1、RCT文獻(xiàn)質(zhì)量評(píng)價(jià)嚴(yán)格文獻(xiàn)評(píng)價(jià) 證據(jù)的質(zhì)量表表 循循征征醫(yī)醫(yī)學(xué)學(xué)證證據(jù)據(jù)分分級(jí)級(jí)水水平平及及依依據(jù)據(jù) 推推薦薦分分級(jí)級(jí) 治治療療、預(yù)預(yù)防防、病病因因的的證證據(jù)據(jù) A A 高高質(zhì)質(zhì)量量 R RC CT Ts s 的的系系統(tǒng)統(tǒng)綜綜述述 B B 單單個(gè)個(gè)大大樣樣本本 R RC CT T C C 未未使使用用 R RC CT T 但但設(shè)設(shè)計(jì)計(jì)很很好好的的隊(duì)隊(duì)列列研研究究、病病例例- -對(duì)對(duì)照照研研究究或或無(wú)無(wú)對(duì)對(duì)照照的的系系列列病病例例觀觀察察 D D 專(zhuān)專(zhuān)家家個(gè)個(gè)人人意意見(jiàn)見(jiàn)、個(gè)個(gè)例例報(bào)報(bào)告告 Meta-分析的基本步驟提出問(wèn)題，制定研究計(jì)劃 檢索資料選擇符合納入標(biāo)準(zhǔn)的研究納入研究的質(zhì)量評(píng)價(jià)提取納入文

2、獻(xiàn)的數(shù)據(jù)信息資料的統(tǒng)計(jì)學(xué)處理敏感性分析形成結(jié)果報(bào)告Tools for assessing quality and risk of biasScales（評(píng)分）, in which various components of quality are scored and combined to give a summary score; Checklists（清單、分類(lèi)）, in which specific questions are asked Cochrane handbook5_0_1第八章: Table 8.5.c: Criteria for judging risk of bias

3、in the Risk of bias assessment toolAssessment of risk of bias Sequence generation（隨機(jī)分配方案產(chǎn)生 ） Allocation concealment（分配方案隱藏） Blinding （盲法） Incomplete outcome data （結(jié)果數(shù)據(jù)不完整） Selective outcome reporting（選擇性報(bào)告結(jié)果 ） Other bias（其它影響真實(shí)性因素：無(wú)其它偏倚）For each entry, an answer Yes indicates a low risk of bias, and

4、 an answer No indicates high risk of bias, and Unclear indicates unclear or unknown risk of bias.(1) Was the allocation sequence adequately generated? Low risk of bias: a random component in the sequence generation process was reported. High risk of bias: a non-random component in the sequence gener

5、ation process was reported. Uncertain risk of bias: insufficient information about the sequence generation process to permit judgment of Yes or No.（1）隨機(jī)分配方案產(chǎn)生 正確：采用隨機(jī)數(shù)字表、計(jì)算機(jī)產(chǎn)生隨機(jī)數(shù)字、拋硬幣、擲骰子或抽簽等方法 不正確： 按患者生日、住院日或住院號(hào)等的末尾數(shù)字的奇數(shù)或偶數(shù) 交替分配方法； 或者根據(jù)醫(yī)生、患者、實(shí)驗(yàn)檢查結(jié)果或干預(yù)措施的可獲得性分配患者入組 不清楚： 根據(jù)干預(yù)措施的可獲得性 文中信息不詳，難以判斷正確與否 (

6、2) Was allocation adequately concealed? Low risk of bias: concealed allocation was completed by clearly described methods so that either participants or investigators could not foresee assignment. High risk of bias: participants or investigators enrolling participants could possibly foresee assignme

