메타분석에있어서무작위대조연구의질분석 http://dx.doi.org/10.7599/hmr.2015.35.1.33 pissn 1738-429X eissn 2234-4446 이승욱 한양대학교의과대학비뇨기과학교실 Meta-Analysis and Quality Assessment of Randomized Controlled Trials Seung Wook Lee Department of Urology, Hanyang University College of Medicine, Seoul, Korea Meta-analysis is the statistical combination of results from two or more separate studies. Potential advantages of meta-analyses include an increase in power, an improvement in precision, the ability to answer questions not posed by individual studies, and the opportunity to settle controversies arising from conflicting claims. However, they also have the potential to mislead seriously, particularly if specific study designs, within-study biases, variation across studies, and reporting biases are not carefully considered. It is important to be familiar with the type of data (e.g. dichotomous, continuous) that result from measurement of an outcome in an individual study, and to choose suitable effect measures for comparing intervention groups. Most meta-analysis methods are variations on a weighted average of the effect estimates from the different studies. Variation across studies (heterogeneity) must be considered. Random-effects meta-analyses allow for heterogeneity by assuming that underlying effects follow a normal distribution. Various judgments are required in the process of preparing a meta-analysis. Especially, quality assessment of randomized controlled trial is essential. There are several methods to assess the methodological quality of clinical trials, including scales, individual markers, and checklists. Analyzing the quality of studies makes the results of meta-analysis more reliable. Sensitivity analyses should be used to examine whether overall findings are robust to potentially influential decisions. Correspondence to: Seung Wook Lee 우 471-701, 경기도구리시경춘로 153, 한양대학교구리병원비뇨기과 Department of Urology, Hanyang University Guri Hospital, Hanyang University College of Medicine, 153 Gyeongchun-ro, Guri 471-701, Korea Tel: +82-31-560-2375 Fax: +82-31-560-2372 E-mail: swleepark@hanyang.ac.kr Received 2 December 2014 Revised 26 December 2014 Accepted 9 January 2015 This is an Open Access article distributed under the terms of the Creative Commons Attribution n-commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. Key Words: Meta-Analysis; Quality Assurance; Health Care; Randomized Controlled Trial 서론메타분석 (Meta-analysis) 은개인이수행하기어려울정도로많은정보속에서근거를종합하여신뢰할수있는객관적근거를제공하는것으로일차연구만으로는어떤문제에결론을내리기어려운상황에서문헌을포괄적으로검토해서결론을내리는것이며, 근거중심의 (Evidence-based medicine, EBM) 의등장과함께의학분야의중요한연구분야로자리잡았다. 메타분석의본격적인등장은 1970년대이후에들어서이루어졌다. 1976년에 Glass [1] 가 메 타분석 이라는용어를사용하기시작하였고, Mulrow [2] 는서술적고찰의오류가능성을지적하고체계적고찰의중요성을실증적으로제시하였다. Glass는문헌들을고찰 (review) 하는것그자체가하나의연구방법론에해당하며, 따라서이고찰역시다른공식적인연구에못지않게체계적으로 (systematically) 수행되어야한다고주장했다. 한편의료행위중많은부분이적절한근거를기반으로이루어지지않고있음을설명한 Archie Cochrane은메타분석분야의선구자중한사람이다. 역학자인그는 1972년그의유명한원고 Effective- http://www.e-hmr.org 2015 Hanyang University College of Medicine 33
