R implementation of a multivariate meta-analysis of randomized controlled trials (RCT) with the difference in restricted mean survival times (RMSTD). Use this package with individual patient level data from an RCT for a time-to-event outcome to determine combined effect estimates according to 4 methods: 1) a univariate meta-analysis using observed treatment effects, 2) a univariate meta-analysis using effects predicted by fitted Royston-Parmar flexible parametric models, 3) multivariate meta-analysis with analytically derived covariance, 4) multivariate meta-analysis with bootstrap derived covariance. This package computes all combined effects and provides an RMSTD curve with combined effect estimates and their confidence intervals.
An R package for meta-analysis of RMSTD
install.packages("devtools") library(devtools) devtools::install_github("iweir/metaRMST") library(metaRMST)
# load library library(metaRMST) # read in built-in dataset data(AorticStenosisTrials) # demonstration of meta-analysis to obtain combined effect by multivariate meta-analysis model (method="mvma") mvma_res <- metaRMSTD(AorticStenosisTrials, time_horizons=c(12,24,36), MA_method="mvma") mvma_res$REresult # generate figure: obj <- RMSTcurves(AorticStenosisTrials, time_horizons=c(12,24,36), tmax=40, nboot=50) RMSTplot(obj, xlim=c(0,40), ylim=c(-0.25,2.75), yby=0.5, ylab="RMSTD (mos)", xlab="Time (mos)")