Intended for healthcare professionals

Tamiflu campaign

The BMJ’s first open data campaign aimed to pressure companies into releasing the underlying clinical trial data for two globally stockpiled anti-influenza drugs, Tamiflu and Relenza. The campaign lasted nearly 4 years and was ultimately successful, and helped galvanize a movement towards increased transparency of clinical trial data.

The bottom line

  • The BMJ’s campaign facilitated the first ever Cochrane review based entirely on clinical study reports and regulatory data.
  • The BMJ’s campaign led to changes in transparency by pharmaceutical companies and triggered inquiries at the national and international level.
  • The BMJ’s campaign heightened awareness of the importance of independent access to underlying trial data, in particular highlighting important discrepancies between journal publications and underlying clinical study reports.

The Tamiflu story

Since the mid-2000s, governments have spent billions of pounds stockpiling two anti-influenza drugs, the neuraminidase inhibitors oseltamivir (Tamiflu) and zanamivir (Relenza). When the so-called “swine flu” H1N1 influenza emerged in 2009, the UK and Australian governments commissioned a rapid update of an existing Cochrane review of the drugs.

As the review team began its work, it received an unexpected criticism and challenge by Keiji Hayashi, a Japanese pediatrician. Hayashi pointed out that the key piece of evidence underpinning the previous Cochrane review’s conclusion--that Tamiflu reduced the risk of secondary complications such as pneumonia--was based on a manufacturer-authored, pooled analysis of 10 manufacturer-funded trials, 8 of which were unpublished.

The Cochrane team set out to respond to Hayashi’s challenge by requesting the unseen trials. Despite the urgency of the situation, however, the reviewers were met with major challenges, as documented in a 2009 BMJ investigation by Deborah Cohen.

John Treanor, lead author of a pivotal Tamiflu treatment trial, told The BMJ: “I did not perform an independent analysis of the primary data, which was not required or requested by JAMA at the time of submission, and I do not have access to the primary data, which I also never requested.”

“When asked a similar question, [Karl] Nicholson,” lead author of the other pivotal Tamiflu treatment trial, “said he did not recall seeing the primary data. He said that the statistical analysis had been conducted by Roche and he analysed the summary data.”

Tom Jefferson, Cochrane review team lead, received a similar response from the authors of the pooled analysis. Jefferson then directed the team’s queries for data to Roche, but was only offered the data under condition of a secret confidentiality agreement, which they refused to sign (that story is
told in this BMJ article published in December 2009).

The campaign

On December 8 2009, Roche publicly promised in a statement printed in The BMJ, to release full study reports to doctors and scientists. But the Cochrane group’s efforts to obtain the promised full clinical study reports were met with numerous refusals. The BMJ launched its first open data campaign to help ensure access to these data for independent scrutiny.

The campaign was based on a key decision to support transparent and accountable analyses and decision making by opposing closed doors and confidentiality clauses.

In 2012, The BMJ intensified its efforts and began using open correspondence as a means of holding specific individuals and organisations to account. Open letters published in The BMJ aimed at (and sometimes achieved) a public response--for example, to NICE (response here) and Roche. And streams of letters published on between the Cochrane researchers and Roche, GSK, WHO, CDC, EMA, and NICE (institutions which which manufactured, licensed, and promoted the drugs)--offers readers the chance to witness attempts to compel greater accountability and responsibility in public health decision making and policy. This was a new tactic that seems to have been successful.

The following facts emerged:

  • WHO was recommending Tamiflu but had not vetted the underlying data.
  • EMA approved Tamiflu, but had not vetted the underlying data.
  • CDC was encouraging the use and stockpiling of Tamiflu on the basis of the 6-page manufacturer funded pooled analysis of 10 clinical trials, but had not vetted the underlying data.
  • CDC’s promotion occurred despite the fact that, since 2000, FDA, which had vetted the underlying data, required Roche to add a statement to Tamiflu’s product labeling: “Serious bacterial infections may begin with influenza-like symptoms or may coexist with or occur as complications during the course of influenza. TAMIFLU has not been shown to prevent such complications.”
  • The majority of Roche's Phase III treatment trials were unpublished a decade after completion.
  • The above facts all remain true today (as of February 2019).

By October 2013, almost 4 years after Roche’s original promise, the Cochrane reviewers had finally received full Clinical Study Reports (CSRs) for 107 studies from the EMA, GSK and Roche (around 150,000 pages), which formed the basis for the latest Cochrane review update published in April 2014 in the Cochrane Library and, in shortened form, in two articles in The BMJ. The reviewers concluded that there was no convincing trial evidence that Tamiflu affected influenza complications (in treatment) or influenza infections (in prophylaxis), and raised new questions about the drug’s harms profile.

In addition to the review and the open correspondence, the Cochrane team made the complete set of 107 full clinical study reports publicly available and The BMJ published all peer reviewers’ comments from 2009, making it possibly the best documented and most transparent review ever undertaken--and certainly the first Cochrane review to use a full regulatory dataset.

Enhanced focus on data sharing and transparency

BMJ’s open data campaign and the Cochrane review raising questions about Tamiflu led some governments to question their policies. In the UK, the Public Accounts Committee put Tamiflu stockpiling and access to clinical trial data on center stage. BMJ editor in chief Fiona Godlee testified before the Committee, and the House of Commons issued a report calling for greater transparency.