HMR Seung Wook Lee Meta-Analysis and Quality Assessment of Randomized Controlled Trials ness and efficiency에서의료와관련된의사결정을하고자하는정책결정자, 전문가및서비스의소비자 (policy makers, professionals, consumers of services) 이현재가용하고, 믿을만하며, 엄격하게정리된근거들을쉽게접근할수있는방법을제공받지못하고있다고언급하였다 [3,4]. 이러한 Cochrane의비판으로모든보건및의료관련연구결과들을데이터베이스화하고자하는임상연구자들, 연구방법론전문가들및사용자들간국제적네트워크인코크란협력단체 (Cochrane Collaboration) 를탄생시키는밑거름이되었을뿐만아니라 [5], 궁극적으로는근거중심의학운동을점화시키는도화선역할을하게되었다 [6]. 이를위해두종류의근거, 즉, 임상가본인의전문적지식인내적근거와다른연구자들의연구결과인외적근거들을종합해임상적인의사결정을하고자하는것이다. 따라서 EBM을시행하기위해서는환자들을보살피는데도움이되는연구들을찾아내고, 이연구결과들을제대로평가할수있어야할것이다 [7]. 체계적고찰에서는통계적인방법을사용해각연구의상대적인공헌도를감안하면서자료들을결합할수도있지만, 연구들간의특성이너무이질적이면자료의결합을시도하지않을수도있다. 이러한자료결합단계를메타분석이라한다. 메타분석은대규모연구들뿐만아니라, 사용된연구방법, 연구대상자들의특성, 결과에영향을미칠수있는다른중요한요인들등에있어서변동을보이는, 소규모로진행된연구들에대해서도사용될수있는방법론이다. 메타분석의목적은활용가능한연구들의장점및제한점들에기초하면서객관적인결론을유도하고자하는것이다 [8]. 즉, 유사한결과를보기위해유사한환자들을대상으로유사한처리를사용한유사한연구들의자료를결합함으로써가능한한최대의통계적검정력 (power) 과정밀성 (precision) 을확보한상태에서해당처리효과를추정하고자하는것이다 [9]. 하지만, 메타분석에서활용가능한연구들의자료의선정에는여러가지고려해야할점이있다. 연구자료수집단계에서활용가능연구자료를모두사용할것인지양질의연구디자인으로수행된높은근거수준을보이는연구의자료들만을택할것인지는항상주의깊게판단해야한다. 대상연구들의디자인의질이낮은경우, 결과를도출하는과정에서오류가발생할가능성이있고이로인해결과가왜곡될수있다. 이러한이유로메타분석의과정에서는대상연구의질평가및자료추출단계가존재하는데, 메타분석은주로결과의신뢰를높이기위해서근거수준이높은무작위대조군연구 (Randomized Controlled Trial, RCT) 를대상으로진행한다. 무작위대조군연구는비뚤림 (bias) 를줄일수있는연구디자인이기때문에연구방법중가장근거가높게평가되며의학적치료의효과를평가하기위한가장믿을만한방법으로고려된다 [10]. 그러나무작위대조군연구라도모든비뚤림을없앨수는없으며연구기획, 수행, 보고혹은적용등여러단계에서비뚤림은발생할수 있으며이러한실수가잘못된결론을도출하게된다 [11]. 그렇기때문에메타분석은대상연구의질적인분석을시행하여높은질의연구를대상으로분석을시행해야그릇된정보가임상에적용되는것을방지할수있다. 이글에서는체계적고찰중메타분석의수행과해석및향후방향과무작위대조군연구의질적분석에대해소개하고자한다. 체계적고찰의자료원메타분석의핵심은해당주제와관련된정보들을얼마나최대한그리고얼마나체계적으로찾아내서요약했는가에있다. 제대로된메타분석을수행하기위해서는연구주제에대한전문적인지식뿐만아니라해당방법론에대한전문적인지식또한필요하다. 따라서메타분석은임상전문가와경험이있는통계전문가간의공동연구로진행되는경우가거의대부분이다. 연구주제와관련된선정기준에해당하는연구들을최대한으로찾아내고메타분석과정중에발생하게될불확실성을줄이기위해서는무엇보다도먼저메타분석주제에관한명확한설정이필요하다. 그다음으로선정기준에부합하는모든적절한연구결과들을파악하기위한문헌탐색이광범위하게이루어져야한다. 탐색된문헌들의수가매우많기때문에이들이메타분석포함기준에적절한지평가하기위해먼저초록을검토한후, 그다음으로자세한평가와자료추출그리고연구의질평가를위해전체논문 / 내용을검토하는과정에서체계적고찰의자료원이필요하며몇가지의종류가있다. 코크란연합에서제공되는코크란라이브러리 (www.thecochranelibrary.com) 는코크란리뷰와 York Database of Abstracts of Reviews of Effectiveness (DARE) 등다른체계적고찰들이수록되어있는데이터베이스이다. 코크란라이브러리에서는 advanced search 를사용하여코크란자료와 DARE의자료를구분하여체계적고찰및메타분석을검색할수있다. 코크란라이브러리는유료이지만, 현재에대부분의우리나라의과대학에서구독하고있어이를통해접근할수있다. 이외체계적고찰이나메타분석만전문적으로검색할수있는검색엔진으로는 TRIP (http://www.tripdatabase.com), Sumsearch (http://sumsearch.uthscsa.edu) 등이있다. 이들은이차검색을통해검색결과를제시하며논문만이아니라기관의체계적고찰보고서등도자료원별로제시하고있다. 메타분석관련프로그램메타분석은소프트웨어를이용하여수행하게되는데몇가지주요한소프트웨어가있다. 코크란연합에서제공하는 Revman이라는프로그램은코크란홈페이지를통해무료로다운로드받을수있는데, 메타회귀분석이나베이지안분석등을제외한대부분의 34 http://www.e-hmr.org
