Manuscripts based on the same database

If editors receive manuscripts from separate research groups or from the same group analyzing the same data set（for example,from a public database,or systematic reviews or meta-analyses of the same evidence）,the manuscripts should be considered independently because they may differ in their analytic methods,conclusions,or both.If the data interpretation and conclusions are similar,it may be reasonable although not mandatory for editors to give preference to the manuscript submitted first.Editors might consider publishing more than one manuscript that overlap in this way because different analytical approaches may be complementary and equally valid,but manuscripts based upon the same dataset should add substantially to each other to warrant consideration for publication as separate papers,with appropriate citation of previous publications from the same dataset to allow for transparency.

Secondary analyses of clinical trial data should cite any primary publication,clearly state that it contains secondary analyses/results,and use the same identifying trial registration number as the primary trial and unique,persistent dataset identifier.

Sometimes for large trials it is planned from the beginning to produce numerous separate publications regarding separate research questions but using the same original participant sample.In this case authors may use the original single trial registration number,if all the outcome parameters were defined in the original registration.If the authors registered several substudies as separate entries in,for example,clinicaltrials.gov,then the unique trial identifier should be given for the study in question.The main issue is transparency,so no matter what model is used it should be obvious for the reader.

如果編輯收到不同研究組或同一研究組分析同一數(shù)據(jù)集的幾篇稿件（例如，來自同一個公共數(shù)據(jù)庫，或?qū)ο嗤C據(jù)的系統(tǒng)綜述和Meta 分析），每篇稿件均應(yīng)分別考慮，因?yàn)槠浞治龇椒?、結(jié)論，或此兩方面均可能有所不同。如果對數(shù)據(jù)的解釋和結(jié)論都相似，編輯較為合理的做法是優(yōu)先發(fā)表先投稿的稿件，盡管并非必須這樣做。編輯可以考慮發(fā)表這類內(nèi)容有重復(fù)的多篇稿件，因?yàn)椴煌姆治鍪侄慰赡苁腔パa(bǔ)的且同樣是正確的?；谙嗤瑪?shù)據(jù)集的稿件應(yīng)在很大程度上可以互相補(bǔ)充，以使分別發(fā)表這些文章具有合理性，并且應(yīng)恰當(dāng)?shù)厥赜孟劝l(fā)表的基于相同數(shù)據(jù)集的文章以保證透明性。

報告對臨床試驗(yàn)數(shù)據(jù)的再次分析應(yīng)守用初次分析發(fā)表的論文，并清楚地說明，報告中包含再次分析及其結(jié)果，而且要與首次分析的試驗(yàn)使用同一臨床試驗(yàn)注冊號和唯一、永久的數(shù)據(jù)集標(biāo)識碼。

有時候，大型臨床試驗(yàn)一開始就計劃針對不同研究問題發(fā)表多篇論文，但使用的是同樣的原始受試者資料。在這種情況下，如果所有結(jié)局指標(biāo)在最初注冊時已確定，則作者可使用最初的那個臨床試驗(yàn)注冊號。如果作者以幾個獨(dú)立子項目分別注冊過，比如在ClinicalTrials.gov 注冊，則應(yīng)分別使用針對研究問題的那個項目的臨床試驗(yàn)注冊號。關(guān)鍵是要透明，不管采用哪種做法，都應(yīng)使讀者清清楚楚。