Articles • forestsearch

Getting Started

Getting started with forestsearch

Methodology

Technical details on the statistical methods

Methodology

Post-Selection Inference

Inference for trial estimands after data-driven subgroup discovery

Interpretable subgroup identification with forestsearch
Post-selection inference for an identified survival subgroup
Post-selection bounds on a nested continuum: the Guo and He (2021) design
Binary and continuous outcomes: the ACTG 175 analyses

Treatment Effect Definitions

Definitions and computation of marginal HR, average hazard ratio (AHR), and controlled direct effect (CDE) estimands.

Treatment Effect Definitions in ForestSearch
Biomarker Treatment Effects in the Weibull AFT/Cox Framework
Causal Interpretation of Hazard Ratios in Survival Analysis

Simulation Studies

Operating characteristics and published benchmarks

setup_gbsg_dgm() and generate_aft_dgm_flex(): Relationship and Usage
Extreme Subgroup Effects Under a Uniform Treatment Benefit

MRCT consistency

Regional consistency in multiregional trials: a harmed subgroup identified outside the region and applied inside it, with the analytic layer and the simulation campaign.

MRCT regional consistency via subgroup identification
MRCT regional consistency: simulations

Reference

The package’s original vignette, kept as a reference for León et al. (2024)

Exploratory subgroup identification with forestsearch (the 2024 workflow)