이승욱 메타분석에있어서무작위대조연구의질분석 HMR 메타분석을다룰수있고, 숲그림 (Forest plot) 이나, 깔때기그림 (Funnel plot) 등을제공하며코크란리뷰를할때유용하다. 상용프로그램으로써 Comprehensive Meta-analysis 나 R 등의프로그램이있다. 이들프로그램은메타분석전용이기에비교적사용하기쉽게다양한방식의기본적인메타분석과그림그리기메뉴를제공하여메타회귀분석또한가능하다. SAS나 STATA 등일반적인통계소프트웨어에서도메타분석을할수있다. 매크로를이용하여메타분석을할수있으며이질성검정, 메타회귀분석등이가능하다. 진단분석등복잡한자료를운영하는데적합하지만, 초보자의경우통계전문가의자문을받을필요가있다. 아울러 1990년이후체계적고찰과메타분석을위한다수의책들이출간되어참고할수있으며, 특히코크란핸드북은 PDF 파일의형태로다운로드가가능하다 [12-16]. 메타분석의수행메타분석은보통다음과같은다섯단계에걸쳐수행하게된다 [14,17]. 1) 연구주제설정 (formulating research questions) 2) 연구포함기준및제외기준설정 (defining inclusion and exclusion criteria) 3) 문헌탐색및논문선택 (literature search and study selection) 4) 연구의질평가및자료추출 (study quality assessment and data extraction) 5) 통계분석및결과제시 (statistical analysis and result presentation) 1. 연구주제설정메타분석의첫단계는연구주제를명확히설정하고연구목적과가설들을자세히명시한프로토콜을작성하는것이다. 연구주제는당연히임상적으로관심이있는것이어야하겠지만, 주로확증된것보다는논란이많은주제가선택되게된다. 연구의목적은구체적이고명확해야하며중재의종류, 비교집단에관한내용, 임상적관심변수등이포함되어야한다 [14]. 2. 연구포함기준및제외기준설정연구주제에합당한모든적절한연구들을찾아내기위해서는먼저연구선정기준을자세히정의해야한다. 여기에는주로연구설계의형태 ( 예 : RCT로한정할것인지아니면코호트연구등관찰연구도포함할것인지 ), 환자군의특성 ( 예 : 연령, 성별, 의학적상태여부등 ), 자료의출판형태 ( 예 : 초록집에만출판된자료를포함시킬것인지, 미출판자료역시포함시킬것인지 ), 사용된언어 ( 예 : 영어권이외에출판된연구들을포함시킬것인지 ), 그리고연구기간 ( 예 : 어 느기간동안에수행된연구들로한정할것인지 ) 등이포함된다. 일반적으로메타분석대상은무작위이중눈가림임상시험들이다. 왜냐하면무작위할당 (random allocation) 이이루어져야만비뚤림이최소화되거나집단간에비뚤림의균형이이루어져서예후에대한비교성이확보될수있기때문이다 [19]. 그리고포함기준은주연구주제에관한결론을내릴수있게끔구성되어야만한다 [17]. 3. 문헌탐색및논문선택연구주제에관한답을하기위해서는주제와관련이있는연구들을가능한한많이찾아내는것이중요하며, 설정된포함기준에해당하는연구들을모두찾아내도록노력해야한다. 또한문헌탐색시사용된특정탐색어들은연구결과보고서 ( 혹은논문 ) 에제시해야한다. 효과적인문헌탐색을위해서는아래의전략들을단계적으로사용하는것이좋다 : 1) Cochrane Library, MEDLINE, EMBASE, Life sciences and Biomedicine (BIOS), Science Citation Index (SCI), Nursing and Allied Health Literature (CINAHL), Allied and Alternative Medicine (AMED), Biological Abstracts (Biosis), Psychological Abstracts (Psyc INFO) 등전자의료관련데이터베이스를사용해상호심사과정을거친논문 (peer-reviewed journal) 을찾는다. 2) 해당주제와관련된출판서적이나관련학술회의의초록집등을검토한다. 3) 국내외관련학회의웹사이트등특정인터넷사이트를조사하고, 해당질병과관련된진료가이드라인등을살펴본다. 4) 전국혹은지역단위의등록체계 (registry) 를조사하고미출판된논문들을찾는다. 이를위해해당질병과관련해명망있는학자나의료센터등을접근해본다. 5) 출판되었거나미출판된연구자료에대한정보를입수하기위해제약회사의의료관련담당자들과연락을취한다. 6) Intute (www.intute.ac.uk) 등과같은의료관련포털사이트및 Google, Yahoo 등과같은일반포털사이트를조사한다. 7) 연구설계의종류 ( 예 : 평행설계 [parallel design] 인지교차설계 [crossover design] 인지 ) 8) 각연구의환자모집단포함기준 ( 예 : 관상동맥질환을가진심장병환자군대관상동맥질환을가지고있지않은환자군, 모든제2형당뇨병환자군대최소한하나이상의심혈관계질환위험인자를가지고있는제2형당뇨병환자군 ) 9) 각연구의환자모집단제외기준 ( 예 : 65세이상인심장병환자또는심장병환자이면서신장질환을가지고있는환자, 제2 형당뇨병환자이면서장기인슐린치료중인환자 ) 10) 각연구가평가한중재방법 ( 예 : 베타차단제대위약인지아니면베타차단제대이뇨제인지, 혈당조절을위한단독요법대 http://www.e-hmr.org 35
HMR Seung Wook Lee Meta-Analysis and Quality Assessment of Randomized Controlled Trials 위약인지, 단독요법대병합요법인지아니면강화혈당조절요법대표준혈당조절요법인지 ) 11) 연구에사용된일차평가변수 ( 예 : 사망인지재입원인지, 주요심혈관계질환의발생률인지심혈관계질환관련사망률인지 ) 4. 연구의질평가및자료추출연구의질에대한평가는매우중요하다. 연구의질이란우리가보기에해당연구가 실제 효과를얼마나잘측정했는지에관한정도이며, 이러한측면의질을타당도 (validity) 라한다. 연구의타당도는내적타당도 (internal validity)- 해당연구결과가원래연구하고자했던상황을얼마나정확히반영하는지에관한정도. 즉, 비뚤림이최소화되는것- 와외적타당도 (external validity)- 해당연구결과가다른상황에까지얼마나정확히일반화될수있는지에관한정도. 