An Introduction to Principal Surrogate Evaluation with the pseval Package

Area of Science:

• Biostatistics
• Clinical Trial Methodology
• Pharmaceutical Research

Background:

• Principal surrogates are crucial for evaluating treatment effects when the primary endpoint is delayed or unobservable.
• Existing methods for surrogate evaluation can be complex to implement in statistical software.
• There is a need for accessible tools to facilitate the analysis of principal surrogates in clinical trial data.

Purpose of the Study:

• To introduce and describe the 'pseval' R package, designed for the evaluation of principal surrogates.
• To provide a flexible and user-friendly interface for implementing statistical models and estimation methods for surrogate evaluation.
• To demonstrate the utility of the package for both experienced and novice R users.

Main Methods:

• The package implements core statistical methods for principal surrogate evaluation within a single clinical trial.
• It supports flexible model definition for risk, surrogate, and counterfactual responses.
• Estimation methods include maximum likelihood and pseudo-score, with bootstrap for statistical inference.

Main Results:

• The 'pseval' package provides a comprehensive framework for principal surrogate evaluation.
• It includes various post-estimation functions such as print, summary, plot, and testing.
• The package is designed to be accessible to users with varying levels of R programming experience.

Conclusions:

• The 'pseval' R package offers a valuable and accessible tool for researchers conducting clinical trial analysis.
• It streamlines the complex process of evaluating principal surrogates, enhancing the efficiency of research.
• The package facilitates robust statistical inference and data visualization for surrogate endpoint analysis.