Introduction of CSTE (ver 0.9.0)
CSTE provides a uniform statistical inferential tool for estimating individualized treatment rule.
CSTE estimates differences of average outcomes between different treatment groups conditional on patients characteristics and provides the corresponding simultaneous confidence bands. Based on the CSTE curve and the simultaneous confidence bands, one can decide the subgroups of patients that benefit from each treatment and then make individualized treatment selections.
The current CSTE can analyze the following types of datasets:
-
Datasets with binary outcomes, 2-arm treatments, and single or multiple (even high-dimensional) covariates.
-
Datasets with right censored time-to-event outcomes, multi-arm treatments, and a single covariate of the biomarker.
CSTE will consider more types of datasets in the future.
R package
CSTE: Covariate Specific Treatment Effect (CSTE) Curve (https://CRAN.R-project.org/package=CSTE)
