english.prescrire.org > Spotlight > Archives : 2013 > Health research: fraud is too widespread!

Spotlight: Archives

Every month, the subjects in Prescrire’s Spotlight.

2013 : 1 | 30 | 60

Health research: fraud is too widespread!

It is high time to put an end to the blind faith placed in researchers’ findings, and to make it possible for their results to be verified.

In recent years, there have been numerous articles in the international biomedical press condemning the pharmaceutical companies’ manipulation of clinical trial data. Conversely, researchers working for the pharmaceutical companies complain about the manipulation of the results of preclinical studies by independent researchers.

A team of researchers from a pharmaceutical company sought to replicate 53 preclinical experiments in the field of haematology and oncology described in major articles. They managed to replicate only 6 of these 53 experiments. Some of the authors of non-replicable experiments admitted having published only the results that confirmed their hypothesis, even though these were non-representative of their overall results. Furthermore, a team from another company estimated that only 25% of the preclinical studies it had tried to verify could be replicated.

An older study revealed that a third of clinician scientists admitted to having carried out misleading manipulations such as ignoring results contradicting their own, or modifying either the trial protocol or the results under pressure from the funder.

This widespread fraud is a consequence of the competitive nature of research, with scientists seeking to further their reputations in view of personal advancement, and to attract funding. It testifies to the prevalence of the unprincipled profit motive over patient welfare.

Now, more and more voices are speaking out around the world to demand a radical overhaul of publication standards for preclinical and clinical studies. Such a change would involve making all the raw data available, independent analyses of the results, and the obligation to publish all the results, including those that do not bear out the sponsors’ hopes.

Access to data was the theme of the discussion-debate at the annual Prescrire Awards ceremony on 31 January 2013.

©Prescrire 1 May 2013

"Research scientists: endemic fraud" Prescrire Int 2013; 22 (138): 116. (Pdf, free)
Download the full review.
Pdf, free