《cochrane纳入的RCT文献质量评价(风险偏倚评估工具)中英文对照版.pdf》由会员分享，可在线阅读，更多相关《cochrane纳入的RCT文献质量评价(风险偏倚评估工具)中英文对照版.pdf（14页珍藏版）》请在taowenge.com淘文阁网|工程机械CAD图纸|机械工程制图|CAD装配图下载|SolidWorks_CaTia_CAD_UG_PROE_设计图分享下载上搜索。
1、中文：偏倚类型偏倚类型选择偏倚随机序列的产生随机序列的产生判断指标判断指标评价员的判断评价员的判断足够详细的描述用于生成分配序列的方法，以评估产生的分组是否具有可比性。足够详细的描述隐藏分配序列的方法，以决定干预的分配在纳入之前或纳入过程中是否可见生成随机序列不充分，发生选择偏倚分配前分配隐藏不充分发生选择偏倚分配隐藏分配隐藏实施偏倚实施者和参与者实施者和参与者双盲双盲 应对每个主要结局进行评估（或分类结局）测量偏倚结局评估中的盲结局评估中的盲法法每个主要结局如果有，描述对参与者和实施者行参与者和实施者了解干盲法，避免其了解干预信息的所有预的相关信息导致实施措施。提供任何与所实施的盲法是偏倚否
2、有效地相关信息。如果有，描述对结局者行盲法，避结局评估者了解分配的免其了解自己所接受的干预信息干预措施将导致测量偏均应评估（或分类的所有措施。提供任何与所实施的倚结局）盲法是否有效地相关信息。失访偏倚不全结局数据不全结局数据每个主要结局均应评估（或分类结局）描述每个主要结局数据的完整性，不全结局数据的数量，包括分析中的自然缺失和排除。这性质，处理方式导致失些缺失数据是否报告，在各个干预访偏倚组的数目（并与总样本量比较），数据缺失以及重新纳入分析的原因发表偏倚SelectiveSelectivereporting.reporting.其它偏倚其它偏倚来源其它偏倚来源说明如何审查选择性报道结局的可
3、能性，以及审查结果选择性报道结局导致发表偏倚说明不包括在上述偏倚中的其它重要偏倚如果特定的问题或条目事先在计划书中指出，应对每一项说明不包括在上述各项中的偏倚随机序列的产生随机序列的产生随机序列产生不充分导致选择偏倚随机序列产生不充分导致选择偏倚判断为低风险的标准研究者描述随机序列产生过程譬如:参考随机数字表使用计算机随机数字生成器扔硬币洗牌的卡片和信封掷骰子抽签最小化*最小化，可实现无随机元素，被认为相当于是随机的。判断为高风险的标准研究者描述序列的产生使用的是非随机的方法。通常是系统的非随机方法，例如：通过奇偶或出生日期产生序列通过入院日期产生序列通过类似住院号或门诊号产生序列相对于上面提
4、到的系统方法，其它非随机的方法少见的多，也更明显。通常包括对参与者进行判断或非随机的方法，例如:临床医生判断如何分配参与者判断如何分配基于实验室检查或系列测试的结果分配基于干预的可获取性进行分配偏倚风险不清楚的判断标准没有足够的信息判断随机序列的产生存在高风险或低风险分配隐藏分配隐藏分配前不充足的分配隐藏导致选择偏倚低风险判断标准参与者以及纳入参与者的研究者因以下掩盖分配的方法或相当的方法，事先不了解分配情况中心分配（包括电话，网络，药房控制随机）相同外形的顺序编号的药物容器；顺序编号、不透明、密封的信封高风险判断标准参与者以及纳入参与者的研究者可能事先知道分配，因而引入选择偏倚，譬如基于如下
5、方法的分配:使用摊开的随机分配表（如随机序列清单）分发信封但没有合适的安全保障（如透明、非密封、非顺序编号）交替或循环出生日期病历号其它明确的非隐藏过程风险未知没有足够信息判断为低风险或高风险。通常因分配隐藏的方法未描述或描述不充分。例如描述为使用信封分配，但为描述信封是否透明密封顺序编号对参与者和实施者的盲法对参与者和实施者的盲法因参与者和实施者了解干预情况而导致实施偏倚偏倚低风险标准任何如下标准:无盲法或盲法不充分，但系统评价员判断结局不太可能受到缺乏盲法的影响参与者和主要实施者均实施可靠的盲法，且盲法不太可能被打破偏倚高风险标准任何如下标准:无盲法或盲法不充分，但系统评价员判断结局很可能
6、受到缺乏盲法的影响尝试对关键的参与者和实施者行盲法，但盲法很可能被打破，结局很可能受到缺乏盲法的影响风险未知任何如下标准:没有足够信息判断为低风险或高风险研究未描述此情况对结局评价实施盲法对结局评价实施盲法结局评价者了解干预分配信息将导致测量偏倚偏倚低风险标准任何如下标准:无盲法或盲法不充分，但系统评价员判断结局不太可能受到缺乏盲法的影响参与者和主要实施者均实施可靠的盲法，且盲法不太可能被打破高风险判断标准任何如下标准:无盲法或盲法不充分，但系统评价员判断结局很可能受到缺乏盲法的影响尝试对关键的参与者和实施者行盲法，但盲法很可能被打破，结局很可能受到缺乏盲法的影响风险未知任何如下标准:没有足够
7、信息判断为低风险或高风险研究未描述此情况结局数据不完整不全结局数据的数量，性质，处理方式导致失访偏倚偏倚低风险标准任何如下标准:无缺失数据缺失数据的产生不大可能与真实结局相关（对于生存数据，删失不大可能引入偏倚）缺失数据的数目在各干预组相当，且各组缺失原因类似对二分类变量，与观察事件的发生风险相比，缺失比例不足以影响预估的干预效应对连续性结局数据，缺失数据的合理效应规模（均数差或标准均数差）不会大到影响观察的效应规模;缺失的数据用合适的方法进行估算高风险判断标准任何如下标准:缺失数据的产生很大可能与真实结局相关,缺失数据的数目及缺失原因在各干预组相差较大对二分类变量，与观察事件的发生风险相比，
8、缺失比例足以影响预估的干预效应对连续性结局数据，缺失数据的合理效应规模（均数差或标准均数差）足以影响观察的效应规模;意向治疗分析中存在实际干预措施与随机分配的干预相违背的情况对缺失数据进行简单的不合适的估算风险未知任何如下标准:没有报道缺失或排除的情况，无法判断高风险或低风险（如未说明随机的数量，未提供数据缺失的原因）研究未描述此情况选择性发表选择性发表选择性发表导致发表偏倚偏倚低风险标准任何如下标准:实验的计划书可获取，系统评价感兴趣的所有首要或次要结局均按计划书预先说明的方式报道实验计划书不可得，但很明显发表的报告包括所有的结局，包括预先说明的结局（这种性质的有说服力的文字可能少见）高风险
9、判断标准任何如下标准:不是所有的预先说明的首要结局均被报道一个或多个首要结局为采用预先说明的测量方法、分析方法或数据子集来报道系统评价感兴趣的一个或多个首要结局报道不全，以至于不能纳入 meta 分析研究未报道此研究应当包含的主要关键结局风险未知没有足够信息判断高风险或低风险，貌似大部分研究会被分为此类OTHER BIASOTHER BIAS不包括在以上五种的其它偏倚偏倚低风险标准研究应未引入其它来源的偏倚高风险判断标准至少有一种重要的偏倚风险，例如：具有与特殊试验设计相关的潜在偏倚来源或被指欺诈或其它问题风险未知可能存在偏倚风险，但存在以下两种中的一种没有足够信息评估是否存在其它重要的偏倚风
10、险没有足够的证据认为发现的问题会引入偏倚Possible approach forsummary assessmentsof the risk of bias for eachimportant outcome(across domains)within and across studiesRisk of biasRisk of biasLow risk ofbias.解释解释合理的偏倚不太可能严重改变结果Unclear risk of 合理的偏倚会对bias.结果产生一定的怀疑High risk of偏倚严重削弱结bias.果的可信度对单个研究对单个研究每一类偏倚均为低风险一类或多类偏倚风险
11、未知一类或多类偏倚为高风险对多个研究整体对多个研究整体绝大多数信息均来自偏倚低风险的研究绝大多数信息均来自偏倚低风险或风险未知的研究来自高偏倚风险研究的信息比例足以影响结果的解释英文：TableTableDomainDomainSelection bias.Random sequenceRandom sequencegeneration.generation.Support for judgementSupport for judgementReview authors judgementReview authors judgementthe allocation sequence in su
12、fficientit should produce comparable groups.allocation to interventions)of a randomised sequence.Selection bias(biasedallocation to interventions)Describe the method used to generateSelection bias(biaseddetail to allow an assessment of whetherdue to inadequate generationAllocationAllocationconcealme
