related study idea Erick Turner, Oregon Health & Science University 13 January 2015 Thank you for undertaking this important work. This is not a comment on this protocol. Rather, I would like to suggest, as a possible future study, an alternate method that could be used to study publication bias in animal research: One could use FDA pharmacology reviews, which largely cover animal studies, then ascertain which studies covered in the review have been published. For each study that is published, one could examine outcome reporting bias by comparing the results in the review with those in the corresponding journal publication. This approach may offer some advantages over using published articles as the starting point. Competing interests I previously worked as a medical reviewer at the FDA and believe that the FDA has access to a great deal of data is informative but often does not find its way into published form. I have published papers showing this to be the case with human data, and I would like to advocate for extending this methodology to preclinical data.