FDA Final Guidance on Real-World Data: Key Considerations for EHR and Medical Claims Data in Drug Development

The Food and Drug Administration (FDA) recently released final guidance on Real-World Data (RWD) and specifically assessing Electronic Health Records (EHR) and Medical Claims Data to support regulatory decision-making for drug and biological products.

The guidance outlines the current thinking of FDA and important considerations on RWD. A key aspect is thoroughly evaluating the appropriateness, strengths, and limitations of the RWD source for the intended analysis. Strategies like data linkage and validation against external reference standards can help address limitations around completeness, accuracy, and potential misclassification. However, as stated in the guidance, RWD sources have inherent constraints compared to controlled clinical trials.

Some key requirements from the FDA guidance on real-world data for medical product development include:

• Selecting appropriate data sources for the study question and design elements. The data source should have information to support key aspects like study population, exposures, outcomes, covariates, etc.
• Having clear conceptual and operational definitions for study variables like inclusion/exclusion criteria, exposures, outcomes. Conduct validation studies to assess how accurately variables can be ascertained from the data source.
• Defining the time periods for ascertaining different study design elements like exposures, outcomes.
• Having quality assurance and quality control procedures for data collection, curation, and creation of the final study dataset. This includes things like characterizing the data, handling missing data, validating key variables.
• Linking different data sources if needed to get more complete information. Using distributed data networks rather than transferring data can help maintain privacy.
• Using computable phenotypes and natural language processing on unstructured data where feasible.
• Conducting validation on key study variables by comparing to an external data source or chart review. Select variables that are critical to the analysis for validation.

In summary, the guidance emphasizes careful consideration of the data source, validation of key variables, and quality control processes for generating the study dataset. The goal is to maximize the accuracy and reliability of real-world data used for medical product development.

The content of this post was generated using Nyquist AI.