Sequential Multiple Assignment Randomized Trials Design and Analyses

Implements the interim augmented inverse probability weighted estimator (IAIPWE) for estimating the value of treatment regimes embedded in sequential multiple assignment randomized trials (SMARTs). The IAIPWE subsumes both the inverse probability weighted estimator (IPWE) and the augmented inverse probability weighted estimator (AIPWE), enabling inference at interim and final analyses. The package provides functions for value estimation, sandwich variance computation, group sequential stopping boundaries, and sample size determination for multi-stage SMARTs with up to two treatment options at each stage. See Manschot, Laber, and Davidian (2023) for additional details.

rsmart

rsmart is an R package to plan and analyze sequential multiple assignment randomized trials (SMARTs)

Installation

install.packages("rsmart")

Alternatively, to use a new feature or get a bug fix, you can install the development version of rsmart from GitHub:

# install.packages("remotes")
remotes::install_github("MSDLLCpapers/rsmart")

Overview

The rsmart package is intended to provide a common package for analyzing SMARTs with normal, binary, and time-to-event endpoints. The current build implements the inverse-probability weighted estimator, augmented IPWE and interim AIPWE of Wu, Wang, and Wahed, Zhang, Tsiatis, Laber and Davidian, and Manschot, Laber, and Davidian. It provides a convenient structure to implement any of these estimators for continuous (asymptotically normal) endpoints.

If you're new to SMARTs and want to read a high-level overview of the key concepts involved, see our vignette titled "Getting Started with SMARTs."

If you want to see an example of how the rsmart package can be used to implement the inverse probability weighted estimator (IAIPWE) approach, see the vignette titled "Full demo of IAIPWE."