synthesizeResults: Conducts a meta-analysis across PLE result sets

In OHDSI/SkeletonComparativeEffectStudy: A Package Skeleton for Comparative Effectiveness Studies

Conducts a meta-analysis across PLE result sets

Description

Conducts a meta-analysis across PLE result sets

Usage

synthesizeResults(
  allDbsFolder,
  maExportFolder = allDbsFolder,
  maxCores = 1,
  method = "BayesianNonNormal",
  resultsZipPattern = "^Results_.*\\.zip",
  addTraditional = TRUE
)

Arguments

allDbsFolder	Folder on the local file system containing the individual zip files across databases (i.e., sites)
maExportFolder	A local folder where the meta-analysis results will be written. If not specified, results will be written to same directory with all other results.
maxCores	Maximum number of CPU cores to be used when computing the meta-analyses.
method	The meta-analysis method to use. Possible values are "BayesianNonNormal" (Schumie et al.) or "DL" (DerSimonian-Laird).
resultsZipPattern	The pattern of the names of the zip files containing the exported results of each database.
addTraditional	Boolean indicating if traditional meta-analysis (i.e., "DL") results should be added to result (if method is "BayesianNonNormal").

Details

Conducts a meta-analysis across result sets generated from a population level effect (PLE) study package. Meta-analysis methodology is based on DerSimonian and Laird (1986) or Schuemie et al. (2021).

Value

Does not return a value, but creates a new zip file in the maExportFolder for the meta-analyses.

References

DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986 Sep;7(3):177-88. doi: 10.1016/0197-2456(86)90046-2

Schuemie M, Chen Y, Madigan D, Suchard M, Combining Cox Regressions Across a Heterogeneous Distributed Research Network Facing Small and Zero Counts. arXiv: 2101.01551, 2021

OHDSI/SkeletonComparativeEffectStudy documentation built on May 13, 2023, 12:48 p.m.