What Is E B M About

Philosophy of Science and EBM

What is it that EBM is about?

- I am wondering whether it could be argued that some (?most) of the philosophical literature available on EBM misconstrue what it is that EBM is about - It seems that the evidence hierarchy can be retained by construing it as pertaining to treatment decisions and being a little more inclusive of other forms of evidence then is typically permitted. There appears to be two points which the philosophical literature assumes EBM proponents will not accept, that I would like to argue are contentious: (i) Randomisation is a relative good. Other methodological issues do matter. (ii) The kind of questions the EBM hierarchy can/should be applied to are far more limited then often assumed. The EBM hierarchy is particularly concerned with decisions related to whether to treat a certain population of patients with a drug. The need for further evidence sources, including the inadequacies of RCTs, for other questions—such as diagnosis, individualisation of treatment etc—are recognised. - Examples of the literature: Worrall, Phil of Sci 2002; Bluhm, Pers Biol and Med 2005; Grossman and Mac Kenzie, Pers Biol and Med 2005 - Worrall and Grossman’s focus appears to be to counter the statement: Evidence from randomised trials is ALWAYS better then evidence from non-randomised trials. While there is some sloppiness in much of the medical literature discussing EBM, I don’t think that too many EBM proponents would have a problem with the revised statement: All things being equal, when considering treatment decisions on the basis of trials testing specific pharmacotherapy, randomised trials provide ‘better’ evidence (in terms of more reliable/less bias) than non-randomised trials. {[green Sure. But the fact that you can find a statement which both sides agree with doesn’t mean that there’s no disagreement. And that would be the case even if the statement you found covered all the ground. You mustn’t forget that people believe contradictory things! And this is important for thesis technique: examiners will notice if you’re too kind to your opponents’ logic. Jason ]} {[blue Thanks, agreed. The reason for the gesture to this statement, however, is that it seems to be the departure point at which the philosophical literature begins. Delineating the areas of substantive disagreement outside of glossily worded statements from EBM proponents seems to be a fertile area in the debate. A ]} As Grossman, states the ceterus paribus clause is doing much work here — but provided it is recognised it does not seem to be an argument against EBM (as construed here). - Bluhm makes a number of great points against Peto et al but seems to spend a lot of time pointing out that the basic sciences (laboratory research) is important in addition to epidemiological research (RCTs). What Bluhm highlights is that the basic sciences is important for diagnosis and individualisation of therapy—points which under my construal of EBM are accepted by proponents.

Having a quick look, many of these points can be seen in Sacket, BMJ 1996.

Counter-arguments/Considerations - I have probably taken a more combative tone here than necessary. The alternative would be to start with the question of what EBM is about and then use the available philosophical literature + some additional arguments to carve out what EBM MUST be in order that it can tell a consistent story. - {[green Yes, that’s the right thing to do … if you think (which I don’t) that it’s a good idea to retain the terminology “evidence-based medicine”. Jason ]} - What I am not particularly interested in is trying to nut out precisely what EBM proponents do, actually, hold—while having said that I do think it is potentially more sophisticated than sometimes assumed in the philosophical literature (but then one does only have what the EBM proponents write to go on) - {[green “EBM proponents” is ambiguous. Personally, I’m with you in not being interested in (1) what the theoreticians of EBM say. But I’m very interested in (2) what effect the EBM movement has on what people do. Now, you obviously don’t have to agree with my interests. But I think it’s very important for you to mention the difference between those two categories. ]} - {[green Also, just in case you’re even a bit interested in what the EBM theoreticians say, I think you’re wrong about them agreeing with you — e.g., Glasziou, who’s one of the more progressive EBM theoreticians, explicitly disagrees with the conclusions of Grossman & Mackenzie. Jason ]} {[blue Thanks, will be great to chat about this sometime; is there something I can read, or is it from verbal communication ]}

Some great case studies here for the importance of both RCT and observational studies— and the way that the basic sciences tells the story in-between (vital for interpretation of both). {[green Yes. ]} See E B Mvs Basic Sciences for some examples of where RCT has improved observational And Hallas et al. BMJ 2006 (pre-rel) on antithrombotics and UGI bleeding for the importance of case-control. RCT highly idealised use — in practice used much more broadly — only good evidence re bleeding risk comes from case-control.