- 1. mcforQ (mcforQ performs exact left-sided testing for Cochran's Q in a meta-analysis of binary data. Only the odds ratio has been implemented). For a Fortran implementation for testing left-sided significance in the risk ratio, see here.
Reference: Ioannidis, Trikalinos, and Zintzaras, J Clin Epidemiol 2006;59:1023-32.
- 2. Software modules for test for excess of significant findings (The zip contains all necessary files. Stata 8 is supported. Users should save the files in the adopath. Binaries for a
necessary plugin are provided for the Mac (OSX, PPC) and for Windows. Source files for the platform-specific plugin are also provided for users of other platforms/architectures.)
Reference: Ioannidis and Trikalinos, Clinical Trials, 2007;4(3):245-53.
- 3. Function for meta-analysis of epidemiological studies: inflation of study-specific confidence intervals according to a credibility ceiling. Script file for S-plus/R. Needs the subroutine ci.plot to be download from http://citeseer.ist.psu.edu/29387.html.
- 4. hetred module for STATA. Hetred reduces I-squared below a prespecified level. Plugin for Windows users is included. Mac and Linux users please contact author for source file.
- 5. Simple Excel calculator of Bayes factor and credibility estimates for associations. For details see Ioannidis JP. Effect of formal statistical significance on the credibility of observational associations. American Journal of Epidemiology 2008 (in press) and for application to agnostic genome-wide associations see Ioannidis JP. Calibration of credibility of agnostic genome-wide associations. American Journal of Medical Genetics B: Neuropsychiatric Genetics 2008 (in press). Download
- 6. PowerQ: STATA module
to perform power calculations for the Cochran Q test and calculate the posterior probabilities for heterogeneity in a given meta-analysis.
- 7. R routine for graphical evaluation of incoherence in a network of randomized trials.
by Georgia Salanti
- 8. Winbugs code used in the meta-analysis published in Am J Epi
'Underlying Genetic Models of Inheritance in Established Type 2
Diabetes Associations'. This is model to estimate meta-analytic
summary effects for genetic variants from case-control studies along
with a parameter reflecting the underlying genetic model. To obtain
the code click here.
Winbugs model and Data using in the analysis published in Journal of Clinical Epidemiology "Graphical methods and numerical summaries for presenting results from multiple-treatments meta-analysis: an overview and tutorial' by Georgia Salanti.
Evaluating novel agent effects in multiple-treatments meta-regression. Exchangeable novel agent effects across networks <click here for WinBUGS code> to model jointly novelty effects in ovarian, colorectal and breast cancer treatment networks <click here for the data>.
- 11.Stata module for estimating the
posterior probability of
heterogeneity based on Q test results (Pereira et al. Research Synthesis
Methods, in press)." (including help file)