FDA issues draft guidance on RWD sourced from EHRs, claims data

Regulatory NewsRegulatory News | 29 September 2021 |  By 

The US Food and Drug Administration (FDA) on Tuesday issued a long-awaited draft guidance detailing recommendations on the use of real-world data (RWD) sourced from electronic health records (EHRs) and medical claims data to support regulatory decision-making for drugs and biologics.
While data from EHRs and medical claims are widely used for pharmacoepidemiologic safety studies, the new draft guidance covers the use of RWD from such sources to support regulatory decisions on both safety and effectiveness.
FDA says the guidance is being issued for comment purposes only as part of its larger real-world evidence (RWE) program and to fulfill a requirement under the 21st Century Cures Act. FDA has also made commitments related to the use of RWE under the current Prescription Drug User Fee Act (PDUFA VI) program and has proposed an even greater focus on the use of RWE in the PDUFA VII commitment letter. (RELATED: FDA expands RWE demonstration project, Regulatory Focus 10 April 2019; PDUFA VII commitment letter: RWE, rare diseases see renewed attention, Regulatory Focus 23 August 2021)
The guidance focuses on three issues related to the use of RWD gathered from EHRs and claims data, including the selection of data sources that appropriately address the study question, development and validation of definitions for study design elements, and data provenance and quality throughout the study lifecycle from accrual to the final study-specific dataset.
The guidance is meant to compliment FDA's 2013 guidance, Best Practices for Conducting and Reporting Pharmacoepidemiologic Safety Studies Using Electronic Healthcare Data, by expanding on aspects of it that relate to the selection of data sources and "provides additional guidance for evaluating the relevance and reliability of both EHRs and medical claims data for use in a clinical study."
"This guidance also provides a broader overview of consideration relating to the use of EHR and medical claims data in clinical studies more generally, including studies intended to inform FDA's evaluation of product effectiveness," FDA writes.
FDA further explains that the guidance does not address study design or type of statistical analysis, nor does it endorse any specific type of data source or study methodology. "For all study designs, it is important to ensure the reliability and relevance of the data used to help support a regulatory decision," FDA writes.
The agency stresses that sponsors of studies relying on RWE sourced from EHRs or medical claims data should submit study protocols and statistical analysis plans to FDA and should request input from or a meeting with the relevant review division before proceeding with the study. Such protocols should identify all data sources that will be included in the study, along with descriptive information relating to the data.
"Each data source should be evaluated to determine whether the available information is appropriate for addressing a specific study hypothesis. Because existing electronic health care data were not developed to support regulatory submissions to FDA, it is important to understand their potential limitations when they are used for that purpose," FDA writes.
The agency provides several examples of potential limitations, such as an EHR not capturing a patient's comprehensive care when their treatment occurs across multiple health systems or networks. Even if EHR systems are modified for the purpose of collecting additional data for prospective clinical studies, FDA cautions that "given the limited ability to add modules to collect extensive additional information, EHR-based data collection may still not be comprehensive."
Determining the relevance of a data source is another key factor to consider. "There are differences in the practice of medicine around the world and between health care systems that may affect the relevance of the data source to the study question. Patients in different types of commercial or government health care payment programs can differ in a range of characteristics, such as age, socioeconomic status, health conditions, risk factors, and other potential confounders," FDA explains. Other factors, such as medication tiering, formulary decisions, and coverage across health care systems have the potential to influence the severity of disease in patients in treated with the same product from one health system to another.
FDA recommends that sponsors provide their reasoning for selecting a particular data source for a specific hypothesis, background information on the health care system and the methods of diagnosis and preferred treatments for the disease of interest, and a description of prescribing and use practices in the health system. When considering foreign data sources, FDA says sponsors should also provide an explanation of how the aforementioned factors might affect the generalizability of the results to the US population.
The guidance also provides recommendations on data capture, including on issues surrounding continuity of coverage, data linkage and synthesis, distributed data networks, computable phenotypes, and unstructured data. Additional suggestions are provided for handling missing data and data validation.
Study design considerations are detailed as well, with FDA offering insights on defining study time periods, selection of study population, and issues related to exposure, outcome and covariate ascertainment and validation.
The guidance also runs through data quality considerations over the course of the data lifecycle, from accrual and curation of data to the transformation and de-identification of data and its eventual storage in a data warehouse and the production of a study-specific dataset.
FDA, Federal Register notice


© 2023 Regulatory Affairs Professionals Society.

Discover more of what matters to you