Evidence Based Medicine, as originally described by Prof Sackett, seemed a good idea at the time. Essentially, he said we should be aware of the sum of the evidence from research studies before we embark upon a specific therapeutic intervention for a person. Sadly, it’s become a hugely distorted idea, not least because experimental evidence from group studies (RCTs) has been given not just top rating, but often, the only rating – some authorities use this blunt measure to cut services – if there aren’t sufficient RCTs (or better, meta-analyses of RCTs) showing a statistically significant treatment effect greater than placebo, then the service should be axed.
This idea assumes that individuals are all the same, and that the results of the RTCs are generalisable to absolutely everyone. Sackett never said that. He was careful to say the research study analysis should be one part of clinical decision making, but the other part should be individualisation to this person – taking their values, needs and preferences into account.
If people had held onto that, they wouldn’t make stupid remarks like “If a patient is prescribed an evidence based drug and say they are no better, they have either not taken the drug, or they are lying” (as a young doctor last year told me they had been taught by a clinical teacher in Glasgow)
But now it appears, the “evidence” on which “evidence based medicine” is based, isn’t quite what its been cracked up to be. So here are four steps we could take to improve this approach.
Firstly, publish ALL the evidence. Many researchers have found that, using Freedom of Information requests, it turns out that drug companies just don’t publish all the data. (see my second point in that link to an earlier post). If only the studies showing better results are published, the “evidence base” is seriously distorted.
Secondly, downgrade the value attributed to trials which only use surrogate health markers. If you only measure one or two parameters, but don’t relate those parameters to change in health experience, you aren’t demonstrating useful impacts on health care.
Thirdly, beware of studies which only show “relative” as opposed to “absolute” changes in outcome. Read any of Gigerenzer’s books – you won’t swallow medical statistics the same way ever again!
Fourthly, and probably most importantly, stop assuming that RCTs and meta-analyses of RCTs give the “truth, the whole truth, and nothing but the truth”. The evidence base is incomplete – it always will be. Publication and statistical choices are chosen by those with vested interest in the treatments. And, finally, we are ALL different. NOBODY except me can tell you whether or not a particular treatment has made me less dizzy, less nauseous, reduced my pain, improved my well-being….in short, made a difference to my life.
Compassionate care requires careful listening to a patient, believing and valuing what they say. Enablement requires a clear presentation of options – yes choices – to a patient and helping them to make their own decisions about treatments.
I’m fed up with the reduction of individuals to a mass to be controlled. “Compliance” and “elimination of variance” are used in authoritarian ways. It’s time to move away from thinking a professional’s views and beliefs (wherever they come from) are the only ones to consider, and to seek genuine partnership instead with professionals who care.
Leave a comment