OPTIMIZE CLINICAL STUDY DESIGN WITH RANDOMIZATION SIMULATION
Sponsored by: Anju Software
- Clinical Study Design
Date: 18 November
Days to go: 23
Time: 3PM London/10AM New York
The power of simulations for initial design and adaptive trial changes
Clinical Trial designs are becoming more complex and the pressure to streamline and accelerate clinical trials continues to increase. Nonetheless, it remains critical to carefully design randomization methods that will result in a final patient distribution that is optimized for statistical analyses. Delays in designing randomization schemes and subsequent validation of methods can negatively impact trial start-up timelines. In adaptive trials, the impact of this delay is multiplied every time the design of the trial is adjusted. To maintain speed and the appropriate rigor in randomization design it is important to be able to quickly and reliably assess multiple designs and rapidly simulate results, even when using dynamic randomization techniques.
Dynamic randomization algorithms can offer several advantages when used in a clinical trial over the more commonly used permuted block method. However, implementing a minimization algorithm, such as Zelen’s method or Pocock & Simon, involves more difficult computational work and requires specifying multiple parameters that affect the outcome of the randomization. It can be a significant challenge to determine the optimum parameters for a particular study design and has the potential to introduce delays in study startup timelines.
A Randomization Simulator permits a statistician to create a simulation of a trial and obtain results from a specified subject listing. Multiple simulations can easily be created adjusting the randomization parameters, or with different random number seeds to assess the impact of adjusting the values and estimating the range of random variation in the results. The results of this analysis can be used to determine the optimum settings (including the weighting of treatment groups and stratification factors). Once the final settings are decided, this type of simulation can also be used to validate the randomization configuration in your RTSM/IRT.
Join us to learn from industry experts to learn more about how automated randomization simulations can support the use of dynamic randomization techniques without negatively impacting your study timelines.
Tony Warren is a Product Manager at Anju Software. Tony holds a B.S. from Purdue University and an M.S. from the University of Arizona. After serving in the US Air Force as a scientific programmer he spent time developing software solutions in the financial and manufacturing industries. His experience in the Life Sciences industry began when he joined an emerging CRO, Kendle International, as a developer for a new phone-based (IVRS) randomization system – TriaLine. Tony spent 13 years with Kendle expanding the use of TriaLine and leading the transition from a phone-based to a web-based solution. In 2018, he joined OmniComm (now part of Anju Software) as Product Manager where he is responsible for the direction and growth of Anju’s RTSM solution, IRTMaster.
Key Learning Objectives
- Components of a system for randomization and clinical trial supplies management
- Introduction to dynamic minimization based on Zelen and Pocock & Simon algorithms
- Building a simulation and archiving results
- Creating variants of existing simulations in a batch mode to enable a large suite of results to be created for detailed analysis.
- Statistical Programming
- Clinical Data Management
- Clinical Operations
- Clinical Trial Supply Management