Consumer health information needs to be rigorous, complete, and relevantBMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7255.240 (Published 22 July 2000) Cite this as: BMJ 2000;321:240
- Ruairidh Milne, senior lecturer in public health medicine (, )
- Nina Booth-Clibborn, research associate,
- Sandy Oliver, research officer
- National Coordinating Centre for Health Technology Assessment, Mailpoint 28, University of Southampton, Southampton SO16 7PX
- Department of Public Health and Epidemiology, University of Birmingham, Birmingham B15 2TT
- Social Science Research Unit, Institute of Education, University of London, London WC1H 0NS
EDITOR—Barker and Gilbert emphasise the importance of evidence being relevant to those who may use it.1 Their example is patient involvement in decision making in health care. As members of a working group for the Centre for Health Information Quality (www.hfht.org/chiq/), we considered what “evidence based” means for consumer health information about treatment effects.2 We suggest that three dimensions need to be considered: rigour, relevance, and completeness.
All health information, including that for consumers, needs rigour or the information is inaccurate: it tells you something that isn't true. Searches that maximise rigour are well developed. But the content of what is found is important: its relevance to patients' concerns and its completeness.
If information is irrelevant it tells you something you didn't want to know—for example, about treatments unavailable locally or not reimbursed by your health system. If information is incomplete it doesn't tell you all you wanted to know. For instance, women often don't take iron pills in pregnancy because they get constipated and assume that this is the effect of the iron. At present the Cochrane review of iron supplementation doesn't mention constipation,3 and the review of methods to prevent or treat constipation doesn't mention iron4; women's concerns may be addressed by a review currently under way.5
We think that it is helpful here to think in terms of the performance of diagnostic tests. Relevance means that what you find when searching is within the scope of the topic you wish to cover. High relevance is therefore equivalent to positive predictive value in a diagnostic test—a/(a+b) in the table. Completeness means that what you find covers as much of the scope of the topic you have to cover as possible. High completeness is therefore equivalent to sensitivity—a/(a+c).
An important issue for busy people preparing consumer health information is the ease with which rigorous, relevant, and complete information can be found. In searching for the evidence on which to base consumer health information the question remains whether it is possible to draw on the information from related systematic reviews (therefore maximising rigour) found by careful searching (maximising relevance) and covering a wide enough area (maximising completeness). Only if all three possibilities are pursued can patient information be evidence based.