Peter Doshi: Roche to publicly post trial protocols—just kidding

“Thirdly, we will post the protocol of our trials as they are started and the results of trials once they are completed on two websites which are available to the public, and”

So states Roche in describing one element of its 2013 policy on clinical trial data sharing.

Such a policy would place the Swiss pharmaceutical company ahead of its peers, even GlaxoSmithKline, generally seen as the pharma leader in trial transparency.

The only problem is that does not appear to be true.

I looked up two random studies on GO28888 (which Roche lists as currently recruiting participants) and GO28753 (which Roche says is “active, not recruiting”). Both are oncology trials of unapproved, investigational drugs. Neither webpage includes a link to the protocol.

I wrote Roche, but they suggested I was looking in the wrong place and directed me to the US and the EMA EUdraCT trial registries”.

This sounds fine, but there is one problem: the website presently does not allow for the uploading and public posting of trial protocols; instead what appears is a structured summary of study details such as the purpose and eligibility criteria. (See the entries for GO28888 and GO28753). Nor are the documents available on the European Clinical Trials Register: GO28888 and GO28753.

I asked Roche to clarify. They said I was correct.

“It is correct that depending on the rules and regulations of the respective sites, protocols/attachments cannot be uploaded. For the time being, protocols can therefore only be requested via this link,” a spokesperson wrote, pointing me to Roche’s online platform for requesting access to clinical study data.

I submitted a request for the two protocols, and received Roche’s response a week later: sorry, but the documents are “out of scope”.

“Unfortunately, the documents that you requested have been deemed out of scope due to the fact that both studies are for non-marketed products for which we have not as yet received a Regulatory decision,” stated the letter from

I had one last question for Roche: “May I ask whether there is any plan to make the protocols public? Or will they remain by request only, as now?”

Their response: “At this point in time, there are only plans to use the request link in order to make the protocols available.”

It is hard not to feel a bit of déjà vu here. In 2004, the New York State Attorney General’s office settled its lawsuit against GlaxoSmithKline regarding consumer fraud related to the anti-depressant paroxetine. In the settlement, GSK agreed to publicly post the clinical study reports of its studies of paroxetine in adolescent and pediatric patients. But for eight years GSK neglected to post the appendices of these trial reports. When I asked for them in 2012, GSK refused. After I raised the point with the New York State Attorney General’s office, the company posted an additional 38,781 pages – documents a RIAT team subsequently used to restore GSK’s infamous Study 329.

Between 2012 and 2014, a wave of new transparency policies from industry seemed to usher in a new era of data sharing. But I remain unconvinced that the policies are actually delivering on their promises. It took AstraZeneca—which announced its own “transparency commitment” in 2015—seven months to inform me and colleagues at Johns Hopkins that it would not be providing us with clinical study reports or patient level datasets for use in our PCORI-funded study of seroquel.

Open data advocate Ben Goldacre often talks about the journey towards clinical trials transparency as one of many “fake fixes,” false promises of change that actually did not sufficiently address the problems they pledged to fix. I think we can now add Roche’s inaccessible protocols to that list.
Competing interests: PD is a co-recipient of a UK National Institute for Health Research grant (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. PD also received €1500 from the European Respiratory Society in support of his travel to the society’s September 2012 annual congress in Vienna, where he gave an invited talk on oseltamivir. PD gratefully acknowledges the American Association of Colleges of Pharmacy for its funding support ($10 000) for a study to analyze written medical information regarding the possible harms of statins. AACP had no involvement in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of this manuscript.

Peter Doshi, associate editor, The BMJ