At the European level, the Council of Europe held hearings in 2010 that questioned the appropriateness and transparency of governmental decision making with respect to H1N1 influenza. Calling for more transparency, the Council questioned the sufficiency of the scientific evidence base to justify governments’ use of pandemic vaccines and antivirals. At the first hearing in January 2010, the WHO defended itself against accusations of conflict of interest. “Let me state clearly for the record. The influenza pandemic policies and responses recommended and taken by WHO were not improperly influenced by the pharmaceutical industry.” In response to the Council of Europe and BMJ investigation, WHO again defended itself but vowed to tighten its procedures.

But months later, another BMJ investigation reported that scientists advising the WHO on its pandemic planning had done paid work for influenza antiviral manufacturers--conflicts of interest that WHO had not publicly disclosed.

The campaign for Tamiflu data also help galvanize the efforts of those campaigning for greater transparency of all clinical trials. Since 2012, major pharmaceutical companies began to announce new policies on third party access to data from their clinical trials, with many erecting systems to streamline the process. In 2013, The BMJ co-founded the AllTrials campaign.

The BMJ also reviewed its own policies following the launch of the open data campaign. In 2012, The BMJ announced that, for drug and medical device trials, they would “be considered for publication only if the authors commit to making the relevant anonymised patient level data available on reasonable request.” That policy was extended in 2015 to all clinical trials.

Public health recommendations

CDC has not changed its recommendations in response to the Cochrane team’s work based on full Clinical Study Reports. In 2012, CDC released a statement on its website “CDC Recommendations for Influenza Antiviral Medications Remain Unchanged.” (Cochrane reviewers Peter Doshi and Tom Jefferson posed a series of questions to CDC about its statement. Their questions - and CDC’s response - can be read on in open correspondence.)

There was however one notable change in recommendations. In 2017, WHO downgraded the status of oseltamivir on its Essential Medicines List. (The Cochrane group twice petitioned to have it removed.) In response, Timothy Uyeki, the CDC’s Chief Medical Officer for its Influenza Division, led a multi-party rapid response to The BMJ objecting to an editorial published in The BMJ. Uyeki, along with co-authors from the European Centre for Disease Prevention and Control, UK Department of Health, Public Health England, World Health Organization, Infectious Diseases Society of America, Pediatric Infectious Diseases Society, American Academy of Pediatrics, and American College of Obstetricians and Gynecologists, clarified their views on the evidence base for Tamiflu, something they say is “critically important to avert confusion among clinicians providing care for influenza patients worldwide.”

In its rapid response, CDC and other public health agencies highlight the conclusions of observational studies over randomised controlled trials.

The Cochrane team countered that CDC et al.’s argument was “weakened by its selective citation of evidence, and the fact that none of their organizations claim to have independently scrutinized the more complete evidence from clinical study reports.”

They also note that “this shift of focus away from randomised evidence has increased in recent years as it has become clear that the efficacy of oseltamivir is limited -- something the official FDA-approved oseltamivir prescribing information has always made clear -- and evidence for important clinical outcomes such as a reduction in pneumonia and death remains unproven.”


8 Dec 2009 Roche point-by-point response from Roche to BMJ questions (Letter).
8 Dec 2009 Complications: tracking down the data on oseltamivir (Investigation)
8 Dec 2009 Neuraminidase inhibitors - the story behind the Cochrane review (Analysis)
8 Dec 2009 Neuraminidase inhibitors for preventing and treating influenza in healthy adults (Research)
4 June 2010 WHO and the pandemic flu “conspiracies” (Investigation)
23 Oct 2012 Health minister agrees to meet academics to discuss access to clinical trial data (News)
29 Oct 2012 Tamiflu: the battle for secret drug data (Feature)
29 Oct 2012 Clinical trial data for all drugs in current use (Editorial)
29 Oct 2012 Open letter to Roche about oseltamivir trial data (Open Letter)
12 Dec 2012 Withdraw approval for Tamiflu until NICE has full data (Open Letter)
7 Feb 2013 GSK backs campaign for disclosure of trial data (News)
9 Apr 2014 The Tamiflu trials (Editorial)
9 Apr 2014 Neuraminidase inhibitors for influenza (Editorial)
9 Apr 2014 Zanamivir for influenza in adults and children (Research)
9 Apr 2014 Oseltamivir for influenza in adults and children (Research)
9 Apr 2014 Oseltamivir: the real world data (Analysis)
10 Apr 2014 Multisystem failure: the story of anti-influenza drugs (Analysis)
10 Apr 2014 Tamiflu: “a nice little earner” (Feature)
10 Apr 2014 The missing data that cost $20bn (Editor's Choice)
5 Feb 2015 Why aren’t the US Centers for Disease Control and Food and Drug Administration speaking with one voice on flu? (Investigation)
12 July 2017 WHO downgrades status of oseltamivir (Editorial)
16 Oct 2017 Co-authored rapid response to The BMJ by co-authors from CDC, ECDC, UK DOH, PHE, WHO, IDSA, PIDS, AAP, and ACOG (Rapid Response)
2 Nov 2017 Cochrane co-authors respond to CDC, ECDC, UK DOH, PHE, WHO, IDSA, PIDS, AAP, and ACOG (Rapid Response)

New! - Tamiflu timeline

We have created an interactive timeline that shows articles about Tamiflu which were published in the BMJ, between 2003 and 2012.

We have so far included the major influenza outbreaks, and will continue to add other external events.