practitioner, the value of these findings is that they are sensitised to issues which could usefully be explored with patients. A number of guidelines now exist to aid both journal editors and readers in assessing qualitative health research.

Conclusion

The argument that qualitative research can contribute answers to questions not easily addressed by randomised controlled trials is not new. Much has been written on the uses and value of sociology for medicine, and indeed there has been a growing acceptance of its methods in healthcare research, including its contribution to randomised controlled trials when these are appropriate.26 However, in the context of the debate about evidence based medicine, it is vital to reiterate that good “evidence” goes further than the results of meta-analysis of randomised controlled trials. The limitations of these trials, and the evidence about barriers to their utilisation by practitioners, should not lead to cynicism about the role of research evidence in health care. We need to be sure that it is the right kind of research to answer the questions posed.

Acknowledgments

Funding: No additional funding.

Conflict of interest: None.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
View Abstract