Dear All? I am learning how to make a meta analysis of binary outcome data. So I chose the Fleiss93 dataset available in the meta package. This is my code: library(meta) data(Fleiss93) MA <- metabin(event.e, n.e, event.c, n.c, data=Fleiss93, studlab=paste(study, year), method = "Inverse",sm="OR") forest(MA) funnel(MA, studlab = TRUE) metabias(res, method="linreg",k.min=7) The funnel plot looks like asymmetric but when I use the function metabias to test this asymmetry I found a p-value=0.399 > 5% indicating that the hypothesis H0 should be accepted 'the funnel plot is symmetric' ??? For me the result is contradictory to what I expect. I misinterpreted the result or?I did not choose the correct method to put in metabias function??? Thank you in advance for your help/ explanations ! Best regards. Safa Aouinti Ph.D. in applied Mathematics and Statistics
[R-meta] How to interpret the result of metabias function of the R package meta?
3 messages · Safa Aouinti, Gerta Ruecker, Guido Schwarzer
Dear Safa, First, the last line of your code should probably read metabias(MA, method="linreg",k.min=7) The reason why the test does not indicate significant asymmetry is probably that there are only 7 studies. Recommendations say that one should test for funnel plot asymmetry only if there are at least 10 studies (Sterne JAC et al. , Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials, BMJ 343, July 22 2011). Best, Gerta Am 02.03.2021 um 16:00 schrieb Safa Aouinti:
Dear All I am learning how to make a meta analysis of binary outcome data. So I chose the Fleiss93 dataset available in the meta package. This is my code: library(meta) data(Fleiss93) MA <- metabin(event.e, n.e, event.c, n.c, data=Fleiss93, studlab=paste(study, year), method = "Inverse",sm="OR") forest(MA) funnel(MA, studlab = TRUE) metabias(res, method="linreg",k.min=7) The funnel plot looks like asymmetric but when I use the function metabias to test this asymmetry I found a p-value=0.399 > 5% indicating that the hypothesis H0 should be accepted 'the funnel plot is symmetric' ??? For me the result is contradictory to what I expect. I misinterpreted the result or?I did not choose the correct method to put in metabias function??? Thank you in advance for your help/ explanations ! Best regards. Safa Aouinti Ph.D. in applied Mathematics and Statistics
_______________________________________________ R-sig-meta-analysis mailing list R-sig-meta-analysis at r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
Dr. rer. nat. Gerta R?cker, Dipl.-Math. Institute of Medical Biometry and Statistics, Faculty of Medicine and Medical Center - University of Freiburg Stefan-Meier-Str. 26, D-79104 Freiburg, Germany Phone: +49/761/203-6673 Fax: +49/761/203-6680 Mail: ruecker at imbi.uni-freiburg.de Homepage: https://www.uniklinik-freiburg.de/imbi-en/employees.html?imbiuser=ruecker
Am 02.03.21 um 16:00 schrieb Safa Aouinti:
[...] metabias(res, method="linreg",k.min=7)
Safa, In addition to Gerta's comments I would like to point out that the classic Egger test is used if argument method = "linreg" (or method = "Egger" from meta, version 4.17-0). It is recommended to use a modification of the classic Egger test for binary outcomes as the log odds ratio and its standard error are correlated. The tests by Harbord, Peters and R?cker are all available in metabias() - see the help file for more information and references. Actually, by default, i.e., for the command metabias(res, k.min = 7), the test by Harbord is used for the odds ratio as effect measure. Best regards Guido