Generalisability for the generalist
We were heartened to read the open letter by Greenhalgh and colleagues, which echoes our own frustrations and those of others who wished they had been signatories. We were disappointed by the editors’ response.
The editors’ suggestion that qualitative research is really only pertinent to a specialist audience is incorrect. As outlined in the open letter and corroborated by many of the rapid responses, we believe that the BMJ’s readers are no less interested in the findings of high-quality qualitative research than the specialists who call for more of this material in the journal. William Osler’s often-quoted dictum ‘Listen to the patient. He is telling you the diagnosis’ is a clear reference to the traditional primacy of qualitative ‘histories’ in ‘help[ing] doctors to make better decisions’ — the stated aim of the BMJ — and all doctors depend on eliciting qualitative evidence from their patients, listening to their symptoms, concerns and circumstances to guide their clinical practice. Although the BMJ does provide an important and welcome platform for patients’ views, for example in its new section on ‘What your patient is thinking’, as scientists we think a leading medical journal should be at least as concerned with scientific enquiry in this area. Qualitative research has established rigorous methods and can inform clinical practice, as Greenhalgh and colleagues point out in respect of the recent ‘top twenty’ BMJ articles:
'The three qualitative papers explored how primary care clinicians develop and use collective “mindlines” instead of written guidelines; what worries parents when their preschool children are acutely ill; and the nature of collusion in the doctor-patient relationship when death is imminent. They have been cited by 572, 197, and 114 subsequent papers respectively (Google Scholar data). In contrast, the three nominated randomised trials have been cited by 321, 78, and 38 subsequent papers.'
Finally, the editors rightly point out that qualitative research does not provide generalisable answers. However, it does not claim to do so, and at least the transferability of insights from qualitative research is rigorously debated and reflected upon — often more so than the transferability of findings from quantitative research of high internal validity and low external validity. Single randomised controlled trials often do not provide generalisable answers either.
1 Greenhalgh T, Annandale E, Ashcroft R, et al. An open letter to The BMJ editors on qualitative research. BMJ 2016;352:i563.
2 Pitkin R. Listen to the patient. BMJ 1998;316:1252.
4 Wilson P, Petticrew M. Why promote the findings of single research studies? BMJ 2008;336:722.
Competing interests: We have had quantitative and qualitative papers accepted and rejected by the BMJ.