Improving biomedical journals' ethical policies: the case of research misconduct

Scientific journals may incur scientific error if articles are tainted by research misconduct. While some journals' ethical policies, especially those on conflicts of interest, have improved over recent years, with some adopting a uniform approach, only around half of biomedical journals, principally those with higher impact factors, currently have formal misconduct policies, mainly for handling allegations. Worryingly, since a response to allegations would reasonably require an a priori definition, far fewer journals have publicly available definitions of misconduct. While some journals and editors' associations have taken significant steps to prevent and detect misconduct and respond to allegations, the content, visibility of and access to these policies varies considerably. In addition, while the lack of misconduct policies may prompt and maintain a de novo approach for journals, potentially causing stress, publication delays and even legal disputes, the lack of uniformity may be a matter of contention for research stakeholders such as editors, authors and their institutions, and publishers. Although each case may need an individual approach, I argue that posting highly visible, readily accessible, comprehensive, consistent misconduct policies could prevent the publication of fraudulent papers, increase the number of retractions of already published papers and, perhaps, reduce research misconduct. Although legally problematic, a concerted approach, with sharing of information between editors, which is clearly explained in journal websites, could also help. Ideally, journals, editors' associations, and publishers should seek consistency and homogenise misconduct policies to maintain public confidence in the integrity of biomedical research publications.