关于研究文献透明化的声明——一种克服不充分研究报告的方法

Douglas G.Altman，David Moher

Douglas G.Altman，David Moher.关于研究文献透明化的声明——一种克服不充分研究报告的方法[J].中国全科医学，2014，17(2)：119-122.[www.chinagp.net]

Douglas G.Altman(道格拉斯G.沃特曼)：英国牛津大学生物统计学教授、生物统计中心主任；国际著名方法学家，Cochrane协作网方法学组负责人，曾任内科学年鉴主编，现任国际方法学专业杂志Trials主编；随机对照试验报告指南(CONSORT)工作组负责人之一；国际促进卫生研究质量与透明化协作网(EQUATOR)主要负责人;国际期刊编辑伦理委员会(COPE)成员。

David Moher(大卫.墨赫尔)：加拿大渥太华大学公共卫生流行病学教授，临床试验中心主任，世界卫生组织国际临床试验注册平台专家组成员。国际著名临床研究方法学家，CONSORT工作组负责人之一，Cochrane协作网替代医学组首席方法学家，EQUATOR特邀专家，COPE成员。

“确保所发表的报告无偏倚、准确地重现研究是每一位参与者的责任”[1]。

研究记录常常会做简化处理，但却存在对患者不利的风险。因此，医学研究界需要变革来保证读者获得所有研究的真相，特别是随机对照试验报告，具有解答什么是对患者最有利治疗的作用。

如果不发表所有研究特别是随机对照试验的结果，将严重地误导循证临床决策。最近一项关于瑞波西汀治疗抑郁的系统评价发现约有四分之三参与研究的患者未在发表试验中[2]。904项治疗缺血性卒中(1955—2008年)的已完成试验中，有五分之一没有被适当地发表，“其中有的大到足以影响临床实践以及系统评价和Meta分析的结果”[3]。

与不发表一样恶劣的是，不完整或误导发表导致的问题更大。发表在同行评审刊物的临床试验结果可能与之前呈报到管理机构的不同[4-6]，阳性结果可能更多。主要测量指标常与研究者在研究计划[7-8]或注册机构[9-10]中所描述的不一样。选择性地不发表不利的有统计学意义的结果，使文献产生偏倚[11-12]。更有甚者，作者常歪曲表述和解释他们的研究结果。一项研究发现，72个主要测量指标差异无统计学意义的随机对照试验都存在同样问题[13]。在诊断准确性试验研究中也有相似的发现[14]。

同行专家评审不能保证杂志发表的文章包含有读者需要的详细的临床和方法学的关键内容。对已发表的随机对照试验报告的评估发现，其常缺乏对干预措施的评估[15-16]、患者纳入标准[17]和结果[18-19]的详细信息；研究方法的详细信息也不充分，特别是关于分组的信息。一项2006年的研究发现，仅三分之一的研究报告描述了随机序列的产生方法、仅四分之一充分描述了隐蔽分组方法[20]。一项包含357个Ⅲ期肿瘤临床试验的综述总结道：“大量的内容在许多试验中都没有报告”[21]。不利事件也常常报告不充分[22-23]。

其他类型研究在结果和报告中的问题可能较随机对照试验更严重。虽然集中研究较少，相似观点已经在有关的流行病学[24-25]、药物流行病学[26]、诊断性研究[27]、预后研究[28]和临床前研究[29-30]中表达过。当然，高质量报告并不等于有同样高质量的研究。但是，完整和清楚的报告可让读者判断研究的可靠性，获取所需的相关信息。有人认为商业资助的研究可能更多地不被发表[2，31]，一旦发表这些试验报告则更加完整[32]。

那么，什么是必须的？报告应该对研究者如何实施其研究及研究结果进行清楚和透明地描述。忽略方法学或研究实施的重要细节应被认为是不可接受的，期刊不应发表这样的报告。虽然找到一些缺失信息需要额外的途径(例如从试验注册机构或研究计划获得信息)，多数缺失信息在稿件中都很明显并能检测出来，尽管报告指南如CONSORT[33]容易获得，但这种状况的改善仍然缓慢[34]。

由于研究者对信息的刻意隐瞒使试验结果不易获得，这与赫尔辛基宣言相违背，不仅不符合当前研究的道义和科学性，而且浪费了大量稀缺资源、降低了研究的社会价值[35]。由于误导或不充分地发表研究结果使研究者和资助组织失去了公众的信任。从科学的角度，这不利于系统评价研究者汇集所有的证据，与使用所有信息的评估比较，只评估部分内容会对有效性和安全性的评估产生偏倚，降低准确性，不利于找到最利于患者的措施。