즉, 결과의일반화가능성-로구분하며, 메타분석에서연구의질평가란선택한논문들에대해이러한타당도를평가하는것을의미한다. 연구의질평가에대한언급은 Glass까지거슬러올라간다 [20]. 메타분석시연구의질에대한평가가반드시이루어져야하는이유는분석에사용된연구들이실제로는질이낮은연구들임에도불구하고분석결과는마치높은수준의연구를통해얻어진결과인것처럼제시되거나보일수있기때문이며, 이보다더중요한문제는질이낮은연구결과들을결합하게되면비뚤림이증가되며, 따라서결과가왜곡 (misleading) 되는결합추정치가얻어지게되기때문이다. Garbage in, garbage out, 즉아무리정교한분석방법을사용해도질이낮은자료의한계점은극복할수없다. 메타분석을수행할때가장간과할수없는형태의비뚤림이바로이연구의질평가와관련된비뚤림이며, 모의실험결과, 연구의질에주로영향을미치는것은부적절한방법론의사용, 그중에서도특히무작위배정에대한은폐 (random allocation concealment) 가제대로이루어지지않았을때가장큰비뚤림 ( 주로처리효과가과장되는방향으로의비뚤림 ) 의결과가초래되는것으로나타났다 [21]. 현재주로사용되고있는질평가척도로는 Chalmers scale [22] 및 Jadad scale [25] 등을들수있는데, Chalmers scale은연구의내적, 외적타당도및자료분석측면등을평가하는전체 30개문항 (100 점만점 ) 으로이루어져있으며, Jadad scale은연구의내적타당 도를평가하는 5개문항, 즉, 1) 무작위화가실시되었다고기술되었는가? 2) 연구대상자들의처리군할당에사용된일련번호들은적절하게생성되었다고기술되었는가? 3) 이중눈가림이실시되었다고기술되었는가? 4) 대조군에사용된처치 ( 예를들면위약 ) 는실험군의처치와구별될수없었다고묘사되었는가? 5) 각집단별로추적실패또는분석에서제외된환자의수등손실내용 (attrition) 및그이유가기술되어있는가? 등으로구성된 5점만점의척도이다. Jadad scale이가장많이사용되는척도이긴하지만, 연구의실제방법론적인측면보다는작성된논문의내용에관한질을평가하는척도라는점, 그리고무작위배정과관련된은폐여부보다는무작위배정일련번호의생성여부에초점이맞추어져있다는것에대한비판도있다 [17,23]. 아래는현재무작위대조연구에대한질평가도구중가장흔히사용되고있는도구들을나열하였다. 1) Jadad scale Jadad scale은 Oxford quality scoring system이라고도불리며, 총 3문항으로구성된 RCT를평가하기위한척도이다. 무작위배정에관한문항 2점, 맹검에관한문항 2점, 탈락에관한문항 1점으로구성된다. 무작위배정에관해서 RCT에서무작위배정을언급하면 1점, 적절한무작위배정을포함하고있으면 1점이추가되며부적절한무작위배정은 -1점으로 scoring되며총 0-2점으로분포된다. 맹검에대해서는양측맹검이언급되면 1점, 적절한맹검이있으면 1점추가되며부적절한맹검이있으면 -1점으로점수화되며최고 2점으로분포된다. 탈락에관해서도탈락에관한내용이언급되었을경우 1점을부여한다. 최종문헌의질평가는총 5점만점으로평가하며 0-2점은문헌의질이 low quality로 3-5점은 high quality 로평가한다 [25] (Table 1). 2) Van Tulder scale Van Tulder scale은 RCT의질평가에적합한도구중의하나로무작위적절성, 치료배정은닉, 기본특성의유사성, 환자맹검, 치료제공자맹검, 관찰자맹검, 동시중재, 규정준수, 탈락 / 탈락률, 결과평가시점, 무작위배정에대한은폐의분석등 11개의요소로평가 Table 1. Jadad scale [25] Item Maximum points Description Randomization 2 1 point if randomization is mentioned 1 additional point if the method of randomization is appropriate Deduct 1 point if the method of randomization is inappropriate (minimum 0) Blinding 2 1 point if blinding is mentioned 1 additional point if the method of blinding is appropriate Deduct 1 point if the method of blinding is inappropriate (minimum 0) An account of all patients 1 The fate of all patients in the trial is known. If there are no data the reason is stated 36 http://www.e-hmr.org
이승욱 메타분석에있어서무작위대조연구의질분석 HMR Table 2. Van Tulder scale [26] Criteria A Was the method of randomization adequate? yes / no / don't know B Was the treatment allocation concealed? yes / no / don't know C Were the groups similar at baseline regarding the most important prognostic indicators? yes / no / don't know D Was the patients blinded to the intervention? yes / no / don't know E Was the care provider blinded to the intervention? yes / no / don't know F Was the timing of the outcome assessment in all groups similar? yes / no / don't know G Were co-interventions avoided or similar? yes / no / don't know H Was the compliance acceptable in all groups? yes / no / don't know I Was the drop-out rate described and acceptable? yes / no / don't know J Was the timing of the outcome assessment in all groups similar? yes / no / don't know K Did the analysis include an intention-to-treat analysis? yes / no / don't know 한다. 