More than numbers: the power of graphs in meta-analysis.
L. Bax, N. Ikeda, N. Fukui, Y. Yaju, H. Tsuruta, and K. Moons. American journal of epidemiology, 169 (2):
249-55(January 2009)5062<m:linebreak></m:linebreak>JID: 7910653; 2008/12/08 aheadofprint; ppublish;<m:linebreak></m:linebreak>Metaanàlisi; Presentació de dades.
DOI: 10.1093/aje/kwn340
Abstract
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'Abbé 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.
%0 Journal Article
%1 Bax2009
%A Bax, Leon
%A Ikeda, Noriaki
%A Fukui, Naohito
%A Yaju, Yukari
%A Tsuruta, Harukazu
%A Moons, Karel G M
%D 2009
%J American journal of epidemiology
%K Communication ComputerGraphics ConfidenceIntervals DataInterpretation EffectModifier Epidemiologic EpidemiologicMethods Humans Journalism Medical Meta-AnalysisasTopic PublicationBias Software Statistical
%N 2
%P 249-55
%R 10.1093/aje/kwn340
%T More than numbers: the power of graphs in meta-analysis.
%U http://www.ncbi.nlm.nih.gov/pubmed/19064649
%V 169
%X 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'Abbé 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.
%@ 1476-6256
@article{Bax2009,
abstract = {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'Abbé 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.},
added-at = {2023-02-03T11:44:35.000+0100},
author = {Bax, Leon and Ikeda, Noriaki and Fukui, Naohito and Yaju, Yukari and Tsuruta, Harukazu and Moons, Karel G M},
biburl = {https://www.bibsonomy.org/bibtex/26be4d6ec3272e7ace1859766edfa376f/jepcastel},
city = {Kitasato Clinical Research Center, Kitasato University, Sagamihara, Kanagawa, Japan. leonbax@kitasato-crc.org},
doi = {10.1093/aje/kwn340},
interhash = {ccd4fb7989398a6a02ec3e07ba4217f2},
intrahash = {6be4d6ec3272e7ace1859766edfa376f},
isbn = {1476-6256},
issn = {1476-6256},
journal = {American journal of epidemiology},
keywords = {Communication ComputerGraphics ConfidenceIntervals DataInterpretation EffectModifier Epidemiologic EpidemiologicMethods Humans Journalism Medical Meta-AnalysisasTopic PublicationBias Software Statistical},
month = {1},
note = {5062<m:linebreak></m:linebreak>JID: 7910653; 2008/12/08 [aheadofprint]; ppublish;<m:linebreak></m:linebreak>Metaanàlisi; Presentació de dades},
number = 2,
pages = {249-55},
pmid = {19064649},
timestamp = {2023-02-03T11:44:35.000+0100},
title = {More than numbers: the power of graphs in meta-analysis.},
url = {http://www.ncbi.nlm.nih.gov/pubmed/19064649},
volume = 169,
year = 2009
}