FDA's final Guidance on the Use of eHRs in Clinical Studies: Real-World Data

The FDA has released final guidance on Real-World Data (RWD) as part of its Real-World Evidence (RWE) program. This guidance provides information on using electronic health records (EHRs) and medical claims data in clinical studies to support regulatory decisions regarding effectiveness or safety. It finalizes the draft guidance issued in September 2021.

Real-World Data Sources and AI:

RWD can come from:

- EHRs

- Medical claims data

- Product and disease registries

- Patient-generated data, including from home use

- Other sources such as digital health technologies

EHRs and medical claims data are used in safety studies and increasingly in effectiveness studies. Advances in artificial intelligence (AI), such as natural language processing, machine learning, and deep learning, enable rapid processing of unstructured data from EHRs. AI can extract data, identify outcomes, and evaluate images or lab results.

When using AI in protocols, specify the algorithm's assumptions, parameters, data sources, whether it is supervised or unsupervised, and validation metrics.