각각의항목에대해,, don t know 로평가하며 5점이상인경우문헌이 high quality로평가한다 [26] (Table 2). 3) Cochrane collaboration risk of bias tool (CCRBT) Cochrane 연합은배정순서생성 (Sequence generation), 무작위배정에대한은폐, 맹검, Incomplete outcome data, Selective outcome reporting, Other potential threats to validity 로구분하여평가하고있다. 각항목별로,, unclear 로평가하며평가자는 6개영역별로 yes, no, unclear에대한세부기준에따라판단하고평가한다. 는 high quality, 는 low quality, unclear 는정보가불충분한경우를의미한다. Low risk of bias는첫 3문항이 yes 이고 no important concerns related to the last three domains were identified로분류하며, moderate risk of bias는 2개이하의 domain에서 unclear or no 로평가된경우로분류하였고 High risk of bias는 3개이상의 domain에서 unclear or no 로평가된경우로분류하였다 [27] (Table 3). 5. 통계분석및결과제시메타분석의마지막단계는적절한통계적방법을사용해연구결과들을결합하는작업이다. 자료의결합은관심변수에관한각연구의개별환자자료 (individual patient data, IPD) 를사용해결합할수도있고 ( 물론이를위해서는각연구의환자자료를얻어야함 ), 각연구의해당변수에대한요약자료를사용할수도있다. 요약자료를사용해처리간효과차이에대한메타분석을하기위해서는각연구에대해서기본적으로다음과같은두가지정보가필요하다. 1) 처리효과차이에대한동일한형태의측정값 : 예를들면, 승산비 (Odds Ratio, OR), 위험비 (Risk Ratio, RR) 또는평균차이 (Mean of Difference, MD) 등 2) 해당처리효과에대한표준오차 ( 또는분산 ) 메타분석에서각 연구의효과차이들을결합할때사용하는주개념은각연구의표본수크기에근거한가중평균을사용하는것이다. 즉, 소규모연구는대규모연구에비해우연한차이로인한영향을더많이받기때문에결합시대규모연구에상대적으로더많은가중치를부여하고자하는것이다 [28]. 결합효과및해당신뢰구간을추정하고결합효과에대한가설검정을실시하기위해메타분석에서주로사용되는통계적모형으로는고정효과모형 (fixed-effects model) 과변량효과모형 (random-effects model) 이있다. 이들두모형의차이는각논문에서얻어진요약자료가어떤분포에서얻어진자료인지에대한통계적인관점의차이라할수있다 [29]. 메타분석결과의제시메타분석연구를실시할때무엇보다도중요한것은모든분석과정을투명하게제시하고, 연구들의포함기준및제외기준을명확히기록해두는것이다. 또한이선택기준들을적용한문헌탐색대상자료원과사용된탐색전략들을명시해주어야하고, 선택된연구들의질을평가하는데사용된기준들에대해서도자세히기술해주어야한다. 이렇게하는이유는연구자가논문을선택하는과정중에본인의의도와는상관없이발생될수도있는비뚤림들을투명하게밝혀서독자들로하여금해당메타분석의결과에기초한근거의한계들, 이러한한계들을극복하기위한노력, 그리고해당결과로부터유도된추론의신뢰성등을평가할수있도록해주기위해서이다. 논문을작성할때에는 RCT를대상으로한메타분석의경우, 연구의질을향상시키고그기준을제시하기위해개발된, QUOROM 지침 [30] 을, 그리고관찰연구를대상으로한메타분석의경우에는 MOOSE 지침 [27] 을준수하는것이좋다. http://www.e-hmr.org 37
HMR Seung Wook Lee Meta-Analysis and Quality Assessment of Randomized Controlled Trials Table 3. Cochrane's assessment of risk of bias [27] Item Sequence generation Allocation concealment Blinding Incomplete outcome data Selective outcome reporting Other potential threats to validity Criteria Referring to a random number table, using a computer random number generator, coin tossing, shuffling cards or envelopes, throwing dice, drawing of lots, minimization* Sequence generated by odd or even date of birth, sequence generated by some rule based on date (or day) of admission, sequence generated by some rule based on hospital or clinic record number Central allocation (including telephone, web-based