Leveraging {targets} and {crew} to simulate clinical trials

In clinical trial design, simulations are essential for evaluating potential designs and optimizing features like sample size, allocation, randomization, milestones, and decision criteria. However, simulations require thousands of repeated experiments and long execution times, and the demand is too overwhelming for workflows that are informal, interactive, or single-threaded. Using {targets}, a pipeline tool for reproducible analysis, and {crew}, a flexible {mirai}-based framework for distributed computing, this pipeline showcases a powerful approach to conduct large-scale clinical trial simulations.

Materials

• Slides: https://wlandau.github.io/rpharma2023
• Code: https://github.com/wlandau/rpharma2023-pipeline

Resources

• targets website: https://docs.ropensci.org/targets/
• targets user manual: https://books.ropensci.org/targets/