In a blistering letter in the current issue of the British Medical Journal, Miran Epstein identifies some of the factors we should consider in assessing the claims of so-called ‘evidence-based medicine’.[1] Nobody rationally disagrees with the suggestion that medicine should have an evidence base, and everybody should agree that in order for medicine to be based on reliable evidence, it should be free of the following ‘polluters’:
- financial conflicts of interest
- inadequately rigorous selection criteria, outcome measures and criteria of statistical significance
- the practice of testing products against placebo or no treatment (rather than current treatment), and then shouting ‘Eureka!’
- recruiting subjects using financial incentives that introduce outcome bias
- marketing campaigns masquerading as research
- research agendas driven by corporate interests rather than patient needs
Yet there’s such pressure on clinicians to be able to assert that their treatment is ‘evidence-based’ that they, in desperation, often give the magic ‘EBM’ endorsement without asking whether the underlying evidence is sound. This has some important ethical corollaries, as Epstein observes. The zeitgeist:
‘…does not regard polluted information as a sufficient condition for rendering disclosure inadequate. Thus it reduces informed consent to a legal fiction and respect for autonomy to a cynical farce.’
I must have written many tens, and very possibly hundreds of thousands of words about informed consent. The journals of medical ethics and medical law are packed full of earnest discussion about informed consent, with long, careful footnotes citing Pico della Mirandola and Hume and Kant and Singer. We’re all so punctilious about our philosophical nuances, and so savage in our peer reviews when we see someone who has overstated the reach of Chester v Afshar or sided too brazenly with Lord Diplock in Sidaway.
Yet isn’t all this care, in the light of the facts, rather ridiculous? Isn’t it classic Titanic deck-chair rearrangement? With (usually) the best of intentions on the part of their clinicians, patients are routinely being fed clinically dangerous misinformation. In the light of a scandal like that, it doesn’t much matter what model of autonomy you’re using, or whether you’re into patient consent or assent or shared decision-making.
It’s unrealistic to expect individual clinicians to conduct an audit of the integrity of the origins of all the information they give to their patients. They have to rely on the summaries they get in the literature.
Journals have a role to play. They need to look beyond the bland and banal declarations of competing interests by authors and peer reviewers, and to look actively for the shady corporate men in suits who stand behind the researchers. But journals can’t do it all. We need a cynical, skeptical, well-funded, well-staffed and ideologically very left-wing regulator. With huge teeth.
References
M. Epstein ‘Wanted: humanistic medicine’, BMJ 2014; 348:g1133 (1 February 2014, p. 23)
Charles
I am in full agreement with you when it comes to the problem of bias research and how this undermines patient autonomy. I hold the somewhat unfashionable belief among some circles, that if we have good quality research and science the number of the ethical and legal issues are greatly reduced and resolution is easier when they do arise.
I assume that your solution to this problem is a joke, or are you seriously suggesting we set up a bloated quango to regulate medical research and science? Maintaining and improving quality is no easy matter, but if you want to do the complete opposite and degrade them the quickest way to do it is regulation. For sure some areas of research need regulation (vivisection, some gene research, use and disposal of chemicals, etc.), but as a general maxim research and science should be unregulated.
As for the clinicians, they do not have to rely upon the summaries they get in the literature. Today’s clinicians have fast access to a mass of information and are not restricted to a few summaries. (Patients have a greater access than they did but this is still limited by the pay-walls set up by the BMJ and other journals. The journals could do a great deal more than they do and I would welcome regulation to tear down the pay-walls.) If clinicians are uncertain about some information (that might be because they spot a suit or two), they should inform their patients and together make a decision.
Not sure what to make of your suggestion that this bureaucracy should be ‘ideologically very left-wing‘. Ideologues of any colour do not quite get science and should be kept well away from having any influence upon it.
Keith: many thanks.
My suggested solution (or part of a solution), was not a joke at all. I wasn’t suggesting a bloated quango, but a lean, predatory one. Nor I am suggesting, as you seem to imply, that ‘research and science’ itself should be regulated. What needs to be regulated is the representation of the results of research.
I’m afraid that your assertion: ‘[a]s for the clinicians, they do not have to rely upon the summaries they get in the literature’ is rather unrealistic. Clinicians in all specialties are terribly hard pressed, the literature is immense, and there is absolutely no way for individual clinicians to be able to evaluate the integrity of the studies about which they read. Sure, there’s ready access to the literature: but that, I’m afraid, is rather more of a problem than a solution.
Charles
Sorry still not convinced. Regulation of the representation of the results is a very effective way of regulating research and science. Part of my research interests many moons ago was the “regulation” imposed by copyright and intellectual property rights. I studied other regulatory systems and broadly came to the conclusion that regulating the dissemination information should be avoided wherever possible.
Not sure I accept the “I’m too busy to do my job properly” excuse some clinicians use. It has never been easier to access and evaluate information, but if they are having problems they should look to their own colleges and “union” to get a change in their working conditions to allow them more reading time. (The ‘RCP Parliamentary Briefing. Medical Workforce: New Deal and EWTD’, 2011, does not address this problem) It is these bodies, universities and journals that should provide more meta-analysis of research results and identify corporate interests. (The emphasis being on “identify“, not banning as the BMJ and other journals do when it comes to the tobacco industry). Obviously better training in statistics and how to quickly read research papers would help. The development of systematic literature surveillance systems would also greatly assist clinicians. (See ‘How much effort is needed to keep up with the literature relevant for primary care?’ B.S. Alper, et al, http://www.ncbi.nlm.nih.gov/pmc/articles/PMC521514/ )
Personally I would be happy if my doctor read my notes.
Where to draw the distinction here between the problem and the solution here is an interesting question, and an informative one.
If the problem is physicians’ reliance on unreliable information, the solution (greater reliability) needs to be brought in at some point. At some point, that is, in the long epistemic chains that lead from researcher, (indeed observer), through publisher […] to the physicians’ acceptance of (total) evidence e (for hypothesis h at t, etc)as reliable. That’s a long chain.
Many proposals (from EBM) emphasis the individual (epistemic) responsibility of physicians. Charles thinks this may be misplaced (if I understand correctly). Keith isn’t so sure. At least if the alternative presented has problems of its own. But is that the problem? A question of whether or not individual physicians are executing their epistemic obligations (and what those obligations are)?
Critical appraisal skills add to the physician’s armoury (filtering publications on a scale of reliability by spotting study design flaws). As Keith rightly points out, current (and future) electronic search tools offer unparalleled reach for the conscientious physician. There is at least a case to me made here that physicians could “do better”.
But how far are we expecting them to go? Is it the place of the individual physician to vet all their information? Some of it? When can they safely rely on research they themselves never conducted? We all have to rely on things without “seeing them for ourselves”. Perhaps most of our knowledge is of this form. Somewhere they are going to have to stop questioning, rely (or trust), and act on that basis.
So is the problem more that, somehow, the processes that place information in physicians’ hands is unreliable? That a lot of garbage (read: bias) gets through? If this is the way to take it, a number of potential solutions present themselves. Watchdogs and agencies, checkboxes in an ethics approvals process, publication criteria, and of course better ways to mark all the likely biases introduced by the big scary corporations and their (eminent) ghostwriters. (Rebecca Kukla is great on this last one.)[1] Of course, we could just ask more of physicians themselves.
An enormous number of potential solutions (obviously) lie unmentioned here. Clarity on the problem will help to limit the search for solutions to those that press on the more critical point in the process.
[1]. Kukla, R. (2012). “”Author TBD”: Radical Collaboration in Contemporary Biomedical Research.” Philosophy of Science 79(5): 845-858.
Comments are closed.