有一个几乎立刻就能实施的计划：我们建议每本杂志在作者投稿时都增加一个所有作者签署的发表透明化声明，在为其他目的提交报告时，如对监管机构，也需做类似的声明(见下文示例)。

透明化声明：负责作者*证明，本文对研究所做的报告是诚实、准确和透明的；没有遗漏重要的内容，对与研究计划(如果有，与所注册的信息)的任何不一致均做了解释。

注：*指文稿的保证人

编辑和编辑团队可以通过更新其“作者须知”来支持本倡议，将完整的发表透明化保证作为投稿程序的一部分。我们认为这与当前广泛实施的要求作者做“利益冲突声明”相似，具有科学上的必要性。采用这样的措施之后还有隐瞒或发表不正确信息的报告，将作为制裁学术不端行为的证据。我们希望此措施将鼓励作者更加仔细地撰写论文，并鼓励他们按照相关报告指南审核自己的文稿。BMJ，我们中的一人(Douglas G.Altman)是其高级统计学编辑，和BMJ开放版(BMJOpen)正在带头实施这项政策。我们邀请其他杂志也实行这项政策，并支持EQUATOR网站(www.equator-network.org)的透明化声明。

科学界和公众对准确和完整的研究报告寄予厚望，我们需要变革来给以保障。广泛认可和实行的发表透明化声明是帮助实现医学研究最大价值的一条途径，然而其影响不了未发表的研究，这仍然是一件耻辱。

DeclarationofTransparencyforEachResearchSrticle

"Itistheresponsibilityofeveryoneinvolvedtoensurethatthepublishedrecordisanunbiased,accuraterepresentationofresearch."1

The research record is often manipulated for short term gain but at the risk of harm to patients.The medical research community needs to implement changes to ensure that readers obtain the truth about all research,especially reports of randomised trials,which hold a special place in answering what works best for patients.

Failure to publish the findings of all studies,especially randomised trials,seriously distorts the evidence base for clinical decision making.A recent systematic review of reboxetine for treating depression found that almost three quarters of included patients were in unpublished trials2.Of 904 completed trials of interventions for acute ischaemic stroke (1955—2008),a fifth were not properly published,"several of which may be large enough to influence clinical practice and the findings of systematic reviews and meta-analyses"3.

Bad as non-publication is,incomplete or misleading publications cause greater problems.Results of clinical trials published in peer reviewed publications may differ from what was previously submitted to regulatory agencies4-6,with the published data being more positive.The primary outcome often differs from what the researchers had stated in the trial protocol7-8or clinical trial registry9-10.Selective non-publication favours statistically significant findings,biasing the literature11-12.Furthermore,authors often distort the presentation and interpretation of their findings.One study found that such "spin" was common in 72 reports of randomised controlled trials with statistically non-significant primary outcomes13.Similar findings have been reported recently for studies of the accuracy of diagnostic tests14.Peer review is failing to ensure that journal articles contain the key clinical and methodological details that readers need.Reviews of published reports of randomised trials have found common deficiencies in the details of the interventions being evaluated15-16,participant eligibility criteria17,and outcomes18-19.Details of study methods are also often inadequate,especially in relation to allocation.A 2006 study found that only a third of trial reports described how the randomisation sequence was generated and only a quarter described an adequate method of allocation concealment20.A review of 357 phase Ⅲ oncology trials concluded that "numerous items remained unreported for many trials"21.Harms too are poorly reported22-23.

The problems associated with publishing and reporting other types of research may be worse than for randomised trials.Although less intensively studied,similar concerns have been expressed in relation to epidemiology24-25,pharmacoepidemiology26,diagnosis research27,prognosis research28,and preclinical research29-30.Of course,good reporting is not the same as high quality research.But a full and clear report allows readers to judge a study′s reliability and relevance.There are concerns that commercially sponsored research may be more likely to remain unpublished2,31,but when published these trials are reported more fully32.

So what is needed? Published research articles should provide a clear and transparent description of how researchers conducted their study and what they found.Omission of important details of methods or study conduct should be deemed unacceptable,and journals should not publish them.Although detection of some deficiencies requires external information (for example,from a trials register or protocol),most deficiencies are inherent in a submitted manuscript and should be detected.Despite the availability of reporting guidelines such as CONSORT33,improvements are slow to materialise34.

