README.md

In CausalMetaR: Causally Interpretable Meta-Analysis

CausalMetaR

The CausalMetaR package provides robust and efficient methods for estimating causal effects in a target population using a multi-source dataset. The multi-source data can be a collection of trials, observational studies, or a combination of both, which have the same data structure (outcome, treatment, and covariates). The target population can be based on an internal dataset or an external dataset where only covariate information is available. The causal estimands available are average treatment effects (ATEs) and subgroup treatment effects (STEs).

Installation

You can install the released version of CausalMetaR from CRAN with:

install.packages("CausalMetaR")

You can install the development version of CausalMetaR from GitHub with:

# install.packages("devtools")
devtools::install_github("ly129/CausalMetaR")

Usage

Refer to Wang et al. (2025) for a detailed guide on using the package. Below, we include the code used in the examples in the paper. The examples are based on the multi-source dataset dat_multisource included in the package.

Loading the package:

library(CausalMetaR)

Specifying the working models:

outcome_model_args <- list(family = gaussian(),
                           SL.library = c("SL.glmnet", "SL.nnet", "SL.glm"))
treatment_model_args <- list(family = binomial(),
                             SL.library = c("SL.glmnet", "SL.nnet", "SL.glm"))
external_model_args = list(family = binomial(),
                           SL.library = c("SL.glmnet", "SL.nnet", "SL.glm"))

Setting a random number seed for reproducibility:

set.seed(1234)

The examples below estimate ATEs and STEs in external and internal target populations. Each example may take a few minutes to run on standard laptop.

Example 1: Estimating the ATE in the external target population

result_ae <- ATE_external(
  Y = dat_multisource$Y,
  S = dat_multisource$S,
  A = dat_multisource$A,
  X = dat_multisource[, 1:10],
  X_external = dat_external[, 1:10],
  outcome_model_args = outcome_model_args,
  treatment_model_args = treatment_model_args,
  external_model_args = external_model_args,
  cross_fitting = TRUE,
  replications = 5)
#> Loading required package: nnls
#> Loading required namespace: nnet
result_ae
#> AVERAGE TREATMENT EFFECT ESTIMATES IN AN EXTERNAL POPULATION
#> 
#> Treatment effect (mean difference) estimates:
#> ---------------------------------------------
#>  Estimate     SE Lower 95% CI Upper 95% CI
#>    6.6294 0.1535       5.8616       7.3972

Example 2: Estimating ATEs in the internal target populations

result_ai <- ATE_internal(
  Y = dat_multisource$Y,
  S = dat_multisource$S,
  A = dat_multisource$A,
  X = dat_multisource[, 1:10],
  outcome_model_args = outcome_model_args,
  treatment_model_args = treatment_model_args,
  cross_fitting = TRUE,
  replications = 5)
result_ai
#> AVERAGE TREATMENT EFFECT ESTIMATES IN INTERNAL POPULATIONS
#> 
#> Treatment effect (mean difference) estimates:
#> ---------------------------------------------
#>  Source Estimate     SE Lower 95% CI Upper 95% CI
#>       A   6.5874 0.1903       6.2145       6.9603
#>       B   7.7556 0.2577       7.2506       8.2606
#>       C   7.2916 0.3594       6.5872       7.9960

Example 3: Estimating STEs in the external target population

result_se <- STE_external(
  Y = dat_multisource$Y,
  S = dat_multisource$S,
  A = dat_multisource$A,
  X = dat_multisource[, 2:10],
  EM = dat_multisource$EM,
  X_external = dat_external[, 2:10],
  EM_external = dat_external$EM,
  outcome_model_args = outcome_model_args,
  treatment_model_args = treatment_model_args,
  external_model_args = external_model_args,
  cross_fitting = TRUE,
  replications = 5)
result_se
#> SUBGROUP TREATMENT EFFECT ESTIMATES IN AN EXTERNAL POPULATION
#> 
#> Treatment effect (mean difference) estimates:
#> ---------------------------------------------
#>  Subgroup Estimate     SE Lower 95% CI Upper 95% CI Lower 95% SCB Upper 95% SCB
#>         a   7.0787 0.3563       5.9088       8.2485        5.5453        8.6121
#>         b   5.5207 0.2321       4.5764       6.4650        4.2830        6.7585
#>         c   7.5709 0.1805       6.7382       8.4037        6.4794        8.6625
#>         d   6.5748 0.2253       5.6446       7.5051        5.3556        7.7941
#>         e   5.3741 0.3382       4.2343       6.5139        3.8802        6.8681

Example 4: Estimating STEs in the internal target populations

result_si <- STE_internal(
  Y = dat_multisource$Y,
  S = dat_multisource$S,
  A = dat_multisource$A,
  X = dat_multisource[, 2:10],
  EM = dat_multisource$EM,
  outcome_model_args = outcome_model_args,
  treatment_model_args = treatment_model_args,
  cross_fitting = TRUE,
  replications = 5)
result_si
#> SUBGROUP TREATMENT EFFECT ESTIMATES IN INTERNAL POPULATIONS
#> 
#> Treatment effect (mean difference) estimates:
#> ---------------------------------------------
#>  Source Subgroup Estimate     SE Lower 95% CI Upper 95% CI Lower 95% SCB
#>       A        a   6.9197 0.5001       5.9395       7.8999        5.6345
#>                b   5.4340 0.3681       4.7126       6.1555        4.4880
#>                c   7.5452 0.3097       6.9383       8.1522        6.7493
#>                d   6.5053 0.3630       5.7939       7.2168        5.5724
#>                e   5.4595 0.5215       4.4373       6.4816        4.1192
#>       B        a   8.2134 0.7554       6.7328       9.6939        6.2720
#>                b   6.8396 0.5381       5.7850       7.8942        5.4567
#>                c   8.7995 0.4232       7.9699       9.6290        7.7118
#>                d   7.7098 0.4529       6.8223       8.5974        6.5460
#>                e   6.4405 0.6975       5.0734       7.8076        4.6479
#>       C        a   8.0544 1.2049       5.6929      10.4159        4.9579
#>                b   6.2151 0.6711       4.8998       7.5304        4.4904
#>                c   8.2620 0.6426       7.0026       9.5215        6.6106
#>                d   7.1427 0.6538       5.8612       8.4242        5.4624
#>                e   6.0910 0.8718       4.3823       7.7997        3.8504
#>  Upper 95% SCB
#>         8.2050
#>         6.3801
#>         8.3411
#>         7.4382
#>         6.7998
#>        10.1547
#>         8.2225
#>         9.8872
#>         8.8737
#>         8.2332
#>        11.1509
#>         7.9398
#>         9.9135
#>         8.8231
#>         8.3316

Reference

To cite this package, use:

Wang G, McGrath S, Lian Y. CausalMetaR: An R package for performing causally interpretable meta-analyses. Research Synthesis Methods. In press.

CausalMetaR documentation built on April 12, 2025, 2:06 a.m.