Improving design and analysis of research: lessons from clinical research

There is a clear need to optimise the use of animals in research and to ensure that the studies that are done make a worthwhile contribution to scientific knowledge. Research that has an inappropriate study design, or is improperly analysed or interpreted, may mislead and is not defensible. Published reviews of the quality of design and analysis in clinical studies are used to see what lessons can be drawn in respect of animal research. The need for clear and full reporting of research is discussed.     

Publishing the findings of clinical research.

Current regulations and practice may not prevent a sponsor of clinical research from delaying or preventing the dissemination of findings that do not support his or her commercial, professional, or managerial interests. Pharmaceutical trials are particularly subject to this concern, but the issue of non-publication has a wider significance. Patients participate in research on the understanding that it will be of public benefit. To be fair to these patients we must confirm that data will be analysed without bias and that findings will be regarded as being in the public domain and hence available to interested reviewers even if the results do not justify publication. Clinical investigators and ethics committees have the power and the duty to enforce this outcome.