How does it integrate with R programming?

It provides standardized code snippets and best practices for using the simtrial and gsDesign2 packages to run and summarize large-scale simulations.

What alpha spending functions are supported?

The skill provides implementation guidance for standard functions including O'Brien-Fleming, Pocock, and the flexible Hwang-Shih-DeCani (HSD) family.

Can I use this for time-to-event trials?

Yes, it includes specific patterns for event-driven trials where the information fraction is calculated based on the number of observed events.

Does it support futility stopping?

Yes, it covers both binding and non-binding futility boundaries, as well as beta-spending approaches to manage Type II error.

Can I apply different statistical tests at different analysis stages?

Yes, the skill demonstrates how to configure unique tests (e.g., different weighted log-rank weights) for each interim look and the final analysis.

Clinical Trial Group Sequential Methods

Name: Clinical Trial Group Sequential Methods
Author: choxos

bychoxos

0•

Data Science & ML

Implements group sequential design methods for clinical trial interim analyses, including alpha spending and futility stopping rules.

This skill provides expert guidance and implementation patterns for designing group sequential clinical trials. It enables researchers to set up interim analyses with sophisticated alpha spending functions like O'Brien-Fleming or Pocock, calculate information fractions, and implement both binding and non-binding futility rules. By integrating with specialized R packages like simtrial and gsDesign2, it helps automate complex simulations to evaluate power, Type I error, and expected sample sizes, ensuring clinical trials are ethically sound, operationally efficient, and statistically robust.

Key Features

010 GitHub stars

02Calculation of information fractions and event-driven analysis timing

03Integration with R packages simtrial and gsDesign2 for GS simulations

04Configuration of binding and non-binding futility stopping boundaries

05Implementation of O'Brien-Fleming, Pocock, and HSD alpha spending functions

06Evaluation of operating characteristics including power and expected sample size

Use Cases

01Calculating updated statistical boundaries based on actual accrued information

02Simulating trial outcomes to determine the optimal timing for interim analyses

03Designing a Phase III trial with interim efficacy looks to enable early stopping

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add choxos/biostatagent group-sequential-methods

For use in Claude.ai and ChatGPT

Key Features

010 GitHub stars

02Calculation of information fractions and event-driven analysis timing

03Integration with R packages simtrial and gsDesign2 for GS simulations

04Configuration of binding and non-binding futility stopping boundaries

05Implementation of O'Brien-Fleming, Pocock, and HSD alpha spending functions

06Evaluation of operating characteristics including power and expected sample size

Use Cases

01Calculating updated statistical boundaries based on actual accrued information

02Simulating trial outcomes to determine the optimal timing for interim analyses

03Designing a Phase III trial with interim efficacy looks to enable early stopping

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add choxos/biostatagent group-sequential-methods

For use in Claude.ai and ChatGPT