FDA Releases Draft Guidance Supporting Computational Modeling

Tablet being held by a doctor showing images of brain scans

April 29, 2022

Establishes process for assessing credibility of models used in medical device submissions

The U.S. Food and Drug Administration (FDA) released a draft guidance for assessing the credibility of computational modeling and simulation (CM&S) in medical device submissions, where credibility is defined as trust in the predictive capability of a computational model. The draft guidance builds on prior standards and guidance documents (e.g., ASME V&V40) to assist with the use of CM&S to support medical device regulatory submissions in multiple applications, including (1) in silico device evaluation, (2) CM&S used within medical device software, (3) in silico clinical trials, and (4) CM&S-based qualified tools, such as those in the Medical Device Development Tools (MDDT) program.

The framework established in the guidance document does not define how to perform modeling studies, nor does it specify appropriate levels of rigor or credibility that should be achieved for any model. Instead, the scope of the document presents a nine-step framework to inform the development of an analysis/testing plan to support the use of the model, along with the appropriate process for assessing the relevant level of credibility necessary to support the desired use of the CM&S results.

The use of CM&S in medical device submissions is becoming more common as enhanced computational methods and capabilities combine to offer greater insight into the safety and efficacy of medical devices. Methods for establishing the credibility of computational models in other industries (e.g., aerospace and automotive) are well documented, owing to their long history of use. However, the complexities associated with medical device development (e.g., the highly variable nature of human physiology) create specific challenges that make the use of CM&S more challenging. The draft guidance leverages established methods — which often require rigorous experimental data collection — with sources of credibility evidence from broader clinical activities. Using credibility evidence from clinical studies is particularly important in life sciences, where clinical studies may be necessary to provide evidence of device safety and efficacy.

To support a medical device submission, the computational model must be documented in a way that allows regulators to assess the credibility of the underlying model before interpreting the results in the context of the broader submission package, which may include clinical and experimental results. By providing this generalized framework, FDA is offering medical device developers a streamlined process for moving through the necessary technical steps to establish the credibility of a model, likely reducing uncertainty about the regulatory acceptance of a model used in a device submission and enhancing the value of the model predictions in support of device safety and efficacy.

Given the diversity of existing CM&S applications and methods in life sciences, the draft guidance process provides opportunities for medical device developers already using CM&S and the general public to provide comments to ensure that current methods are appropriately considered in the final guidance, thus potentially reducing the need to modify existing analysis methods and plans.

How Exponent Can Help

Exponent has extensive experience conducting advanced CM&S, as well as developing and executing experimental methods with accompanying data analysis methods for use in regulatory submissions for medical devices. This deep expertise uniquely positions Exponent to help clients develop and implement CM&S activities and assess model credibility for medical device submissions. Exponent consultants have been active members of the ASME V&V40 subcommittee since its inception, including holding multiple leadership positions. Our regulatory experts regularly work with device developers to understand critical conditions driving device performance, develop computational models to characterize the device's performance and interactions with associated tissues/organs/external electromagnetic fields (where appropriate), and design relevant benchtop experiments to provide insight into model, test, and clinical results.