An article reporting the results of pharmacoepidemiological studies on the use of hydroxychloroquine (Plaquenil°) in covid-19 has been retracted by its main authors, shortly after its publication in The Lancet. Another study in patients with covid-19, using the same database, was published in the New England Journal of Medicine, then retracted.

The main reasons cited by the authors for these retractations centred on the fact that access to the raw data was so restricted that it was impossible to respond to criticisms of the study on hydroxychloroquine.

Lack of transparency over raw data from drug evaluations is a recurrent problem with potentially harmful consequences for patients and for the acquisition of knowledge about drugs. Prescrire was already questioning the scientific community back in 2012 as to whether authors from university hospitals who are named on articles reporting the results of clinical trials are granted access to the raw data for their analyses. A study published in 2011 showed for example that many clinical trial protocols stated that the company funding the trial owned the data. And only 26 of the 39 authors who responded to a survey reported having had access to the raw data. The remaining authors had access to data that had already been processed to some degree (coded, compiled, analysed, etc.).

Access to raw data from clinical studies and pharmacovigilance activities is the only way of examining how the data were coded, collated and analysed, in order to verify their reliability and veracity. For example, in the early 2010s, a Cochrane Review Group had to fight for several years to obtain data held - and withheld - by Roche on oseltamivir (Tamiflu° or other brands) in the treatment of influenza. When the Group finally obtained the data, their analysis revealed major methodological flaws in the drug's evaluation that had been concealed by the company, while various countries, including France, had been stockpiling oseltamivir on the basis of these trials.

In 2015, independent researchers obtained access to the raw data from a trial to evaluate paroxetine (Seroxat° or other brands) in adolescents with depression. Their re-analysis of the data showed that in the original article reporting the results of this trial, the authors and the drug company had exaggerated the drug's efficacy while playing down its adverse effects and, in particular, concealing the risk of suicide.

A European Regulation on clinical trials adopted in April 2014 stipulated that data in clinical study reports are not trade secrets. Despite this, it remains difficult to obtain full transparency over the evaluation of medicinal products from pharmaceutical companies and public medicines agencies.

In practice, drug evaluation has various weaknesses that were accentuated in this case by hasty publication in response to the covid-19 pandemic. Many interests are at stake. Healthcare professionals must constantly be prepared to adjust their practice, in view of the continually evolving state of knowledge.

Sources:

• "Défaut d'accès aux données d'évaluation des médicaments : un problème récurrent, non spécifique du covid-19" Application Prescrire 8 June 2020.
• "Retraction - Hydroxychloroquine or chloroquine with or without a macrolide for treatment of covid-19 : a multinational registry analysis" thelancet.com 5 June 2020 > HERE
• "Retraction: Cardiovascular Disease, Drug Therapy, and Mortality in Covid-19" nejm.org 4 June 2020 > HERE

For more information:

• "Covid-19: Follow Prescrire's independent, evidence-based analysis of the pandemic" > HERE