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 = "grid with gradients")
  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

1. 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.

2. 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