Are clinical trial data shared sufficiently today? NoBMJ 2013; 347 doi: https://doi.org/10.1136/bmj.f1880 (Published 09 July 2013) Cite this as: BMJ 2013;347:f1880
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
I recently posted a response to "Are clinical trial data shared sufficiently today? Yes"
I would like to make the following comments to B. Goldacre paper which I also want to endorse.
(1) Data posted in the response to Castellani (US phase 3 trials) also endorses Ben Goldacre statements.
(3) Data (ClinicalTrials.gov & other registries + different study phases) endorses Ben Goldacre comments except for the following quotation:
"The most current review—with no cherry picking permitted—estimates that around half of all trials for the treatments being used today have gone unpublished; and that trials with positive results are twice as likely to be disseminated."
- The percentage of unpublished interventional studies is higher outside the US, in other registries, and phases.
- There is also evidence of unpublished & unregistered clinical trials.
- "around half" is probably underestimation according to data:
Tamsulosin >75% unpublished
Dabigatran >80% unpublished
Agomelatine >80% unpublished
Liraglutide: 68.5% - three times as likely to be disseminated (3.1 to be specific) & sample sizes doesn't make any sense.
- Evidence of unpublished & unregistered clinical trials.
- Many examples of misleading clinical trials.
- Many examples of publication bias in clinical trials.
- Clinical trials attempt to test interventions assuming linearity when in reality we (humans) are complex and chaotic (i.e., nonlinear dynamics).
- We are probably more complex than a soccer match & this is the best we got in World Cup predictions:
- Why not all observational studies & other types of pharmacological evidence?
- Why yes? many reasons.
- Evidence can be collected from anywhere (e.g., Google, Google Trends, PubMed, ClinicalTrials.gov, etc).
(4) Online pharmacies (new web sites in the last hour - google search query):
We are engaging in long discussions and debates while 30-50 online pharmacies are created per hour matching names of drugs displayed above. Those links show new websites to sell each one of this drugs (present time to 1h before). Be careful anyone because some of those URLs have security and privacy threats.
I think experimental interventions are also needed outside. "Conceptuality is subjective; realization is objective." -- B. Fuller
Addendum: response to Jacobs A: about "zombie statistics" (words included in the previous title of this letter) **
* Not an error: I'm not citing my response in this URL
** I asked for retraction but no response
Competing interests: I promote open data & transparency I endorse and support AllTrials & the BMJ OpenData Campaign
Goldacre is absolutely right to call for all trials to be published. As he says, transparency is an integral part of the scientific methods.
However, his use of statistics to make his point seems questionable. He states:
The most current review—with no cherry picking permitted—estimates that around half of all trials for the treatments being used today have gone unpublished
I'm really not sure it does say that. I couldn't find the place in the review Goldacre cites where it says that around half of all trials for the treatments being used today have gone unpublished. If I missed it, perhaps Goldacre could point to which page of the article makes that claim?
Also of note is that although the review was published in 2010, much of the evidence it cites was generated much earlier than that. And despite what Goldacre seems to believe, publication practices have changed considerably in recent years. More recent evidence suggests that more like 80% of trial results are disclosed these days.
So when calling for better quality evidence, is it too much to ask that Goldacre bases his own statistics on good quality evidence?
Competing interests: I run a company that provides professional medical writing services to pharmaceutical companies and academic researchers, but more often to the former.
Ben Goldacre lists four levels of clinical trial transparency, three of which are priorities for the AllTrials campaign in his July 9, 2013 “Head-to-Head” remarks. While we agree with and support his laudable efforts, we would like to point out that - and this is amazing - the first three of these priorities have been available for years for drugs approved or discussed by advisory committees in the US, yet, these free publically available documents are rarely used.
These documents are prepared by US FDA scientists to comply with The Freedom of Information Act, which applies to the entire government, not just the drug approval process. Approval Packages are the decisional documents generated from the data, often “raw” data submitted by manufacturers in support of new drug approvals. These documents are rigorous analyses, not simple reviews of application submissions. They are accessible on the Internet at http://www.accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm for drugs approved since 1998.
In addition, Briefing Documents are prepared for discussion for drugs that are, in general, not yet approved in the US, as well as for agents discussed by advisory committees for drugs requiring advisory committee meetings since 2000. These documents are important outside the US as some drugs denied approval by the FDA on safety grounds might be marketed in other countries. The atypical antipsychotic sertindole is an example. The FDA analysis found that sertindole increased the risk of sudden cardiac death compared to risperidone with no trade-off in improved efficacy.
Safety issues involving rofecoxib and rosiglitazone were identifiable in Approval Packages and Briefing Documents for these drugs. The 1999 Approval Package for oseltamivir indicated that the manufacturer had not submitted data that the drug would reduce the serious complications of influenza. This has been reflected in the professional product labeling for the drug since at least 2000.
Even as we advocate for increased trial transparency, we urge the clinical community to make use of these currently available documents a top priority, as this can have immediate consequences for patient safety and well-being.
Larry D. Sasich, PharmD, MPH, FASHP
North Bay, ON, Canada
Elia Abi-Jaoude, MD
University of Toronto
Competing interests: No competing interests
I wish you had included data from Care Pathways which include treatment recommendations - the most skewed and biased subjective assessment goes on - the 'Liverpool Care Pathway' being a classic example. Or doesn't collecting and analyzing 'variances' in treatment and analyzing their effect on outcome constitute 'research' that affects patients ? According to the Round 1 -3 National Audits of Care for the Dying, ALL patient datasets were anonymised before submission, yet no one has been given access to these apart from the copyright holders - hardly the level of 'transparency' I would like to see from the Department of Health, as a patient or a taxpayer.
Competing interests: No competing interests