Understanding the FDA’s Guidelines for Evaluating the Reliability of Computational Modeling and Simulation: Examining Credibility Evidence – Part 1
Computational modeling and simulation have become increasingly important tools in various industries, including healthcare and pharmaceuticals. These techniques allow researchers and scientists to simulate complex systems, predict outcomes, and make informed decisions. In the field of drug development, computational modeling and simulation can play a crucial role in evaluating the safety and efficacy of new drugs before they are tested on humans.
The U.S. Food and Drug Administration (FDA) recognizes the potential of computational modeling and simulation in drug development and has developed guidelines to evaluate the reliability of these techniques. These guidelines aim to ensure that computational models and simulations are credible and can be used as evidence to support regulatory decisions.
In this two-part article series, we will explore the FDA’s guidelines for evaluating the reliability of computational modeling and simulation. In Part 1, we will focus on examining credibility evidence, which is one of the key aspects considered by the FDA.
Credibility evidence refers to the evidence that supports the reliability and accuracy of a computational model or simulation. The FDA requires that credibility evidence be provided to demonstrate that the model or simulation is fit for its intended purpose. This evidence should address various aspects, including model development, verification, validation, and uncertainty quantification.
Model development involves the creation of a computational model that accurately represents the system being studied. The FDA expects that the model development process follows established best practices and is well-documented. This includes providing a clear description of the model’s assumptions, equations, algorithms, and input parameters. The FDA also emphasizes the importance of transparency in model development, encouraging researchers to share their models and code to facilitate independent evaluation.
Verification is the process of ensuring that the computational model has been implemented correctly. It involves checking that the model’s equations and algorithms have been coded accurately and that the software implementation is free from errors. Verification can be achieved through various methods, such as code review, unit testing, and comparison with analytical solutions or known experimental data.
Validation is the process of assessing the accuracy and reliability of a computational model by comparing its predictions with experimental or observational data. The FDA expects that validation studies are conducted using relevant and representative data. The data used for validation should cover a range of conditions and scenarios that the model is intended to simulate. The FDA also emphasizes the importance of uncertainty quantification in validation studies, which involves assessing and quantifying the uncertainties associated with the model’s predictions.
Uncertainty quantification is an essential aspect of credibility evidence. It involves characterizing and quantifying the uncertainties associated with a computational model’s predictions. Uncertainties can arise from various sources, such as measurement errors, model assumptions, and parameter variability. The FDA expects that uncertainty quantification is performed rigorously and transparently, providing a clear understanding of the limitations and reliability of the model’s predictions.
In conclusion, credibility evidence plays a crucial role in evaluating the reliability of computational modeling and simulation in drug development. The FDA’s guidelines emphasize the importance of model development, verification, validation, and uncertainty quantification to ensure that computational models and simulations are credible and fit for their intended purpose. In Part 2 of this article series, we will delve deeper into the FDA’s guidelines and explore additional aspects related to evaluating the reliability of computational modeling and simulation.