tvmedg

The tvmedg package implements causal mediation analysis using g-computation in longitudinal settings with time-varying exposure, mediators, and confounders. It extends the g-computation framework to decompose total effects into the randomized interventional analogues of natural direct and indirect effects, in the presence of time-varying confounding affected by prior exposures or mediators. This approach builds upon the ‘mediational g-formula’ introduced by VanderWeele and Tchetgen Tchetgen (2017).

The current version of tvmedg supports multiple mediators, both binary and continuous exposures, and spline-based functional forms for continuous variables. The package also enables parallel computing for efficiency in large-scale analyses. Besides the core modeling functions, tvmedg includes functions for visualizing and diagnosing model results.

Installation

The development version of tvmedg can be installed from GitHub:

# install.packages("devtools")
devtools::install_github("causalepi/tvmedg")

Quick Start

library(tvmedg)
library(doParallel)
#> Loading required package: foreach
#> Loading required package: iterators
#> Loading required package: parallel

Simulation data

head(sim_data)
#>   id mm Ap Mp L1       L2        L3 Yp      age sex ow risk lastid
#> 1  1  1  0  0  0 100.0000  80.00000  0 16.52949   0  0    0      0
#> 2  1  2  0  0  0 125.1296 102.02885  0 16.52949   0  0    0      0
#> 3  1  3  0  0  0 116.4990  98.99688  0 16.52949   0  0    0      0
#> 4  1  4  0  0  0 131.9247 104.07117  0 16.52949   0  0    0      0
#> 5  1  5  0  0  0 109.5959 103.24813  0 16.52949   0  0    0      0
#> 6  1  6  0  0  0 124.2403  95.82793  0 16.52949   0  0    0      0

Run model

cl <- makeCluster(8)
registerDoParallel(cl)

op <- tvmedg(data = sim_data,
       id = "id",
       basec = c("age","sex","ow","risk"),
       expo = c("Ap"),
       med = c("Mp"),
       tvar = c("L1","L2","L3"),
       outc = c("Yp"),
       time = c("mm"),
       lag = 2,
       cont_exp = F,
       mreg = "binomial",
       lreg = c("binomial","gaussian","gaussian"),
       yreg = "binomial",
       sp_list = c("mm"),
       sp_type = c("bs"),
       sp_df = c(3),
       followup = 12,
       seed = 123,
       montecarlo = 1000,
       boot = F,
       parallel = TRUE)
#> Q(a,a): 0.188 
#> Q(a,a*): 0.037 
#> Q(a*,a*): 0.008 
#> Indirect (rIE): 0.151 
#> Direct (rDE): 0.029 
#> Total (rTE): 0.18 
#> Proportional explain: 0.839 
#> Total time elapsed: 5.774524 mins

stopCluster(cl)