and pharmacy-controlled randomization), sequentially numbered drug containers of identical appearance, sequentially numbered, opaque, sealed envelopes Using an open random allocation schedule(e.g. a list of random numbers), assignment envelopes were used without appropriate safeguards, alternation or rotation, date of birth, case record number, any other explicitly unconcealed procedure blinding, but the review authors judge that the outcome and the outcome measurement are not likely to be influenced by lack of blinding; Blinding of participants and key study personnel ensured, and unlikely that the blinding could have been broken; Either participants or some key study personnel were not blinded, but outcome assessment was blinded and the non-blinding of others unlikely to introduce bias blinding or incomplete blinding, and the outcome or outcome measurement is likely to be influenced by lack of blinding; Blinding of key study participants and personnel attempted, but likely that the blinding could have been broken; Either participants or some key study personnel were not blinded, and the non-blinding of others likely to introduce bias missing outcome data; Reasons for missing outcome data unlikely to be related to true outcome; Missing outcome data balanced in numbers across intervention groups, with similar reasons for missing data across groups; For dichotomous outcome data, the proportion of missing outcomes compared with observed event risk not enough to have a clinically relevant impact on the intervention effect estimate; For continuous outcome data, plausible effect size among missing outcomes not enough to have a clinically relevant impact on observed effect size; Missing data have been imputed using appropriate methods Reason for missing outcome data likely to be related to true outcome, with either imbalance in numbers or reasons for missing data across intervention groups; For dichotomous outcome data, the proportion of missing outcomes compared with observed event risk enough to induce clinically relevant bias in intervention effect estimate; For continuous outcome data, plausible effect size among missing outcomes enough to induce clinically relevant bias in observed effect size; As-treated analysis done with substantial departure of the intervention received from that assigned at randomization; Potentially inappropriate application of simple imputation The study protocol is available and all of the study s pre-specified (primary and secondary) outcomes that are of interest in the review have been reported in the pre-specified way; The study protocol is not available but it is clear that the published reports include all expected outcomes, including those that were pre-specified t all of the study s pre-specified primary outcomes have been