By not making results of their research easily accessible,researchers are withholding knowledge,in contravention of the Declaration of Helsinki.Not only are current practices questionable on moral and scientific grounds,failure to publish all research findings is a massive waste of scarce resources and diminishes the social value of the research35.Researchers and funding organisations also fail the public when research findings are published in a misleading or inadequate way.Scientifically,this harms systematic reviewers who want to aggregate all of the evidence.Reviewing a partial picture provides biased and less precise estimates of effectiveness and safety than when the full information is used,and it may compromise the identification of what works best for patients.

We have a proposal that can be acted on almost immediately.We suggest that authors should sign a publication transparency declaration (box) as part of every journal submission.The same declaration could be appropriate for submissions in other contexts-for example,to regulatory agencies.

Transparencydeclaration：Theleadauthor*affirmsthatthismanuscriptisanhonest,accurate,andtransparentaccountofthestudybeingreported;thatnoimportantaspectsofthestudyhavebeenomitted;andthatanydiscrepanciesfromthestudyasplanned(and,ifrelevant,registered)havebeenexplained.

*Themanuscript′sguarantor.

Editors and editorial groups can support this initiative by updating their instructions to authors so that a completed publication transparency pledge is required as part of the submission process.We see this action as a necessary scientific analogue of the current widespread practice of asking authors about conflicts of interest.Subsequent revelation of withheld or incorrect information would be evidence of scientific misconduct for which various actions could be taken.We hope that this step will encourage authors to reflect more carefully on how they write their article and encourage them to check that they have adhered to relevant reporting guidelines.TheBMJ,for which one of us (DGA) is the senior statistics editor,andBMJOpenare leading the way by implementing this policy immediately.We invite other journals to do likewise and support the transparency declaration on the EQUATOR website (www.equator-network.org).

The scientific community and the public at large deserve an accurate and complete record of research;we need to make changes to ensure that we will get one.Widespread endorsement and implementation of a publication transparency declaration is one way to help to get the maximum value from medical research.It will,however,have no influence on the non-publication of studies,which is a continuing disgrace.

1 PLoS Medicine Editors.An unbiased scientific record should be everyone′s agenda[J].PLoS Med,2009,24(6):e1000038.

2 Eyding D,Lelgemann M,Grouven U,et al.Reboxetine for acute treatment of major depression:systematic review and meta-analysis of published and unpublished placebo and selective serotonin reuptake inhibitor controlled trials[J].BMJ,2010,341:c4737.

3 Gibson LM,Brazzelli M,Thomas BM,et al.A systematic review of clinical trials of pharmacological interventions for acute ischaemic stroke (1955—2008) that were completed,but not published in full[J].Trials,2010,11:43.

4 Turner EH,Knoepflmacher D,Shapley L.Publication bias in antipsychotic trials:an analysis of efficacy comparing the published literature to the US Food and Drug Administration database[J].PLoS Med,2012,9(3):e1001189.

5 Turner EH,Matthews AM,Linardatos E,et al.Selective publication of antidepressant trials and its influence on apparent efficacy[J].N Engl J Med,2008,358(3):252-260.

6 Vitry AI.Reporting of studies on new medicines in major medical journals:a case study in breast cancer[J].Clin Pharmacol Ther,2010,87(4):398-400.

7 Dwan K,Altman DG,Cresswell L,et al.Comparison of protocols and registry entries to published reports for randomised controlled trials[J].Cochrane Database Syst Rev,2011(1):MR000031.

8 Nankervis H,Baibergenova A,Williams HC,et al.Prospective registration and outcome-reporting bias in randomized controlled trials of eczema treatments:a systematic review[J].J Invest Dermatol,2012,132(12):2727-2734.

9 Mathieu S,Boutron I,Moher D,et al.Comparison of registered and published primary outcomes in randomized controlled trials[J].JAMA,2009,302(9):977-984.

10 Rosenthal R,Dwan K.Comparison of randomized controlled trial registry entries and content of reports in surgery journals[J].Ann Surg,2013,257(6):1007-1015.

11 Song F,Parekh S,Hooper L,et al.Dissemination and publication of research findings:an updated review of related biases[J].Health Technol Assess,2010,14(8):1-193.

12 Dwan K,Kirkham JJ,Williamson PR,et al.Selective reporting of outcomes in randomised controlled trials in systematic reviews of cystic fibrosis[J].BMJ Open,2013,3(6):e002709.

13 Boutron I,Dutton S,Ravaud P,et al.Reporting and interpretation of randomized controlled trials with statistically nonsignificant results for primary outcomes[J].JAMA,2010,303(20):2058-2064.