7、nts Uncertain risk of bias: insufficient information to permit judgment of Yes or No of concealment（2）分配方案隱藏 完善： 中心隨機(jī)，包括采用電話、網(wǎng)絡(luò)和藥房控制的隨機(jī) 按順序編號(hào)或編碼的相同容器 按順序編碼、密封、不透光的信封 不完善： 公開(kāi)隨機(jī)分配序列如列出隨機(jī)數(shù)字 未密封、透光或未按順序編號(hào)的信封 交替分配 根據(jù)住院號(hào)、生日等末尾數(shù)字的奇數(shù)或偶數(shù) 不清楚： 未提及分配方案隱藏 提供的信息不能判斷是否完善，如使用信封，但未描述是否按順序編碼、密封、不透光隨機(jī)分配方案的隱藏隨機(jī)分配方案的隱藏

8、 隨機(jī)分配受試對(duì)象的過(guò)程中，受試對(duì)象和選擇合格受試對(duì)象的研究人員不能預(yù)先知道隨后的分配方案，目的在于防止選擇性偏倚 未隱藏分配方案或分配方案隱藏不完善的試驗(yàn)，常常夸大治療效果（3041）(3) Was knowledge of the allocated intervention (s) adequately prevented during the study? （ Blinding?） Low risk of bias: blinding of participants and key study personnel, or the outcome measurement was not

9、likely to be influenced by lack of blinding; High risk of bias: no blinding of participants and key study personnel, or the outcome measurement was likely to be influenced by lack of blinding Uncertain risk of bias: insufficient information to permit judgment of Yes or No of blinding盲法盲法 (Blindness)

10、(Blindness) 臨床試驗(yàn)中，試驗(yàn)的研究者或受試者，都不知道試驗(yàn)對(duì)象的分配情況，即不知道受試對(duì)象在試驗(yàn)組還是對(duì)照組，接受的是試驗(yàn)措施還是對(duì)照措施。 盲法主要用于研究措施的實(shí)施者、 結(jié)果測(cè)量，也可用于資料分析和報(bào)告 目的：避免測(cè)量性偏倚盲法的分類(lèi)盲法的分類(lèi)單盲（Singleblind）：只是受試對(duì)象不知道他們是在試驗(yàn)組或?qū)φ战M。檢驗(yàn)人員、病理醫(yī)師等不知道受試對(duì)象的診斷情況也屬于單盲。雙盲（Double-blind）：受試對(duì)象和試驗(yàn)措施實(shí)施者雙方都不知道干預(yù)措施分組情況。三盲（Triple-blind）：受試對(duì)象、研究者和資料分析或報(bào)告者都不知道受試對(duì)象分配在哪一組和接受哪一種干預(yù)措施。 不

11、能把“分配隱藏”與盲法混淆 分配隱藏實(shí)質(zhì)是產(chǎn)生隨機(jī)序列者和決定分配序別者不能參與納入受試者，也不宜參與以后的試驗(yàn)過(guò)程，尤其不能參與結(jié)果的測(cè)量。隱蔽分組（分配隱藏）和盲法的概念、實(shí)施與報(bào)告 吳泰相 劉關(guān)鍵中國(guó)循證醫(yī)學(xué)雜志, 2007, 7(3): 222225.盲法 正確： 沒(méi)有采用盲法，但結(jié)果判斷和測(cè)量不會(huì)受影響 對(duì)患者和主要研究人員采用盲法，且盲法不會(huì)被破壞 對(duì)結(jié)果測(cè)量者采用盲法，未對(duì)患者和主要研究人員采用盲法，但不會(huì)導(dǎo)致偏倚 不正確： 未采用盲法或盲法不完善，結(jié)果判斷或測(cè)量會(huì)受影響 對(duì)患者和主要研究人員采用盲法，但盲法可能被破壞 對(duì)患者和主要研究人員均未采用盲法，可能導(dǎo)致偏倚： 不清楚 信

12、息不全，難以判斷是否正確 文中未提及盲法(4) Were incomplete outcome data adequately addressed? Low risk of bias: no missing outcome data, or the data exist but do not have an impact on observed clinical effects. High risk of bias: missing outcome data do exist and have an impact on observed clinical effects. Uncertain