reported; One or more primary outcomes is reported using measurements, analysis methods or subsets of the data (e.g., subscales) that were not pre-specified; One or more reported primary outcomes were not pre-specified (unless clear justification for their reporting is provided, such as an unexpected adverse effect); One or more outcomes of interest in the review are reported incompletely so that they cannot be entered in a meta-analysis; The study report fails to include results for a key outcome that would be expected to have been reported for such a study Had a potential source of bias related to the specific study design used; or Stopped early due to some data-dependent process (including a formal-stopping rule); or Had extreme baseline imbalance; or Has been claimed to have been fraudulent; or Had some other problem Insufficient information to assess whether an important risk of bias exists; or Insufficient rationale or evidence that an identified problem will introduce bias *Minimization may be implemented without a random element, and this is considered to be equivalent to being random. 결론기존의표준요법과비교해새로운치료법이더효과적인지에대해중간규모의임상시험연구들간에서로불일치한결과를보이는경우, 또는이연구들에서보여준효과크기들간에서로차이가나는경우등에있어서메타분석은해당요법이나중재에대한실제효과를좀더정확히측정할수있게끔해줄것이다. 그러나체계적 고찰및메타분석결과는임상적논리를보강해주는데도움을주지만이를대체할수는없다. 따라서메타분석결과를무비판적으로수용하거나무감각하게적용해서는안될것이며, 임상가의지식, 경험, 가치관및임상적근거등을통합한의사결정이이루어져야할것이다 [31]. 또한, 메타분석을수행하거나받아드리는경우에있어서도대상연구의질적분석에더많은관심을갖고정보를얻어야메타분석의결과에대한비선택적수용을막을수있을것이다. 38 http://www.e-hmr.org
이승욱 메타분석에있어서무작위대조연구의질분석 HMR 결론적으로질적분석을통해올바르게시행된메타분석의결 과는임상적근거를설정하는데중요한역할을한다. 비록방법론 의복잡성이메타분석의한계점중하나이긴하지만, 특정임상적 질문에대해명확한결론을내려주는대규모무작위임상시험결 과가없는한, 메타분석은해당질문에대해답을해줄수있는차 선책중하나이다. REFERENCE 1. Glass GV. Primary, secondary, and meta-analysis of research. Educ Res 1976:3-8. 2. Mulrow CD. The medical review article: state of the science. Ann Intern Med 1987;106:485-8. 3. Cochrane AL, Fellowship RC. Effectiveness and efficiency: random reflections on health services: Nuffield Provincial Hospitals Trust London; 1972. 4. Cochrane AL. 1931-1971: a critical review, with particular reference to the medical profession. Medicines for the year 2000. London: Office of Health Economics; 1979:1-11. 5. Chalmers I, Sackett D, Silagy C, Maynard A, Chalmers TC. n-random reflections on health services research. London: BMJ Publishing Group, 1997. 6. Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn t. BMJ 1996;312:71-2. 7. Sackett DL, Straus SE, Richardson WS, Rosenberg W, Haynes RB. Evidence-based medicine: how to practice and teach EBM. 2nd ed. Churchill -Livingston, London, 2000 8. Egger M, Smith GD. Meta-Analysis. potentials and promise. BMJ 1997; 315:1371-4. 9. DerSimonian R, Laird N: Meta-analysis in clinical trials. Control Clin Trials 1986;7:177-86. 