Is there support for handling multiplicity in trial endpoints?

Yes, it includes guidance on using graphical multiplicity methods with the gMCPLite package to manage Type I error across primary and secondary hypotheses.

Does it support interim analysis and sample size re-estimation?

Yes, it includes detailed code patterns for both blinded and unblinded sample size re-estimation (SSR) and interim analysis results using the rpact package.

Can this skill help with Bayesian historical data integration?

Yes, it provides implementation patterns for Meta-Analytic Predictive (MAP) priors and robust priors using the RBesT package to integrate historical information while protecting against prior-data conflict.

What types of adaptive designs are covered?

It covers a wide range of modern designs including Platform, Basket, Umbrella, Response-Adaptive Randomization (RAR), and Multi-Arm Multi-Stage (MAMS) designs.

Which R packages does this skill support for clinical trials?

This skill covers major industry packages including adaptr for platform trials, rpact for group sequential designs, RBesT for Bayesian synthesis, basket for basket trials, and MAMS for multi-stage designs.

Advanced Adaptive Trial Design in R

Name: Advanced Adaptive Trial Design in R
Author: choxos

bychoxos

Ciencia de Datos y ML

Design, simulate, and analyze complex adaptive clinical trials using industry-standard R packages and Bayesian methods.

Acerca de

This skill provides specialized guidance for implementing advanced clinical trial methodologies in R, covering platform trials, basket and umbrella designs, and multi-arm multi-stage (MAMS) trials. It empowers biostatisticians and data scientists to perform interim analyses, response-adaptive randomization, and sample size re-estimation using robust packages like rpact, RBesT, and adaptr. Whether integrating historical data via Bayesian priors or managing multiplicity with graphical methods, this skill ensures best practices in modern clinical research design and simulation.

Características Principales

0 GitHub stars
Simulate platform trials with dynamic arm addition and dropping rules using adaptr
Design Multi-Arm Multi-Stage (MAMS) trials for continuous and binary outcomes
Implement response-adaptive randomization using Thompson sampling and constraints
Perform sample size re-estimation and interim analyses for group sequential designs
Analyze basket and umbrella trials with Bayesian hierarchical and MEM models

Casos de Uso

Integrating historical clinical data into current trials using Bayesian Meta-Analytic Predictive (MAP) priors
Conducting interim analyses to determine if a trial should stop early or requires a sample size adjustment based on conditional power
Designing a platform trial with multiple treatment arms and early stopping rules for futility or superiority

Acerca de

Características Principales

0 GitHub stars
Simulate platform trials with dynamic arm addition and dropping rules using adaptr
Design Multi-Arm Multi-Stage (MAMS) trials for continuous and binary outcomes
Implement response-adaptive randomization using Thompson sampling and constraints
Perform sample size re-estimation and interim analyses for group sequential designs
Analyze basket and umbrella trials with Bayesian hierarchical and MEM models

Casos de Uso

Integrating historical clinical data into current trials using Bayesian Meta-Analytic Predictive (MAP) priors
Conducting interim analyses to determine if a trial should stop early or requires a sample size adjustment based on conditional power
Designing a platform trial with multiple treatment arms and early stopping rules for futility or superiority