Oseltamivir: over 15 years of data retention and systematic stonewalling.

Data retention by the company; complicit or complacent drug regulators. In 2009, while preparing its meta-analysis of oseltamivir, the Cochrane Acute Respiratory Infections Group found that only 40% of the assessment data had led to published reports, and that most of the data were unverifiable. With support from the British Medical Journal (BMJ), the Cochrane group lobbied Roche and EMA in order to obtain these missing data (3-6). In 2011, EMA provided the Cochrane group with incomplete reports and refused to ask Roche for the missing data (4). The Cochrane group then filed a complaint with the EU ombudsman, and in 2012, EMA acknowledged that Roche had concealed a number of suspected adverse effects of oseltamivir from regulatory agencies (4). Finally, in 2013, Roche released 77 clinical study reports relating to the 82 trials of oseltamivir that the company had funded and that had been carried out between 1997 and 2001 (2,3). This lengthy delay between the end of the trials and publication of their detailed results is unjustifiable and undermines their credibility. Furthermore, they would never have seen the light of day without an administrative appeal. Faced with this lack of transparency, the Cochrane group adapted its standard method ology to the specific case of oseltamivir. In particular, they focused on clinical study reports rather than on published articles. They also launched a broader reflection on the type of documents to be used by Cochrane for future meta-analyses (see p. 52) (3). Roche reacted quickly to publication of the unfavourable results of the Cochrane meta-analysis in March 2014. In particular, the company sent the Cochrane group 69 pages of comments criticizing the results of this meta-analysis in October 2014, and even had the cheek to complain that the authors had failed to ask the company to clarify the data (7).