10. Uetani K, Nakayama T, Ikai H, Yonemoto N, Moher D. Quality of reports on randomized controlled trials conducted in Japan: evaluation of adherence to the CONSORT statement. Intern Med (Tokyo, Japan) 2009;48:307-13. 11. Begg C, Cho M, Eastwood S, Horton R, Moher D, Olkin I, et al. Improving the quality of reporting of randomized controlled trials: the CON- SORT statement. JAMA 1996;276:637-9. 12. Petitti DB. Meta-analysis, decision analysis, and cost-effectiveness analysis: methods for quantitative synthesis in medicine. New York: Oxford University Press; 1999. 13. Sutton AJ, Abrams KR, Jones DR, Jones DR, Sheldon TA, Song F. Methods for meta-analysis in medical research. 1st ed. New York. John Wiley & Sons Ltd; 2000. 14. Egger M, Davey Smith G, Altman D. Systematic Reviews in Health Care: Meta-Analysis in Context, 2nd ed. London, BMJ Publishing Group, 2001. 15. Glasziou P, Irwig L, Bain C, Bain C, Colditz G, Irwig L. Systematic reviews in health care. Cambridge, Cambridge University Press 2001. 16. Stangl D, Berry DA. Meta-analysis in medicine and health policy. Basel: Marcel Dekker; 2000. 17. L Abbe KA, Detsky AS, O Rourke K. Meta-analysis in clinical research. Ann Intern Med 1987;107:224-33. 18. Meade MO, Richardson WS. Selecting and appraising studies for a systematic review. Ann Intern Med 1997;127:531-7. 19. Lee JY. Medical Statistics at a Glance. Seoul, Epublic 2007. 20. Glass GV. Primary, secondary, and meta-analysis of research. Educ Res 1976;5:3-8. 21. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical evidence of bias. Dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 1995;273:408-12. 22. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1-12. 23. Juni P, Witschi A, Bloch R, Egger M. The hazards of scoring the quality of clinical trials for meta-analysis. JAMA 1999;282:1054-60. 24. Egger M, Smith GD, Phillips AN. Meta-analysis: principles and procedures. BMJ 1997;315:1533-7. 25. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1-12. 26. Van Tulder M, Furlan A, Bombardier C, Bouter L. Updated method guidelines for systematic reviews in the cochrane collaboration back review group. Spine (Phila Pa 1976) 2003;28:1290-9. 27. Higgins JP, Green S, et al. Cochrane handbook for systematic reviews of interventions. Ver. 5.1.0. The Cochrane Collaboration; 2011 [cited 2011 Jan 7]. Available from: http://www.cochrane-handbook.org. 28. Hedges LV. Statistical Methodology in Meta-Analysis. Educational Testing Service, Princeton, 1982. 29. Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF, et al. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Lancet 1999;354:1896-900. 30. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MO- OSE) group. JAMA 2000;283:2008-12. 31. Cook DJ, Mulrow CD, Haynes RB. Systematic reviews: synthesis of best evidence for clinical decisions. Ann Intern Med 1997;126:376-80. http://www.e-hmr.org 39