13、nt.concealment.Describe the method used to concealthe allocation sequence in sufficientdetail to determine whether intervention due to inadequateallocations could have been foreseen in concealment of allocationsadvance of,or during,enrolment.Performance bias.Blinding ofBlinding ofparticipants andpar
14、ticipants andpersonnelpersonnelbe made for eachmain outcome(orclass of outcomes).Detection bias.prior to assignment.Describe all measures used,if any,toPerformance bias due toblind study participants and personnelknowledge of the allocatedfrom knowledge of which intervention ainterventions by partic
15、ipantsand personnel during thestudy.information relating to whether theintended blinding was effective.Assessments shouldparticipant received.Provide anyDetection bias due toBlinding of outcomeBlinding of outcome Describe all measures used,if any,toassessmentassessmentbe made for eachmain outcome(or
16、class of outcomes).Attrition bias.blind outcome assessors fromparticipant received.Provide anyinformation relating to whether theintended blinding was effective.knowledge of the allocatedinterventions by outcomeassessors.Assessments shouldknowledge of which intervention aIncomplete outcomeIncomplete
17、 outcome Describe the completeness of outcomeAttrition bias due to amount,datadata Assessmentsshould be made foreach main outcomedata for each main outcome,includingnature or handling ofattrition and exclusions from theanalysis.State whether attrition andin each intervention group(comparedwith total
18、 randomized participants),reasons for attrition/exclusions wherereported,and any re-inclusions inanalyses performed by the reviewauthors.Reporting bias.incomplete outcome data.(or class of outcomes).exclusions were reported,the numbersReporting bias due toSelective reporting.Selective reporting.Stat
19、e how the possibility of selectivereview authors,and what was found.Other bias.Other sources ofOther sources ofbias.bias.outcome reporting was examined by the selective outcome reporting.not addressed in the other domains inthe tool.If particular questions/entries werepre-specified in the reviews pr
20、otocol,responses should be provided for eachquestion/entry.covered elsewhere in thetable.State any important concerns about biasBias due to problems notRANDOM SEQUENCE GENERATIONRANDOM SEQUENCE GENERATIONSelection bias(biased allocation to interventions)due to inadequate generation of aSelection bia
21、s(biased allocation to interventions)due to inadequate generation of arandomised sequence.randomised sequence.Criteria for a judgementThe investigators describe a random component in the sequenceof Low risk of bias.generation process such as:Referring to a random number table;Using a computer random
22、 number generator;Coin tossing;Shuffling cards or envelopes;Throwing dice;Drawing of lots;Minimization*.*Minimization may be implemented without a random element,andthis is considered to be equivalent to being random.Criteria for theof bias.The investigators describe a non-random component in the se
23、quencesystematic,non-random approach,for example:Sequence generated by odd or even date of birth;Sequence generated by some rule based on date(or day)ofadmission;Sequence generated by some rule based on hospital or clinicrecord number.Other non-random approaches happen much less frequently than thes
24、ystematic approaches mentioned above and tend to be obvious.Theyusually involve judgement or some method of non-randomcategorization of participants,for example:Allocation by judgement of the clinician;Allocation by preference of the participant;Allocation based on the results of a laboratory test o
25、r a seriesof tests;Criteria for therisk of bias.Allocation by availability of the intervention.judgement of High risk generation process.Usually,the description would involve someInsufficient information about the sequence generation process tojudgement of Unclearpermit judgement of Low risk or High
