Package: EvidenceSynthesis 0.5.0
Martijn Schuemie
EvidenceSynthesis: Synthesizing Causal Evidence in a Distributed Research Network
Routines for combining causal effect estimates and study diagnostics across multiple data sites in a distributed study, without sharing patient-level data. Allows for normal and non-normal approximations of the data-site likelihood of the effect parameter.
Authors:
EvidenceSynthesis_0.5.0.tar.gz
EvidenceSynthesis_0.5.0.tar.gz(r-4.5-noble)EvidenceSynthesis_0.5.0.tar.gz(r-4.4-noble)
EvidenceSynthesis_0.5.0.tgz(r-4.4-emscripten)EvidenceSynthesis_0.5.0.tgz(r-4.3-emscripten)
EvidenceSynthesis.pdf |EvidenceSynthesis.html✨
EvidenceSynthesis/json (API)
NEWS
| # Install 'EvidenceSynthesis' in R: |
| install.packages('EvidenceSynthesis', repos = 'https://cloud.r-project.org') |
Bug tracker:https://github.com/ohdsi/evidencesynthesis/issues3 issues
Pkgdown site:https://ohdsi.github.io
- ncLikelihoods - Example profile likelihoods for negative control outcomes
- ooiLikelihoods - Example profile likelihoods for a synthetic outcome of interest
Last updated 2 years agofrom:4b1759c186. Checks:1 OK, 2 WARNING. Indexed: no.
| Target | Result | Latest binary |
|---|---|---|
| Doc / Vignettes | OK | Mar 27 2025 |
| R-4.5-linux | WARNING | Mar 27 2025 |
| R-4.4-linux | WARNING | Mar 27 2025 |
Exports:approximateHierarchicalNormalPosteriorapproximateLikelihoodapproximateSimplePosteriorbiasCorrectionInferencecomputeBayesianMetaAnalysiscomputeConfidenceIntervalcomputeFixedEffectMetaAnalysiscreateSimulationSettingscustomFunctiondetectApproximationTypefitBiasDistributionplotBiasCorrectionInferenceplotBiasDistributionplotCovariateBalancesplotEmpiricalNullsplotLikelihoodFitplotMcmcTraceplotMetaAnalysisForestplotPerDbMcmcTraceplotPerDbPosteriorplotPosteriorplotPreparedPspreparePsPlotsequentialFitBiasDistributionsimulatePopulationsskewNormalsupportsJava8
Dependencies:AndromedaBeastJarbitbit64blobbootcachemclicliprcodacolorspaceCompQuadFormcpp11crayonCyclopsDBIdbplyrdigestdistributionaldplyrEmpiricalCalibrationfansifarverfastmapgenericsggdistggplot2gluegridExtragtableHDIntervalhmsisobandlabelinglatticelifecyclelme4magrittrMASSmathjaxrMatrixmemoisemetametadatmetaformgcvminqamunsellnlmenloptrnumDerivpbapplypillarpkgconfigplogrprettyunitsprogresspurrrquadprogR6rbibutilsRColorBrewerRcppRcppEigenRcppParallelRdpackreadrreformulasrJavarlangRSQLitescalesstringistringrsurvivaltibbletidyrtidyselecttzdbutf8vctrsviridisLitevroomwithrxml2zip
Bayesian adaptive bias correction using profile likelihoods
Rendered fromBayesianBiasCorrection.Rmdusingknitr::rmarkdownon Mar 27 2025.Last update: 2023-05-08
Started: 2023-05-08
Code used in the video vignette
Rendered fromVideoVignette.Rmdusingknitr::rmarkdownon Mar 27 2025.Last update: 2023-05-08
Started: 2023-04-05
Effect estimate synthesis using non-normal likelihood approximations
Rendered fromNonNormalEffectSynthesis.Rmdusingknitr::rmarkdownon Mar 27 2025.Last update: 2023-05-08
Started: 2020-11-19
Citation
To cite package ‘EvidenceSynthesis’ in publications use:
Schuemie M, Suchard M, Bu F (2023). EvidenceSynthesis: Synthesizing Causal Evidence in a Distributed Research Network. R package version 0.5.0, https://CRAN.R-project.org/package=EvidenceSynthesis.
Corresponding BibTeX entry:
@Manual{,
title = {EvidenceSynthesis: Synthesizing Causal Evidence in a
Distributed Research Network},
author = {Martijn Schuemie and Marc A. Suchard and Fan Bu},
year = {2023},
note = {R package version 0.5.0},
url = {https://CRAN.R-project.org/package=EvidenceSynthesis},
}
Readme and manuals
EvidenceSynthesis
EvidenceSynthesis is part of HADES.
Introduction
This R package contains routines for combining causal effect estimates and study diagnostics across multiple data sites in a distributed study. This includes functions for performing meta-analysis and forest plots.
Features
- Perform a traditional fixed-effects or random-effects meta-analysis, and create a forest plot.
- Use non-normal approximations of the per-data-site likelihood function to avoid bias when facing small and zero counts.
Example
# Simulate some data for this example:
populations <- simulatePopulations()
# Fit a Cox regression at each data site, and approximate likelihood function:
fitModelInDatabase <- function(population) {
cyclopsData <- Cyclops::createCyclopsData(Surv(time, y) ~ x + strata(stratumId),
data = population,
modelType = "cox")
cyclopsFit <- Cyclops::fitCyclopsModel(cyclopsData)
approximation <- approximateLikelihood(cyclopsFit, parameter = "x", approximation = "custom")
return(approximation)
}
approximations <- lapply(populations, fitModelInDatabase)
approximations <- do.call("rbind", approximations)
# At study coordinating center, perform meta-analysis using per-site approximations:
estimate <- computeBayesianMetaAnalysis(approximations)
estimate
# mu mu95Lb mu95Ub muSe tau tau95Lb tau95Ub logRr seLogRr
# 1 0.5770562 -0.2451619 1.382396 0.4154986 0.2733942 0.004919128 0.7913512 0.5770562 0.4152011
Technology
This an R package with some parts implemented in Java.
System requirements
Requires R and Java.
Getting Started
-
Make sure your R environment is properly configured. This means that Java must be installed. See these instructions for how to configure your R environment.
-
In R, use the following commands to download and install EvidenceSynthesis:
install.packages("EvidenceSynthesis")
User Documentation
Documentation can be found on the package website.
PDF versions of the documentation are also available:
- Vignette: Effect estimate using non-normal likelihood approximations
- Package manual: EvidenceSynthesis.pdf
Support
- Developer questions/comments/feedback: OHDSI Forum
- We use the GitHub issue tracker for all bugs/issues/enhancements
Contributing
Read here how you can contribute to this package.
License
EvidenceSynthesis is licensed under Apache License 2.0
Development
This package is being developed in RStudio.
Development status
Beta
