What is research and publication misconduct

Misconduct can be committed at any point along the research-publication continuum. Although investigators/authors are usually considered to be the main perpetrators of research and publication misconduct, any person involved in the process is a potential offender. Peer-reviewers may fail to declare conflicts of interest when reviewing a manuscript or research proposal.5 COPE has considered a number of cases in which authors have complained about the quality of reviews and expressed concerns that the individual concerned may be using anonymity to hide behind an inappropriately destructive report because of an unhealthy wish to retard the progress of the manuscript and the author's research group. COPE has also seen examples where reviewers have abused the confidentiality entrusted to them and plagiarised the material or ideas contained within a paper or grant proposal.

Similarly editors may breach ethical standards particularly with respect to conflicts of interest. In the same way that authors are now required to declare competing interests, notably commercial affiliations, financial interests, and personal connections, so must editors. Editors can influence the chances of acceptance or rejection of a paper by selecting "hawks" or "doves" to review the paper. Editors have also abused their position and published their own fabricated papers, sometimes bypassing the usual peer-review process.

There are a multitude of misdemeanours that are commonly brought together under the headings of research and publication misconduct.4 At the milder end of the spectrum there are sins of omission; an example would be when experimental design is inadequate to answer the questions raised in the study. Similarly the inappropriate use of statistical analyses may produce inaccuracies both in the final results and in the conclusions reached. This occurs through ignorance but also by intent. Selective presentation of results through data suppression or exclusion may similarly influence the research findings to produce apparently "clear cut" results but that are at the same time misleading. It is always easier to write a paper when everything fits neatly together! Finally the major offences of fabrication, falsification, and plagiarism constitute outright research fraud. It is these major crimes that have been picked up avidly by the daily press, and in the UK have resulted in doctors being struck off by the GMC.

Some scientists argue that even overtly fraudulent research is not as dangerous as many would like to have us believe, since others will repeat the work and, if it is found to be false, will eventually put the record right and make this clear in the biomedical literature. I would argue that this "soft line" makes a complete nonsense of the fundamental premise on which scientific research is based, namely "honesty". In addition, it is a deplorable waste of time and resources to prove that fraudulent research is false. In addition, damage can be done before the truth about falsified research eventually comes to light as illustrated by a paper published in 1993 in the British Medical Journal that was subsequently retracted last year after five years' exposure in the public domain.6 This paper was of sufficient public health importance to change the delivery of services in a region of the UK.

There are instances when there are no questions about a way in which a piece of research was conducted but, in the process of preparing the manuscript for publication and the manner in which the paper is presented to a journal, the authors commit publication misconduct. A common problem centres on authorship. It is now well recognised that individuals' names often appear on a paper whose contribution to the work is questionable, so-called "gift authorship". The Vancouver Group of Editors has published guidelines on what constitutes authorship, which, for example, would exclude a widely held custom to include the head of department as a "courtesy" (Box 18.2).7 There is now a move away from authorship towards contributorship, in which each individual outlines at the end of the paper their individual contributions to the work. There would also be a guarantor who would take overall responsibility for the veracity of the paper.8 Disputes between authors are also common, ranging from failure to get approval from all authors before the final manuscript is submitted to a journal, to changes in authorship or the order of authors during manuscript revisions.

Some authors still submit papers simultaneously to two journals (dual submission) while others attempt to identify "the minimal publishable

Box 18.2 Authorship*

• Authorship credit should be based only on substantial contributions to:

• conception and design or analysis and interpretation of data;

• drafting the article or revising it critically for important intellectual content;

• final approval of the version to be published.

• These conditions must all be met.

• Participation solely in the acquisition of funding or the collection of data does not justify authorship.

• General supervision of the research group is also not sufficient for authorship.

unit" sometimes referred to as "salami-slicing". Most editors would prefer to see a substantial manuscript containing a cohesive story rather than a string of minor contributions in a variety of different journals.

Failure to declare conflicts of interest is another aspect of publication misconduct. Conflicts include direct or indirect financial support from the study, consultancy agreement with a study sponsor, a holding of any patents relating to the study, and any other mechanisms by which financial benefits might accrue as a result of publication of the study. It has been said, "disclosure is almost a panacea".

Was this article helpful?

0 0

Post a comment