26、 risk.ALLOCATION CONCEALMENTALLOCATION CONCEALMENTSelection bias(biased allocation to interventions)due to inadequate concealment ofSelection bias(biased allocation to interventions)due to inadequate concealment ofallocations prior to assignment.allocations prior to assignment.Criteria for a judgeme
27、ntParticipants and investigators enrolling participants could not foreseeof Low risk of bias.assignment because one of the following,or an equivalent method,was used to conceal allocation:Central allocation(including telephone,web-based andpharmacy-controlled randomization);Sequentially numbered dru
28、g containers of identicalappearance;Criteria for theof bias.Sequentially numbered,opaque,sealed envelopes.Participants or investigators enrolling participants could possiblyallocation based on:Using an open random allocation schedule.a list of randomnumbers);Assignment envelopes were used without ap
29、propriatesafeguards.if envelopes were unsealed or nonopaque or notsequentially numbered);Alternation or rotation;Date of birth;Case record number;Any other explicitly unconcealed procedure.judgement of High risk foresee assignments and thus introduce selection bias,such asCriteria for therisk of bia
30、s.Insufficient information to permit judgement of Low risk or High risk.or not described in sufficient detail to allow a definite judgement forexample if the use of assignment envelopes is described,but itremains unclear whether envelopes were sequentially numbered,opaque and sealed.judgement of Unc
31、learThis is usually the case if the method of concealment is not describedBLINDING OF PARTICIPANTS AND PERSONNELBLINDING OF PARTICIPANTS AND PERSONNELPerformance bias due to knowledge of the allocated interventions by participants andPerformance bias due to knowledge of the allocated interventions b
32、y participants andpersonnel during the study.personnel during the study.Criteria for a judgementAny one of the following:of Low risk of bias.No blinding or incomplete blinding,but the review authorsjudge that the outcome is not likely to be influenced by lack ofblinding;Blinding of participants and
33、key study personnel ensured,andunlikely that the blinding could have been broken.Criteria for theof bias.Any one of the following:No blinding or incomplete blinding,and the outcome is likelyto be influenced by lack of blinding;Blinding of key study participants and personnel attempted,but likely tha
34、t the blinding could have been broken,and theoutcome is likely to be influenced by lack of blinding.Criteria for thejudgement of Unclearrisk of bias.Any one of the following:Insufficient information to permit judgement of Low risk orHigh risk;judgement of High riskThe study did not address this outc
35、ome.BLINDING OF OUTCOME ASSESSMENTBLINDING OF OUTCOME ASSESSMENTDetection bias due to knowledge of the allocated interventions by outcome assessors.Detection bias due to knowledge of the allocated interventions by outcome assessors.Criteria for a judgementAny one of the following:of Low risk of bias
36、.No blinding of outcome assessment,but the review authorsjudge that the outcome measurement is not likely to beinfluenced by lack of blinding;Blinding of outcome assessment ensured,and unlikely thatthe blinding could have been broken.Criteria for theof bias.Any one of the following:No blinding of ou
37、tcome assessment,and the outcomemeasurement is likely to be influenced by lack of blinding;Blinding of outcome assessment,but likely that the blindingcould have been broken,and the outcome measurement islikely to be influenced by lack of blinding.Criteria for thejudgement of Unclearrisk of bias.Any
