An open API service indexing awesome lists of open source software.

https://github.com/boehringer-ingelheim/bprinstrattte

Causal Effects in Principal Strata Defined by Antidrug Antibodies
https://github.com/boehringer-ingelheim/bprinstrattte

bayesian-methods causal-inference clinical-trial estimand mcmc-methods pharmaceutical-development principal-stratification r-package simulation stan time-to-event

Last synced: 13 days ago
JSON representation

Causal Effects in Principal Strata Defined by Antidrug Antibodies

Awesome Lists containing this project

README

          

---
output: github_document
bibliography: "inst/references.bib"
csl: "inst/asa.csl"
---

```{r, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#",
fig.path = "man/figures/README-",
out.width = "100%"
)
```

# BPrinStratTTE BPrinStratTTE website

[![](https://img.shields.io/badge/lifecycle-experimental-orange.svg)](https://lifecycle.r-lib.org/articles/stages.html#experimental)
[![CRAN status](https://www.r-pkg.org/badges/version/BPrinStratTTE)](https://CRAN.R-project.org/package=BPrinStratTTE)

Bayesian models to estimate causal effects of biological treatments on time-to-event endpoints in clinical trials with principal strata defined by the occurrence of antidrug antibodies.

## Scope

- The package contains functions to fit Bayesian principal stratification models and to perform clinical trial simulations to determine operating characteristics for given scenarios.
- Two-arm clinical trials of biological therapies are considered
- with an intercurrent event (determining the principal stratum of interest) that can only occur in the treated arm (such as the development of antidrug antibodies), and
- with a time-to-event endpoint that is assumed to follow an exponential distribution.
- Effect estimators are hazard ratios and restricted mean survival times.
- Potential predictors of the intercurrent event can be taken into account.
- The models are fitted by Monte Carlo Markov Chain (MCMC) sampling, they are coded in [Stan](https://mc-stan.org/) and precompiled.
- More flexible time-to-event distributions (piecewise-exponential and Weibull) will be considered in future versions of the package.

## Principal stratification methodology

- Principal stratification is an approach to estimate causal effects in partitions of subjects determined by post-treatment events. It was introduced in the biostatistical literature by @Frangakis2002.
- The ICH E9 (R1) addendum on estimands and sensitivity analysis in clinical trials proposed principal stratification as one approach to deal with intercurrent events in clinical trials (@ICHE9R1Guideline).
- Principal stratum membership is typically not known with certainty. A Bayesian approach may be particularly suited to deal with this type of uncertainty. Following a proposal by @Imbens1997, principal stratum membership can be treated as a latent mixture variable.
- Motivated by scientific questions arising in clinical trials of biological therapies, in this package the approach by @Imbens1997 is adapted to a specific clinical trial setting with a time-to-event endpoint and the intercurrent event only occurring in the treated group.
- For recent reviews of applications to clinical trials see @Lipkovich2022 and @Bornkamp2021.

References:

## Installation

The current stable version of the package can be installed from CRAN with:

```{r cran-installation, eval=FALSE}
install.packages("BPrinStratTTE")
```

The development version of the package can be installed from GitHub with:

```{r gh-installation, eval=FALSE}
if (!require("remotes")) {install.packages("remotes")}
remotes::install_github("Boehringer-Ingelheim/BPrinStratTTE")
```