Research misconduct

Last year, 51 cases of major research misconduct were reported to COPE, the Committee on Publication Ethics in the UK.The Association of British Pharmaceutical Industries (ABPI) has identified a further significant number of high-profile cases of scientific fraud, and the General Medical Council (GMC) in the UK is overloaded with similar cases.There can be little doubt that our current mechanisms for preventing, identifying and acting on research misconduct are inadequate, and this was the reason for the third COPE seminar, set up to debate the proposition that the UK needs an independent review body to deal with biomedical research misconduct.Fraud and misconduct can take place at many points along the research continuum, between the research question and the research publication.The process begins with trial design, issues of consent and confidentiality, and the need to obtain Research Ethics Committee approval, antedating the research process itself.The attitudes, beliefs and, indeed, honesty of researchers are likely to be tested along this route; given the alarming findings about the lack of honesty of a cohort of undergraduate medical students in the UK, it is perhaps unsurprising that misconduct occurs. 1 Further opportunities for falsification occur at the output end of the research process, and peer review itself, particularly the selection of conference abstracts and papers, is fraught with problems.Finally, the publication of research papers offers a range of opportunities for unethical behaviour, including salami, prior, duplicate and multiple publication, gift authorship and falsification of results.The definition of research misconduct needs to embrace many of these potential lapses, but also needs to be precise enough to ensure that any body engaged in its prevention or investigation knows exactly what it is doing.In 1999, the Royal College of Physicians' report included phrases such as 'piracy, plagiarism and fraud' and 'theft, fabrication and destruction'. 2A more recent definition would include fabrication, falsification, plagiarism and destruction of data in the course of proposing, carrying out and reporting research, including failure to follow protocols and collusion with others, permitting research misconduct.Publication misconduct also includes failure to publish, failure to declare conflict of interest and causing deliberate delays in the peer review process, often by bringing forward false criticisms of papers under review.