Editorials

Applying research evidence to individual patients

Evidence based case reports will help

Clinical review p 1660 

At the heart of clinical medicine is an unresolved conflictbetween the essentially case based nature of clinical practice and the mainly population based nature of the research evidence. While clinicians are exhorted to use up to date research evidence to give patients the best possible care, actually doing so in individual patients is difficult. The reasons are well known.¹ The research literature is poorly organised,² largely of poor quality and irrelevant to clinical practice,³ often conflicting, and often not there at all.⁴ The most valid and, at first sight, relevant information may be based on highly selected groups of patients bearing little resemblance to the patient in front of you. And statistical probabilities may mean little to you or your patient. Steering your way through the evidence jungle takes time, skill, and perseverance.

To help readers develop the increasingly necessary art of using research evidence in practice, the BMJ is launching a new type of articlethe evidence based case report. In the first of these (p 1660), Glasziou describes how searching and interpreting the published literature helped him to reach an informed diagnosis in a woman with a chronic cough.⁵

Evidence based case reports will attempt to show how evidence can be applied at all stages of patient care. Information from cohort studies about the frequency of different conditions can suggest the most likely diagnosis. Decisions about which tests to order can be guided by information on the sensitivity and specificity of different tests and, in the case of invasive tests, their adverse effects and acceptability to patients. Decisions about which interventions to advise, if any, can be informed by randomised controlled trials and systematic reviews looking at the comparative effectiveness, safety, and acceptability of the various options. Information on long term or rare side effects can be gleaned from well designed cohort or case control studies. When a thorough search of the literature fails to find appropriate, high quality evidence, the case reports' contributors are encouraged to say so, since it may be helpful to know that the gaps are in the evidence rather than in your own knowledge.

Case reports have long been used to report new findings and to give educational impact to review articles. Evidence based case reports will not report new findings. General medical journals have largely stopped publishing original case reports of this sort,⁶ recognising that these can give a clinical audience no reliable information on the cause of a condition, its prognosis, the usefulness of a diagnostic test, or the effectiveness and safety of an intervention. Exceptions to this rule are reports dealing with adverse events, where journals have a duty to report any reasonable information they receive. This includes case reports in which the adverse event might be reasonably linked to an intervention on the basis of temporal relation, dose response relation, reversibility on withdrawal, recurrence on rechallenge, or physiological rationale.

Instead of presenting new findings, evidence based case reports are intended to illustrate a process. Contributors are being asked to take an approach now familiar to students of critical appraisalto define the clinical question; search the literature for studies of appropriate relevance, design, and quality; apply the information; and audit the result.⁷ Explicit methods will allow readers to see how the authors reach their conclusions. Because they will be based on real patients seen by generalist clinicians in primary and secondary care, the case reports will also, we hope, provide reliable updates on the management of common clinical problems.

The first few evidence based case reports have been commissioned, but we hope many others will be submitted. All published contributions will be peer reviewed. We prefer reports on common or important conditions relevant to a general clinical audience. Guidance to contributors will shortly be appearing on the BMJ's website (www.bmj.com).