14 Ochodo EA,de Haan MC,Reitsma JB,et al.Overinterpretation and misreporting of diagnostic accuracy studies:evidence of "spin"[J].Radiology,2013,267(2):581-588.

15 Duff JM,Leather H,Walden EO,et al.Adequacy of published oncology randomized controlled trials to provide therapeutic details needed for clinical application[J].J Natl Cancer Inst,2010,102(10):702-705.

16 Glasziou P,Meats E,Heneghan C,et al.What is missing from descriptions of treatment in trials and reviews?[J].BMJ,2008,336(7659):1472-1474.

17 Shapiro SH,Weijer C,Freedman B.Reporting the study populations of clinical trials.Clear transmission or static on the line?[J].J Clin Epidemiol,2000,53(10):973-979.

18 Vera-Badillo FE,Shapiro R,Ocana A,et al.Bias in reporting of end points of efficacy and toxicity in randomized,clinical trials for women with breast cancer[J].Ann Oncol,2013,24(5):1238-1244.

19 Le Cleach L,Chassany O,Levy A,et al.Poor reporting of quality of life outcomes in dermatology randomized controlled clinical trials[J].Dermatology,2008,216(1):46-55.

20 Hopewell S,Dutton S,Yu LM,et al.The quality of reports of randomised trials in 2000 and 2006:Comparative study of articles indexed in PubMed[J].BMJ,2010,340:c723.

21 Peron J,Pond GR,Gan HK,et al.Quality of reporting of modern randomized controlled trials in medical oncology:A systematic review[J].J Natl Cancer Inst,2012,104(13):982-989.

22 Bagul NB,Kirkham JJ.The reporting of harms in randomized controlled trials of hypertension using the CONSORT criteria for harm reporting[J].ClinExp Hypertens,2012,34(8):548-554.

23 Shukralla AA,Tudur-Smith C,Powell GA,et al.Reporting of adverse events in randomised controlled trials of antiepileptic drugs using the CONSORT criteria for reporting harms[J].Epilepsy Res,2011,97(1/2):20-29.

24 Knol MJ,Egger M,Scott P,et al.When one depends on the other:Reporting of interaction in case-control and cohort studies[J].Epidemiology,2009,20(2):161-166.

25 Lee W,Bindman J,Ford T,et al.Bias in psychiatric case-control studies:literature survey[J].Br J Psychiatry,2007,190:204-209.

26 Holmes MV,Shah T,Vickery C,et al.Fulfilling the promise of personalized medicine? Systematic review and field synopsis of pharmacogenetic studies[J].PLoS One,2009,4(12):e7960.

27 Zintzaras E,Papathanasiou A,Ziogas D,et al.The reporting quality of studies investigating the diagnostic accuracy of anti-CCP antibody in rheumatoid arthritis and its impact on diagnostic estimates[J].BMC Musculoskelet Disord,2012,13:113.

28 Kyzas PA,Denaxa-Kyza D,Ioannidis JP.Quality of reporting of cancer prognostic marker studies:association with reported prognostic effect[J].J Natl Cancer Inst,2007,99(3):236-243.

29 Kilkenny C,Parsons N,Kadyszewski E,et al.Survey of the quality of experimental design,statistical analysis and reporting of research using animals[J].PLoS One,2009,4(11):e7824.

30 Dirnagl U,Macleod MR.Stroke research at a road block:the streets from adversity should be paved with meta-analysis and good laboratory practice[J].Br J Pharmacol,2009,157(7):1154-1156.

31 Lundh A,Sismondo S,Lexchin J,et al.Industry sponsorship and research outcome[J].Cochrane Database Syst Rev,2012,12:MR000033.

32 Khan NA,Lombeida JI,Singh M,et al.Association of industry funding with the outcome and quality of randomized controlled trials of drug therapy for rheumatoid arthritis[J].Arthritis Rheum,2012,64(7):2059-2067.

33 Schulz KF,Altman DG,Moher D.CONSORT 2010 Statement:updated guidelines for reporting parallel group randomised trials[J].BMJ,2010,340:c332.

34 Turner L,Shamseer L,Altman DG,et al.Consolidated standards of reporting trials (CONSORT) and the completeness of reporting of randomised controlled trials(RCTs) published in medical journals[J].Cochrane Database Syst Rev,2012,11:MR000030.

35 Chalmers I,Glasziou P.Avoidable waste in the production and reporting of research evidence[J].Lancet,2009,374(9683):86-89.