BMJ  2004;329 (30 October), doi:10.1136/bmj.329.7473.0-h

Editor's choice

From optimism to hubris

This theme issue—on whether evidence based medicine makes a difference—makes balanced nods in the direction of both advocates and critics of EBM. It begins with optimism and ends with hubris. Yet among the systematic reviews, randomised controlled trials, anecdotes, and speculation a picture does emerge. There are positive answers to many questions that are asked about EBM: Does it benefit patients? How do you teach it? How do you change practice?

The clearest messages are probably about how to teach it and how to change practice. Arri Coomarasamy and Khalid Khan's systematic review (p 1017) shows that when teaching EBM is integrated into clinical practice it improves skills, attitudes, and behaviour as well as knowledge. The importance of integration with clinical practice fits well with the ethnographic analysis by Gabbay and le May (p 1013) of how general practices handle and integrate evidence. Their answer, through two years of study, is that clinicians rely on "mindlines"—collectively reinforced, internalised, tacit guidelines, formed through interactions with each other and wider networks of "communities of practice." This might sound like jargon—but clinicians will recognise the behaviour.

Gabbay and le May's observations also fit with the results of the analysis of NICE guidelines by Trevor Sheldon and colleagues (p 999). They looked at 12 pieces of "tracer" guidance issued by NICE and tracked changes in practice in UK trusts. Unsurprisingly, they found that implementation was patchy. Guidelines were more likely to be followed when the evidence was strong and stable and clinicians already moving in that direction—and the change was not too expensive or difficult to implement.

Yet it's the sceptics in this issue who have the best quotes. Nick Freemantle wonders about the real return on the substantial resources used to produce NICE guidance (p 1003) and quotes what might serve as an epitaph to many guidelines: "Nothing is impossible for the man who doesn't have to do it himself."

And even when evidence is strong it doesn't move magically from Cochrane review to clinical practice. In his commentary on the patchy adoption of video assisted thoracic surgery (shown in a systematic review to be superior to thoractomy for penumothorax and minor resections (p 1008)) Peter McCulloch applies G K Chesterton's comment on Christianity to EBM (p 1012): [it has] "not been tried and found wanting; it has been found difficult and left untried."

Unlike the high priests in their ivory towers on our cover tending their magic brew, doing EBM well—like most worthwhile endeavours—is hard. As Hilda Bastian says in her personal view, EBM supporters need to learn from their mistakes. A "diehard evidence based enthusiast," she catalogues the errors: weak information on adverse effects; systematic reviews jumping to conclusions too soon; clinicians adopting changes too soon; enthusiasts being arrogant and snobby; hubris.