R-Forge Logo

Welcome to Causal Effects Sensitivity Analysis project!

Graphical Sensitivity Analysis for Assessing Surrogate Endpoints via Causal Inference

This package provides a graphical tool to facilitate sensitivity analysis for assessing the surrogate value of a biomarker. It is based on the principal stratification framework.

A surrogate endpoint is a biomarker measurement made shortly after a treatment is administered that reliably predict future clinical endpoints. Several statistical methods have been proposed to assess the surrogate value of biomarkers. Most seek to estimate counterfactual-based quantities that represent causal effects. Since the estimands underpinning these methods generally are not statistically identifiable without strong, untestable assumptions, sensitivity analysis plays a key role.

The project summary page you can find here.