38、one of the following:Insufficient information to permit judgement of Low risk orHigh risk;The study did not address this outcome.judgement of High riskINCOMPLETE OUTCOME DATAINCOMPLETE OUTCOME DATAAttrition bias due to amount,nature or handling of incomplete outcome data.Attrition bias due to amount
39、,nature or handling of incomplete outcome data.Criteria for a judgementAny one of the following:of Low risk of bias.No missing outcome data;Reasons for missing outcome data unlikely to be related totrue outcome(for survival data,censoring unlikely to beintroducing bias);Missing outcome data balanced
40、 in numbers acrossintervention groups,with similar reasons for missing dataacross groups;For dichotomous outcome data,the proportion of missingoutcomes compared with observed event risk not enough tohave a clinically relevant impact on the intervention effectestimate;For continuous outcome data,plau
41、sible effect size(differencein means or standardized difference in means)amongmissing outcomes not enough to have a clinically relevantimpact on observed effect size;Criteria for theof bias.Missing data have been imputed using appropriate methods.Any one of the following:Reason for missing outcome d
42、ata likely to be related to trueoutcome,with either imbalance in numbers or reasons formissing data across intervention groups;For dichotomous outcome data,the proportion of missingoutcomes compared with observed event risk enough toinduce clinically relevant bias in intervention effect estimate;For
43、 continuous outcome data,plausible effect size(differencein means or standardized difference in means)amongmissing outcomes enough to induce clinically relevant bias inobserved effect size;As-treated analysis done with substantial departure of theintervention received from that assigned at randomiza
44、tion;Potentially inappropriate application of simple imputation.judgement of High riskCriteria for thejudgement of Unclearrisk of bias.Any one of the following:Insufficient reporting of attrition/exclusions to permitjudgement of Low risk or High risk.number randomized notstated,no reasons for missin
45、g data provided);The study did not address this outcome.SELECTIVE REPORTINGSELECTIVE REPORTINGReporting bias due to selective outcome reporting.Reporting bias due to selective outcome reporting.Criteria for a judgementAny of the following:of Low risk of bias.The study protocol is available and all o
46、f the studyspre-specified(primary and secondary)outcomes that are ofinterest in the review have been reported in the pre-specifiedway;The study protocol is not available but it is clear that thepublished reports include all expected outcomes,includingthose that were pre-specified(convincing text of
47、this naturemay be uncommon).Criteria for theof bias.Any one of the following:Not all of the studys pre-specified primary outcomes havebeen reported;One or more primary outcomes is reported usingmeasurements,analysis methods or subsets of the data.judgement of High risksubscales)that were not pre-spe
48、cified;One or more reported primary outcomes were notpre-specified(unless clear justification for their reporting isprovided,such as an unexpected adverse effect);One or more outcomes of interest in the review are reportedincompletely so that they cannot be entered in ameta-analysis;The study report
49、 fails to include results for a key outcome thatwould be expected to have been reported for such a study.Criteria for therisk of bias.Insufficient information to permit judgement of Low risk or High risk.judgement of UnclearIt is likely that the majority of studies will fall into this category.OTHER
50、 BIASOTHER BIASBias due to problems not covered elsewhere in the table.Bias due to problems not covered elsewhere in the table.Criteria for a judgementThe study appears to be free of other sources of bias.of Low risk of bias.Criteria for theof bias.There is at least one important risk of bias.For ex