Brede meta-analysis


DTU Informatics >>> Neuroinformatics > Services > Brede meta-analysis

Input

Brede Wiki CSV title: 


Column names Number Mean SD Events
Experimentals:
Controls:

Go to: Article | Data | CSV

Original data

First AuthorYearSubgroupNumber of PatientsNumber of ControlsPatient MeanPatient SDControl MeanControl SDPatient Age% Female PatientsAge of OnsetHam-D rating scaleAntidepressants %Mood stabilizers %Antipsychotics %Drug Free %ImagingMRI Field Strength (T)Slice Thickness (mm)PMID
Lacerda AL200431345.411.125.891.46077.427.9414.07100MRI1.51.514960287
Janssen J200428414.641.2651.210100.033.0461251421MRI1.51.215576058
Chen HH200827266.91.46.41.514.463.011.7520.1000100MRI1.51.519108659

Interpreted data and analysis

Study Experimentals Controls Effects
Mean SD Events N Mean SD Events NSDpooled Effect SE CI Z P Weight WeightDSL
Lacerda AL 2004 5.410000 1.120000 nan 31 5.890000 1.460000 nan 341.309155 -0.362 0.250 -0.853 0.129 -1.446 0.148 15.937655 35.073% 7.450257 34.234%
Janssen J 2004 4.640000 1.260000 nan 28 5.000000 1.210000 nan 411.230394 -0.289 0.246 -0.772 0.194 -1.174 0.240 16.461516 36.226% 7.562762 34.751%
Chen HH 2008 6.900000 1.400000 nan 27 6.400000 1.500000 nan 261.449882 0.340 0.277 -0.203 0.882 1.227 0.220 13.042051 28.701% 6.749728 31.015%
Meta-analysis (fixed effect) 86 101 -0.134 0.148 -0.425 0.156 -0.906 0.365 100%
Meta-analysis (random effects, DSL) -0.119 0.214 -0.539 0.301 -0.556 0.578 100%
σa: 0.267356 Q: 4.154726 dof: 2.0 P-value: 0.125

Forest plot

-2 -1 0 1 2 Lacerda AL 0.000000 Janssen J 1.000000 Chen HH 2.000000 Meta-analysis (fixed effect) Meta-analysis (random effects)

Funnel plot

Effect size Standard error -0.10 -0.20 -0.30 -0.40 0.10 0.20 0.30 0.40 0.08 0.16 0.24 0.32

Warning

Column not matched "number_of_experimental_events". Column not matched "number_of_control_events".

Meta-analysis

This webservice will download a specified page from the Brede Wiki. The page should contain a comma-separated values (CSV) file with information for a meta-analysis. Study effect sizes and meta-analytic effect sizes will be computed.

The CSV file should provide information about the two groups: "experimentals" and "controls". The information you need to supply depends on whether your data is categorical (binary) or continuous:

You will usually need to specify the column names used for the data, but an automated guess will be performed if columns are not specified.

Symbols

M: mean, SD: standard deviation of data, SDpooled: Pooled standard deviation of data, N: number of data (subjects, etc.), MD: mean difference, SMD: standardized mean difference, CI: 95% confidence interval.

Examples

Acknowledgment

Matthew Kempton. Lundbeck Foundation Center for Integrated Molecular Brain Imaging (CIMBI)
Finn Årup Nielsen, CIMBI, DTU Informatics.