In Reply: Drs Johnson and Lassere emphasize the importance of prespecified analyses in the reporting and interpretation of randomized clinical trials. The protocol for a trial serves as the centerpiece, and its analysis plan needs to specify not only the type and method of analysis but the efficacy and safety outcomes as well. Information about amendments, their timing, and the rationales should become part of the protocol. Ideally, this full protocol should be available to reviewers and editors when the manuscript for the corresponding trial is submitted to a journal. As Johnson and Lassere point out, the FDA Amendments Act requirement to post “the full protocol or such information on the protocol for the trial as may be necessary to help to evaluate the results of the trial”1 may prove especially useful.
Psaty BM, Kronmal RA. Guest Authorship, Mortality Reporting, and Integrity in Rofecoxib Studies—Reply. JAMA. 2008;300(8):900–906. doi:10.1001/jama.300.8.904-b