Access to trial data: done and dusted?

March 21, 2016


Honorary Research Fellow Tom Jefferson talks Tamiflu, trials and transparency in his latest blog for CEBM.


Tom Jefferson

Tom Jefferson

In 2009 my Cochrane review group received the support of the BMJ. This was no hairy-fairy vague “I love you too” kind of handshake. Over the period of 4 years the BMJ, and by reflex, the world media, reported our efforts to clarify how many clinical trials of antivirals for influenza had been carried out and to access their full clinical study reports. We were only interested in the “full reports” i.e. the lengthy and complex documents prepared by regulators because we had worked out that the journal versions were biased, ghost-written and selectively summarised – and because Roche had pledged to provide them in the pages of the BMJ.

The BMJ launched its Open Data campaign. Part of it was the publishing of hundreds of pages of correspondence with influenza antiviral manufacturers, politicians, WHO, CDC, EMA and industry key opinion leaders such as the European Scientific Working Group on Influenza (ESWI) and its alter ego MUGAS. Readers could read for themselves the questions we asked and the evasive answers we got. The documentation is still available here.

Partly because of the BMJ and numerous other campaigns, and partly because of other factors described elsewhere, we were buffeted by august organisations and even pharma in a mad scramble to set up web-based resources and documents dispensing knowledge. The BMJ campaign seems now to have been subsumed by a rush of organisations and individuals to join the ranks of those agreeing that data from experiments on humans cannot be kept secret. The European regulators EMA agree to this, after having been stimulated to face reality by the Ombudsman, a mere official of the much-despised and derided EU. Remarkably since 2010 the EMA has released millions of pages of clinical trials data, fulfilling the pledge of its Executive Director in November 2012. Now we can also be comforted by the confirmation that from this autumn we will benefit from EMA policy 0070, when all we will need to do is register on a website and we will be able to download everything. Well, that is, everything that’s available following redaction of items judged commercial in confidence) and everything which is included in certain parts of pharmaceutical licensing applications and not in other parts which EMA do not have or are automatically released under policy 0070.

“Everybody seems comfortable in the knowledge that the battle is nearing its end and the forces of light and progress have won. So, how is it that I am struggling to feel a sense of victory, and instead think “we” have actually been very deftly shunted to a dead end?”

Now the spotlight has moved on and the BMJ campaign webpages have not been updated. Everybody seems comfortable in the knowledge that the battle is nearing its end and the forces of light and progress have won. So, how is it that I am struggling to feel a sense of victory, and instead think “we” have actually been very deftly shunted to a dead end?

Evidence on the effect of the bow wave of consensus for releasing trial data is thin on the ground.

EMA has certainly released millions of pages under Policy 0043, but access is laborious and getting even more so as each month goes past. Waiting over a year to get the EMA CSR holdings for Gardasil 4 vaccine is not conducive to timely research, especially if what you get is akin to a PDF jigsaw puzzle. But who else is releasing clinical trial data following simple requests? The answer is no one. Pharma data sharing policies require varied degrees of information, protocols and study plans that must be pre-approved by committees because, they say, they have to ensure “responsible” analyses. They obviously have different standards from EMA, and it is curious that in over 200 different (mainly unredacted or minimally redacted) CSRs I’ve read so far, I could not spot any obviously commercially confidential information nor any trial participants’ names (which by law have to be anonymised).

Canada passed a transparency law in 2014, Vanessa’s Law. Its spirit is that of allowing the public greater understanding of how decisions on pharmaceuticals are made, but the law was left to the regulator (Health Canada) to apply. At present it looks as if getting ingots out of Fort Knox may be easier.

The US FDA does not release clinical trial data but makes approval packages freely available. Approval packages contain additional information of the licensing application including detailed FDA officers’ reviews of the trials, but they never contain the actual clinical study reports.

Distortions have also been allowed to creep into the discourse. For example although since 2010, EMA has been releasing usable individual participant data (IPD) with minor redactions, discussion of IPD data sharing has been under the hammer because of concerns of the potential for misuse or abuse (e.g. re-identification of participants) by anyone “unqualified” to access it. But we don’t know of any re-identification or misuse from the data EMA has released (possibly because they are in paper/PDF form, not electronic). In addition, the conceptual separation of individual data from its parent clinical study report is a dangerous nonsense, as they are part of the same trial. I use the term parent to mean that any portion of data from a clinical trial cannot be divorced from its design and methods which are reported in the full clinical study report.

EMA remains the best source of clinical trial data in the world, albeit with the limitations described.

The huge steps forward achieved in the last
few years risk being compromised by the fickleness of the media and researchers’ lack of intellectual rigour.

EMA’s efforts are not enough because EMA has limited holdings for those drugs and biologics they have made a regulatory decision on. Industry and FDA need to open access to their holdings with no barriers and minimal redactions. Otherwise it feels like history is repeating itself.

Cite as Jefferson T (2016): Access to Trial Data: Done and Dusted? CEBMJ

DOI: 10.13140/RG.2.1.2785.4489


Tom Jefferson is a co-recipient of a UK National Institute for Health Research grant for the systematic review discussed in this article (HTA – 10/80/01 Update and amalgamation of two Cochrane Reviews: neuraminidase inhibitors for preventing and treating influenza in healthy adults and children— This review focused on oseltamivir, manufactured by Roche, and zanamivir, manufactured by GSK.  Tom Jeffersn also co-recipients of a Cochrane Methods Innovations Fund grant to develop guidance on the use of regulatory data in Cochrane reviews.

Tom Jefferson receives royalties from his books. Tom Jefferson is occasionally interviewed by market research companies about phase I or II pharmaceutical products. In 2011-13, Tom Jefferson acted as an expert witness in a litigation case related to oseltamivir and in a labour case on influenza vaccines in healthcare workers in Canada. He has acted as a consultant for Roche (1997-99), GSK (2001-2), Sanofi-Synthelabo (2003), and IMS Health (2013) and in 2014 was retained as a scientific adviser to a legal team acting on oseltamivir. In 2014-15 Tom Jefferson was a member of two advisory boards for Boerhinger. He is a member of an independent data monitoring committee for a Sanofi Pasteur clinical trial.

Tom Jefferson and Kamal Mahtani are co-facilitating a Regulatory Data Workshop as part of Evidence Live.

On Tuesday 21st June 2016 we are running a number of pre-conference workshops at a speacial rate of £155 each. Course fees include catering and course materials.

Workshops will ruimages-2--300x143n at several venues in and around the Radcliffe Observatory Quarter.

CEBM & Evidence Live promote small group learning for all of these workshops

Places are limited, so please book early to avoid disappointment.

One comment on “Access to trial data: done and dusted?

  1. Pingback: Why institutions should audit their trial publication record - CEBM

Leave a Reply

Your email address will not be published. Required fields are marked *