13、risk of bias: insufficient reporting of attrition and exclusions to permit judgment of Yes or No of missing data and its effects.結(jié)果數(shù)據(jù)不完整 完整： 無(wú)缺失數(shù)據(jù)； 缺失數(shù)據(jù)不影響結(jié)果分析（生存分析中缺失值）； 組間缺失的人數(shù)和原因相似； 缺失數(shù)據(jù)不足以對(duì)效應(yīng)值產(chǎn)生重要影響； 缺失數(shù)據(jù)采用恰當(dāng)方法賦值 不完整： 組間缺失的人數(shù)和原因不平衡； 缺失數(shù)據(jù)足以對(duì)效應(yīng)值產(chǎn)生重要影響； 采用“as-treated”分析(實(shí)際治療分析 )，但改變隨機(jī)入組時(shí)干預(yù)措施的人數(shù)較多；

14、不恰當(dāng)應(yīng)用簡(jiǎn)單賦值 不清楚： 信息不全，難以判斷數(shù)據(jù)是否完整（缺失人數(shù)或原因未報(bào)告）； 文中未提及數(shù)據(jù)完整性問(wèn)題 (5) Are reports of the study free of any suggestion of selective outcome reporting? Low risk of bias: all outcomes of interest in the review have been reported in a pre-specified way by the authors High risk of bias: not all outcomes of intere

15、st in the review have been reported in a pre-specified way by the authors Uncertain risk of bias: insufficient information to permit judgment of Yes or No of this question.選擇性報(bào)告結(jié)果 無(wú)選擇性報(bào)告結(jié)果： 有研究方案，且系統(tǒng)評(píng)價(jià)關(guān)心的方案中預(yù)先指定的結(jié)果指標(biāo)（主要和次要結(jié)果）均有報(bào)告； 沒(méi)有研究方案，但所有期望的結(jié)局指標(biāo)，包括在發(fā)表文獻(xiàn)中預(yù)先指定的指標(biāo)均有報(bào)告 有選擇性報(bào)告結(jié)果： 未報(bào)告所有預(yù)先指定的主要結(jié)局指標(biāo)； 報(bào)告的

16、一個(gè)或多個(gè)主要結(jié)局指標(biāo)采用預(yù)先未指定的測(cè)量和分析方法； 報(bào)告的一個(gè)或多個(gè)主要結(jié)局指標(biāo)未預(yù)先指定； 系統(tǒng)評(píng)價(jià)關(guān)心的一個(gè)或多個(gè)結(jié)局指標(biāo)報(bào)告不完善，以致不能納入行meta-分析； 未報(bào)告重要的結(jié)局指標(biāo) 不清楚：信息不全，難以判斷是否存在選擇性報(bào)告結(jié)果 (6) Was the study apparently free of other problems that could place it at risk of bias? Low risk of bias: the study appears to be free risk of bias. High risk of bias: there is

17、 at least one important risk of bias existing in the study. Uncertain risk of bias: insufficient information to assess whether an important risk of bias exists.其它影響真實(shí)性因素：無(wú)其它偏倚 無(wú)：納入研究無(wú)其它偏倚來(lái)源 有：至少存在一種重要偏倚風(fēng)險(xiǎn) 與使用的研究設(shè)計(jì)方案相關(guān)的偏倚 提前終止研究（數(shù)據(jù)原因或正規(guī)終止原則） 明顯基線不平衡 聲稱(chēng)有欺騙行為 其它問(wèn)題 不清楚 信息不全，難以判斷是否存在重要偏倚 發(fā)現(xiàn)的問(wèn)題是否導(dǎo)致偏倚，理由或依

18、據(jù)不足Summary assessments of the risk of bias for each study Low risk of bias: Low risk of bias for all key domains; any plausible bias is unlikely to seriously alter the study results. Unclear risk of bias: Unclear risk of bias for one or more key domains; any plausible bias raises some doubt about th

19、e study results. High risk of bias: High risk of bias for one or more key domains, any plausible bias may seriously weaken confidence in the results.總分評(píng)分細(xì)則 在各個(gè)方面將試驗(yàn)評(píng)為A(是)、B(不清楚)、C(否)3級(jí) 在這里，可以把A、B和C分別解讀為L(zhǎng)ow, Moderate和High risk of bias 如各評(píng)價(jià)條目均為A級(jí), 則為低度偏倚, 發(fā)生各種偏倚的可能性最小,所有評(píng)價(jià)指標(biāo)均有詳細(xì)描述; 質(zhì)量評(píng)為A級(jí); 若有一個(gè)條目或多個(gè)條目

