More Than Numbers: The Power of Graphs in Meta-Analysis

被引:131
作者
Bax, Leon [1 ,2 ,5 ]
Ikeda, Noriaki [2 ]
Fukui, Naohito [3 ]
Yaju, Yukari [4 ]
Tsuruta, Harukazu [2 ]
Moons, Karel G. M. [5 ]
机构
[1] Kitasato Univ, Kitasato Clin Res Ctr, Sagamihara, Kanagawa 2288555, Japan
[2] Kitasato Univ, Dept Med Informat, Sagamihara, Kanagawa 2288555, Japan
[3] Japan Clin Res Support Unit, Tokyo, Japan
[4] Univ Tokyo, Sch Hlth Sci & Nursing, Dept Epidemiol & Biostat, Tokyo, Japan
[5] Univ Med Ctr Utrecht, Julius Ctr Hlth Sci & Primary Care, Utrecht, Netherlands
关键词
PUBLICATION BIAS; FUNNEL-PLOT; EXPLORING HETEROGENEITY; FILL METHOD; TRIM;
D O I
10.1093/aje/kwn340
中图分类号
R1 [预防医学、卫生学];
学科分类号
1004 ; 120402 ;
摘要
In meta-analysis, the assessment of graphs is widely used in an attempt to identify or rule out heterogeneity and publication bias. A variety of graphs are available for this purpose. To date, however, there has been no comparative evaluation of the performance of these graphs. With the objective of assessing the reproducibility and validity of graph ratings, the authors simulated 100 meta-analyses from 4 scenarios that covered situations with and without heterogeneity and publication bias. From each meta-analysis, the authors produced 11 types of graphs (box plot, weighted box plot, standardized residual histogram, normal quantile plot, forest plot, 3 kinds of funnel plots, trim-and-fill plot, Galbraith plot, and L'Abbe plot), and 3 reviewers assessed the resulting 1,100 plots. The intraclass correlation coefficients (ICCs) for reproducibility of the graph ratings ranged from poor (ICC = 0.34) to high (ICC = 0.91). Ratings of the forest plot and the standardized residual histogram were best associated with parameter heterogeneity. Association between graph ratings and publication bias (censorship of studies) was poor. Meta-analysts should be selective in the graphs they choose for the exploration of their data.
引用
收藏
页码:249 / 255
页数:7
相关论文
共 43 条
[1]  
[Anonymous], MIX COMPREHENSIVE FR
[2]  
[Anonymous], SUMMING
[3]  
[Anonymous], 1982, PRIM CARDIOL
[4]   A graphical method for exploring heterogeneity in meta-analyses:: application to a meta-analysis of 65 trials [J].
Baujat, B ;
Mahé, C ;
Pignon, JP ;
Hill, C .
STATISTICS IN MEDICINE, 2002, 21 (18) :2641-2652
[5]   A systematic comparison of software dedicated to meta-analysis of causal studies [J].
Bax, Leon ;
Yu, Ly-Mee ;
Ikeda, Noriaki ;
Moons, Karel G. M. .
BMC MEDICAL RESEARCH METHODOLOGY, 2007, 7 (1)
[6]   OPERATING CHARACTERISTICS OF A BANK CORRELATION TEST FOR PUBLICATION BIAS [J].
BEGG, CB ;
MAZUMDAR, M .
BIOMETRICS, 1994, 50 (04) :1088-1101
[7]   EFFICACY OF BCG VACCINE IN THE PREVENTION OF TUBERCULOSIS - METAANALYSIS OF THE PUBLISHED LITERATURE [J].
COLDITZ, GA ;
BREWER, TF ;
BERKEY, CS ;
WILSON, ME ;
BURDICK, E ;
FINEBERG, HV ;
MOSTELLER, F .
JAMA-JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION, 1994, 271 (09) :698-702
[8]   A nonparametric "trim and fill" method of accounting for publication bias in meta-analysis [J].
Duval, S ;
Tweedie, R .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 2000, 95 (449) :89-98
[9]   Trim and fill: A simple funnel-plot-based method of testing and adjusting for publication bias in meta-analysis [J].
Duval, S ;
Tweedie, R .
BIOMETRICS, 2000, 56 (02) :455-463
[10]   Bias in meta-analysis detected by a simple, graphical test [J].
Egger, M ;
Smith, GD ;
Schneider, M ;
Minder, C .
BMJ-BRITISH MEDICAL JOURNAL, 1997, 315 (7109) :629-634