Package: DRsurvCRT
Type: Package
Title: Doubly-Robust Estimation for Survival Outcomes in
        Cluster-Randomized Trials
Version: 0.0.1
Authors@R: c(
    person(given = "Xi", family = "Fang", role = c("aut","cre"),email = "x.fang@yale.edu"),
    person(given = "Fan", family = "Li", role = c("aut"), email = "fan.li@example.edu")
  )
Description: Cluster-randomized trials (CRTs) assign treatment to groups rather than individuals, so valid analyses must distinguish cluster-level and individual-level effects and define estimands within a potential-outcomes framework. This package supports right-censored survival outcomes for both single-state (binary) and multi-state settings. For single-state outcomes, it provides estimands based on stage-specific survival contrasts (SPCE) and restricted mean survival time (RMST). For multi-state outcomes, it provides SPCE as well as a generalized win-based restricted mean time-in-favor estimand (RMT-IF). The package implements doubly robust estimators that accommodate covariate-dependent censoring and remain consistent if either the outcome model or the censoring model is correctly specified. Users can choose marginal Cox or gamma-frailty Cox working models for nuisance estimation, and inference is supported via leave-one-cluster-out jackknife variance and confidence interval estimation. Methods are described in Fang et al. (2025) "Estimands and doubly robust estimation for cluster-randomized trials with survival outcomes" <doi:10.48550/arXiv.2510.08438>.
License: MIT + file LICENSE
Encoding: UTF-8
LazyData: true
Imports: Rcpp, frailtyEM, survival, ggplot2, pracma, abind
LinkingTo: Rcpp, RcppArmadillo
RoxygenNote: 7.3.3
Depends: R (>= 3.5)
NeedsCompilation: yes
Packaged: 2025-12-19 03:14:10 UTC; fancy
Author: Xi Fang [aut, cre],
  Fan Li [aut]
Maintainer: Xi Fang <x.fang@yale.edu>
Repository: CRAN
Date/Publication: 2025-12-30 19:00:06 UTC
Built: R 4.4.3; x86_64-w64-mingw32; 2025-12-31 00:50:26 UTC; windows
Archs: x64
