Scientific misconduct is worryingly prevalent in the UK, shows BMJ survey

One in seven UK based scientists or doctors has witnessed colleagues intentionally altering or fabricating data during their research or for the purposes of publication, found a survey of more than 2700 researchers conducted by the BMJ.

The survey, which was emailed to 9036 academics and clinicians who had submitted articles to the BMJ or acted as peer reviewers for the journal (response rate 31%), found that 13% of these researchers admitted knowledge of colleagues “inappropriately adjusting, excluding, altering, or fabricating data” for the purpose of publication. Just over one in 20 (6%) said they were aware of cases of possible misconduct within their own institutions that remained insufficiently investigated.

The survey, which aimed to describe the extent of research fraud in the UK, was conducted in advance of a high level summit organised by the BMJ and the Committee on Publication Ethics (COPE) on 12 January. The meeting brings together institutions, researchers, and funders to address the problem of research misconduct in the UK.

“UK science and medicine deserve better. Doing nothing is not an option,” said Fiona Godlee, BMJ editor in chief.

Dr Godlee added: “While our survey can’t provide a true estimate of how much research misconduct there is in the UK, it does show that there is a substantial number of cases and that UK institutions are failing to investigate adequately, if at all. The BMJ has been told of junior academics being advised to keep concerns to themselves to protect their careers, being bullied into not publishing their findings, or having their contracts terminated when they spoke out.”

This view is echoed by COPE Chair, Elizabeth Wager, who notes: “This survey chimes with our experience from COPE where we see many cases of institutions not cooperating with journals and failing to investigate research misconduct properly.”

The meeting seeks to arrive at a consensus among members of the UK scientific establishment to both deal with the size of the problem and formulate possible solutions.

The results of the BMJ survey mirror those of a 2001 study conducted among newly appointed hospital consultants in Merseyside, which found about 10% had witnessed their peers altering or fabricating data for the purpose of publication, and nearly 6% admitted they had personally been involved in research fraud (J Med Ethics 2001;27:344-6).

The prevailing attitude that such instances are rare and anomalous is countered by Dr Godlee and Dr Wager in a recent BMJ editorial: “There are enough known or emerging cases to suggest that the UK’s apparent shortage of publicly investigated examples has more to do with a closed, competitive, and fearful academic culture than with Britain’s researchers being uniquely honest.” (BMJ 2012;344:d8357 doi:10.1136/bmj.d8357).