Reporting animal research - worthy of a rethink?

Back in September 2006 the ALPSP held a one-day meeting on ‘publication ethics’. During the day I learnt much about how journal editors and publishers had worked with other parties to improve the reporting of research involving human subjects. I was left wondering if there were similar initiatives for the reporting of research involving animals given that both areas of research can be controversial and give rise to issues of great public concern. To answer this question I began by looking at whether journals had any policies relating to the animal research they publish, and if so what these policies were. This marked the beginning of what has become an annual survey of journal editorial policies relating to the use of research animals (see Osborne et al.1 for survey protocol and analysis details). So, four years on what have I learnt . . . ? Since July 2005 there has been a twofold increase in the number of articles published per year that involve animal research (125,280 in the last year surveyed). There has also been a similar increase in the number of journals publishing four or more such articles per year (currently 2,131 journals). The publication of animal research is therefore an area that would appear increasingly relevant to journal editors and publishers. To date the survey has analysed the ‘publication policies’, ‘guidance’ and/or ‘instructions to authors’ of 868 journals. This sample represents 40% of the total number of English language journals currently publishing primary research articles involving the use of animals, and of these, nearly half (44.6%) do not mention animal research in their policies at all. Of the journals that do have a policy relating to animal research, 17.5% do nothing more than include the word ‘animal’ at some point within the text, for example in a heading referring to ‘Human and Animal Studies’, without then stating any policies specific or relevant to animal studies. 19% simply suggest that animal research should be conducted according to legal standards, and/or local guidelines, and 12.6% suggest that animal use should have undergone an ethical review. I could go on, but what I hope this illustrates is that the vast majority of journals publishing animal research currently do not have policies that are anywhere near equivalent to those that many journals apply when publishing research involving human subjects. So what role, or even responsibility, do I think scientific journals have in improving the reporting of animal research? First of all, I am assuming that journals wish to ensure that the research they publish is well conducted and scientifically valid. If this is true, then the inclusion of details of study design, animal care and treatment in submitted articles is essential if editors and reviewers are to make a sound judgement on the suitability of manuscripts for publication. Unfortunately, a recent analysis by Kilkenny et al.2 has shown that such information, covering all aspects of study design and practice, is lacking in many research reports. Good study design, based on sound statistical principles, not only reduces the number of animals used to a minimum but is essential if valid conclusions are to be drawn from the results. Similarly, pain, suffering, and distress in experimental animals, caused by poor welfare or the experimental procedures used, can have an adverse effect on the validity of experiments. Optimizing welfare and reducing suffering is therefore an important element of good scientific practice. It is also important that these details are included in the published articles, or as supplementary information, so that readers will be able to judge the validity of the Reporting animal research – worthy of a rethink? 5