The press agencies like nothing more than bad news – disaster sells. But the opposite seems true in clinical research, according to a recent paper in JAMA (Journal of the American Medical Association).

The paper, entitled “Out of Sight, Out of Mind, Out of the Peer-Reviewed Literature” cites 57 clinical trials in which the US FDA found evidence of wrongdoing.  These amounted to breaches of virtually every rule of Good Clinical Practice and included:

• falsifying data
• inadequate adverse event reporting
• protocol violations
• inadequate or inaccurate recordkeeping
• failure to protect the safety of patients, and
• problems with informed consent.

Apparently these trials led to 78 published academic papers, only three of which mentioned the negative findings. There were virtually no retractions or corrections issued.

The paper accuses the FDA of repeatedly hiding evidence of fraud from both the public and trusted scientific advisers. In at least one case, the author claims that falsified data in a trial comparing chemotherapies led to a patient’s death.

The question that occurs to me is just why would the FDA go out of its way to hide such misconduct – what could it possibly gain from a deception like this?

Rather than hinting at an underlying conspiracy, the paper suggests that a lot of this is the result of publication bias. Peer-reviewed publication usually favours research that is deemed to be “successful” – which misses the entire point of science in my opinion!

Whilst a paper that fails to shed new light on a subject perhaps because of a failing in method, might not be the feather in the cap of the researcher in question, it is arguably at least as important in the overall advancement of science.

How many billions in funds and years of researcher lives have been spent researching dead-end techniques and treatments that others had proven unworkable years ago?

It also fails the rules of Good Clinical Practice in another fundamental way – failure to disseminate negative findings means that future participants in trials will inevitably put at a greater and, crucially, avoidable risk.

In 2013, The Economist reported that negative findings only accounted for around 14% of papers published in science journals – down from 30% in 1990.

Share on LinkedIn