If I mention evidence-based medicine do your eyes light up, or do your shoulders slump? I’m still an enthusiast for evidence-based medicine — after all, if we don’t use best available evidence in our practice, what else do we have? I still get rather too much enjoyment from a good PubMed search, and I still get excited about journal articles that might improve my practice. But what I see in practice doesn’t seem to match the expectations of students coming through my consultations, or even the expectations of policy makers.
The students have been taught that the path to an answer in any consultation is to formulate a PICO — patient, intervention, comparison, outcome — question, followed by a MEDLINE search and a critical appraisal. The result will be ‘The Answer’. Clinical scenarios can be carefully constructed to demonstrate how the evidence gives you The Answer and tells you ‘What To Do’.
Policy makers, on the other hand frequently seem frustrated at our inability to do what ‘The Evidence’ tells us. While we can be a very frustrating profession, I’m sure, and we sometimes do things we shouldn’t, there’s a general under-appreciation of the complexity of what happens in a consultation and the role of evidence in that.
We’ve been misled if we think the evidence always gives us an answer or tells us what to do. While sometimes that happens, often it doesn’t. Instead of thinking about The Evidence giving us The Answer, we should think of evidence as being a spectrum. At one end of the spectrum we do get a clear answer about what we should do. Low dose aspirin after a heart attack would be an example, where the evidence is clear, and we’d be foolish not to follow. At the other end of the spectrum the evidence shows treatments where harms clearly outweigh benefits. It’s easy to decide not to use medications in this group, and sometimes, as with rosiglitazone or rofecoxib, the drug is removed from the market.
These are probably the minority of cases, though. Far more often, we get information without an answer. It’s often useful information. Knowing the benefits and harms of treating to lower blood pressure targets, or knowing how many fewer hours of tonsillitis are suffered with antibiotics can be really helpful in making decisions. Acting on information like this requires judgement. At other times the evidence is equivocal, or absent. But there’s still a patient and a doctor sat in a room, and between you, you still need to decide on a course of action. That, too, requires judgement.
Saying a treatment ‘doesn’t work’ is shorthand for saying it works no better — and in many cases no worse — than placebo. We could say ‘this treatment works as well as placebo.’ Look at all those who improve in the placebo arm of a trial.
There are also many other questions that come up in a consultation that have no answers in MEDLINE. ‘I’m lonely’ does not equate to any particular diagnosis, but still ends up in symptoms and a consultation. Finding the form of words for people in these situations, even ones to manage expectations without a cure, is a matter of more judgement, but also about being human.
This has always been advocated by those who understood evidence-based medicine, but somewhere an evangelism crept in, making The Evidence our holy grail.
How desirable is that? I’ll let you be the judge.
- © British Journal of General Practice 2016