For many years it was possible to start a clinical trial, on a drug, without telling anyone that you were doing so. Then, if it turned out to be negative you could just slip it under the carpet and never let anyone know you were doing it. This is what happened with many antidepressant trials. Positive, publish. Negative, bury. Unsurprisingly, the results for antidepressant treatment looked pretty damned good.
Another little trick was to keep the primary end-point of the trial secret. To explain. Say you started a study on statins where you most wanted to look at the effect on overall mortality – whether taking statins meant more people were alive at the end of the study, than those taking a placebo. In this case overall mortality would be the ‘primary end-point’.
You could also measure other outcomes, or end-points. For example, you see how many people were admitted to hospital with angina, or how many people needed an angiogram and/or stent. Or how many people suffered a non-fatal heart attack or stroke. You can, in fact, measure many different things. These would usually be called secondary end-points.
Up until fairly recently, if you failed to reach your primary end-point – the term normally used for this sorry state of affairs would be ‘failed to reach statistical significance’ – you didn’t need to let anyone know. You could just say. ‘Oh look, the number of episodes of angina was significantly reduced, as was hospitalisation for chest pain and the rate of non-fatal strokes.’ Success!
Man on Clapham omnibus: ‘But, but, I thought you said the trial was going to look at overall mortality.’
Pharmaceutical company: ‘How do you know that, we never told anyone what the primary end-point would be.’
Yes, I know, pharmaceutical companies cannot speak. But you get the general idea.
Now, if you start trawling around your data, you can almost always find something somewhere got better. And if something else got worse, you can just fail to mention it. Essentially, therefore, unregistered clinical trials are not worth the paper that they are written on. Especially, of course, if they never got written on any paper at all.
In the year 2000 the major US group the National Heart, Lung, and Blood Institute (NHLBI) decided that any studies they were going to fund (these are non pharma company studies) must register all end-point/primary outcomes, before the study started. This means that the trial investigators could not manipulate the results post-hoc.
A group of researchers recently looked at 55 large clinical studies funded by the NHLBI between 1970 and 2012 to see if the transparency rules had made any difference. What they found should shake the foundations of medical research…but it almost certainly won’t:
- 57% of studies (17/30) published before 2000 showed a significant benefit in the primary outcome
- 8% (2/25 trials published after 2000 showed a significant benefit in the primary outcome
As the researchers said ‘The requirement of prospective registration in ClinicalTrials.gov is most strongly associated with the trend towards null clinical trials. The prospective declaration of the primary outcome variable required when registering trials may eliminate the possibility of researchers choosing to report on other measures included in a study. Almost half of the trials [published after 2000] might have been able to report a positive result if they had not declared a primary outcome in advance.1’
Pharmaceutical companies have been asked to register trials since 2005.
At this point I am going to try and join two thoughts together. Almost every study done on blood pressure lowering, blood sugar lowering and cholesterol lowering was done before the year 2005. I only choose these three areas as they are the three area of maximum drug prescribing in the world. Billions upon billions are spent in these areas, hundreds of millions are ‘treated’.
The evidence used for this mass medication of the Western World is demonstrably, horribly, biased. Had companies been forced to register their trials prior to publication, positive results would have been reduced by at least 49%. Almost certainly far more. You could put this another way around and say that it very likely that only 8% of studies would have been positive.
We do not know which trials would have been positive, or which negative. Yet we have based the entire edifice of drug treatment, of hundreds of millions of people, on unreliable nonsense. The study in PLOS is only the latest demonstration of this fact. The database of medical research – everything until at least 2005 is a gigantic festering mess. It needs to be stripped out and cleansed.
Do you think this is too strong?
Well I shall now quote Dr Marcia Angell, Dr Richard Horton and Dr Richard Smith. Editors of, respectively, the New England Journal of Medicine, the Lancet and the British Medical Journal. The three highest impact factor journals in medical research.
‘It is simply no longer possible to believe much of the clinical research that is published, or to rely on the judgement of trusted physicians or authoritative medical guidelines. I take no pleasure in this conclusion, which I reached slowly and reluctantly over my two decades as editor of the New England Journal of Medicine.’ Marcia Angell.
‘The case against science is straightforward: much of the scientific literature, perhaps half, may simply be untrue. Afflicted by studies with small sample sizes, tiny effects, invalid exploratory analyses, and flagrant conflicts of interest, together with an obsession for pursuing fashionable trends of dubious importance, science has taken a turn towards darkness.’ Richard Horton
‘The poor quality of medical research is widely acknowledged, yet disturbingly the leaders of the medical profession seen only minimally concerned about the problems and make no apparent efforts to find a solution.’ Richard Smith
Who, in a position of power, will finally wake up and realise that the vast database of medical research stinks of bias and manipulation. Who can we call upon to take up the gigantic and painful task of clearing out the Augean stables?