Pardonable revisions and protocol reviews

This week sees a few alterations to the ordering of events in The Lancet. Letters to the Editor are now divided into peer-reviewed Research Letters, promoted to join their companion original research towards the front of the journal, and Correspondence. No longer will briefly reported original observations from the clinic or laboratory be camouflaged amongst the vigorous comment on which we thrive. Also, we begin a new weekly series of Seminars that aim to provide extensively referenced reviews on topics of general medical and surgical interest. And we introduce Dissecting Room. Aside from providing a new venue for reviews of books, software, and interesting web sites (no, not an oxymoron), together with poetry and art, these pages offer the diversions of Lifelines and Jabs and Jibes. There is another idea that The Lancet wishes to promote in 1997—namely, protocol reviews. We invite those responsible for conducting randomised trials and metaanalyses across all disciplines to consider submitting their study protocols to us. We will undertake to evaluate these and, if they satisfy our usual criteria of originality, methodological strength, and appropriateness for our readership, we will offer a provisional commitment to publish. We will also post a summary of the protocol on our web site and invite comment from readers. Why are we doing this? For two main reasons. First, by opening negotiations with authors early on, we hope to attract the best research papers and to assist investigators in placing their study in the right journal. If that journal is ours, protocol review should accelerate a paper’s path to publication. Second, we are conscious that journals may harm the process of systematically reviewing the research literature through publication bias. Organisations such as the Cochrane Collaboration, for example, are actively seeking details from researchers in academic and industry settings of all continuing and completed randomised phase III trials. The Lancet can play a useful part in this. Reviewing protocols before the study is completed will help to register ongoing work and to avoid bias against important negative studies.