The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

Binary Endpoint Estimation with Covariate Adjustment

Lifecycle: stable CRAN status R-CMD-check test-coverage

The goal of beeca is to provide an implementation solution with a simple user interface to estimate marginal estimands in a binary endpoint setting with covariate adjustment. The primary aim of this lightweight implementation is to facilitate quick industry adoption and use within GxP environments. A secondary aim is to support the simulation studies included in the manuscript Magirr et al. (2024).

Installation

Type Source Command
Release CRAN install.packages("beeca")
Development GitHub remotes::install_github("openpharma/beeca")

Methodology

Motivated by the recent FDA guidance (2023) on “Adjusting for Covariates in Randomized Clinical Trials for Drugs and Biological Products Guidance for Industry”” and its recommendations on robust variance estimation, we implemented two approaches, namely Ge et al. (2011) and Ye et al. (2023), to perform variance estimation for a treatment effect estimator based on g-computation under a logistic regression working model.

Scope

The package is designed to estimate marginal (unconditional) estimands in a binary endpoint setting with covariate adjustment. It is suited for clinical trials with or without covariate adaptive (stratified permuted block or biased coin) randomization where the summary measure of the marginal estimand is one of (risk difference, odds ratio, risk ratio, log odds ratio, log risk ratio). For practical considerations on the implications of covariate adjustment in superiority vs non-inferiority trials, please see Nicholas et al. (2015) and Morris et al. (2022).

Example

This is a basic example which shows how to obtain the point and variance estimates of a marginal estimand with covariate adjusted working model:

library(beeca)

## Set treatment to a factor
trial01$trtp <- factor(trial01$trtp)

## Fit a logistic regression working model and pass it to beeca
fit1 <- glm(aval ~ trtp + bl_cov, family="binomial", data=trial01) |>
  get_marginal_effect(trt="trtp", method="Ye", contrast="diff")

## View the results in Analysis Results Data (ARD) structure
fit1$marginal_results

Package documentation

The package documentation can be found here. For a brief overview of the different estimands and their estimation, please see vignette vignette("estimand_and_implementations").

Quality checks

Where possible we have cross checked the {beeca} package with alternative implementations in SAS and R. For example, the Ge et al. method which applies the delta method has been cross checked against the SAS %margins macro and the R packages {margins} and {marginaleffects}. The Ye et al. method has been cross checked against {RobinCar}.

Package authors

Acknowledgments

Our lightweight implementation was inspired and aided by the more comprehensive {RobinCar} package, developed by Marlena Bannick, Ting Ye et al. We thank the ASA-BIOP Covariate Adjustment Scientific Working Group for valuable feedback and discussions.

Further development of covariate adjustment software is by the Software Subteam of ASA-BIOP Covariate Adjustment Scientific Working Group.

References

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.