20、為B, 則該試驗(yàn)有發(fā)生相應(yīng)偏倚的中等度可能性,所有評(píng)價(jià)指標(biāo)中至少有1項(xiàng)指標(biāo)提及但未詳細(xì)描述; 質(zhì)量評(píng)為B級(jí); 如其中有一個(gè)條目或多個(gè)條目為C,則該試驗(yàn)有發(fā)生相應(yīng)偏倚的高度可能性, 所有評(píng)價(jià)指標(biāo)中至少1項(xiàng)指標(biāo)為不充分或未使用; 質(zhì)量評(píng)為C級(jí);EntryEntryJudgementJudgementDescriptionDescriptionAdequate sequence generation?Yes.Quote: “patients were randomly allocated.”Comment: Probably done, since earlier reports from the

21、same investigators clearly describe use of random sequences (Cartwright 1980).Allocation concealment?No.Quote: “.using a table of random numbers.”Comment: Probably not done.Blinding? (Patient-reported outcomes)Yes.Quote: “double blind, double dummy”; “High and low dose tablets or capsules were indis

22、tinguishable in all aspects of their outward appearance. For each drug an identically matched placebo was available (the success of blinding was evaluated by examining the drugs before distribution).”Comment: Probably done.Blinding? (Mortality)Yes.Obtained from medical records; review authors do not

23、 believe this will introduce bias.Incomplete outcome data addressed? (Short-term outcomes (2-6 wks)No.4 weeks: 17/110 missing from intervention group (9 due to lack of efficacy); 7/113 missing from control group (2 due to lack of efficacy).Incomplete outcome data addressed? (Longer-term outcomes (6

24、wks)No.12 weeks: 31/110 missing from intervention group; 18/113 missing from control group. Reasons differ across groups.Free of selective reporting?No.Three rating scales for cognition listed in Methods, but only one reported.Free of other bias?NoTrial stopped early due to apparent benefit.Jadad sc

25、ore 1) Was the study described as randomized?2) Was the method used to generate the sequence of randomization described and appropriate?3) Was the study described as double blind?4) Was the method of double blinding described and appropriate?5) Was there a description of withdrawals and dropouts?質(zhì)量評(píng)

26、價(jià)注意事項(xiàng)質(zhì)量評(píng)價(jià)注意事項(xiàng)兩個(gè)以上評(píng)價(jià)人獨(dú)立評(píng)價(jià)；兩個(gè)以上評(píng)價(jià)人獨(dú)立評(píng)價(jià)；專(zhuān)業(yè)人員和方法學(xué)家；專(zhuān)業(yè)人員和方法學(xué)家；不同意見(jiàn)討論解決，如仍有分歧，咨詢第三不同意見(jiàn)討論解決，如仍有分歧，咨詢第三位評(píng)價(jià)人；位評(píng)價(jià)人；仍不能解決，咨詢?nèi)圆荒芙鉀Q，咨詢COCHRANECOCHRANE專(zhuān)業(yè)組專(zhuān)業(yè)組非隨機(jī)對(duì)照試驗(yàn)的質(zhì)量評(píng)價(jià) 不是所有的研究類(lèi)型都能進(jìn)行隨機(jī)對(duì)照試驗(yàn) 病因?qū)W研究、診斷學(xué)研究、預(yù)后研究 不是所有的治療性研究都能進(jìn)行或者需要進(jìn)行隨機(jī)對(duì)照試驗(yàn) 倫理、可行性、經(jīng)費(fèi) 非隨機(jī)對(duì)照試驗(yàn)的系統(tǒng)評(píng)價(jià)方法 基本步驟：同隨機(jī)對(duì)照試驗(yàn) 但目前沒(méi)有公認(rèn)的、規(guī)范的手冊(cè)和方法Tools for assessing quality and