FDA offers draft guidance for registries as RWD

The US Food and Drug Administration (FDA) released draft guidance on the use of registries in regulatory decision making, adding another piece to the puzzle of how the agency will consider real-world data (RWD) when considering a new drug indication or assessing post-approval requirements.
 
The draft guidance, which was published Monday, outlines considerations for sponsors and stakeholders when proposing to design a registry or use an existing registry to support regulatory decisions about the safety and effectiveness of drugs or biologics. The guidance does not offer recommendations on study design or type of statistical analysis when using registry data. The public can comment on the draft guidance until 28 February 2022.
 
This is the latest in a series of guidances from FDA as part of its framework for using Real-World Evidence in regulatory decisions. The new guidance comes about two months after the agency released draft guidance on the use of electronic health records (EHRs) and claims data as RWD. (RELATED: FDA issues draft guidance on RWD sourced from EHRs, claims data, Regulatory Focus 29 September 2021)
 
The EHR/claims data guidance focused on the selection of data sources that appropriately address the study question, the development and validation of definitions of study design elements, and data provenance and quality. Several trade associations, including the Pharmaceutical Research and Manufacturers of America (PhRMA), the Real-World Evidence (RWE) Alliance, the Biotechnology Innovation Organization (BIO), and the Alliance for Regenerative Medicine have all asked FDA to extend the comment period so that they can continue to review the EHR guidance and see how it aligns with other RWD guidance documents.  
 
Supporting regulatory decisions
Registries have varying degrees of usability in the regulatory context, depending on how the data are intended to be used, the patient population enrolled, the data collected, as well as how the datasets were created, maintained, curated, and analyzed. For example, registry data collected to obtain comprehensive clinical information on a particular disease may or may be appropriate to examine a drug-outcome association in a subset of those patients, according to the FDA draft guidance.
 
Registries have some advantages over other RWD sources since registries collect structured and predetermined data elements that can provide longitudinal, curated data on a defined patient population and their course of disease, complications, and medical care. Registries may also collect patient-reported data, which is ordinarily lacking from EHR datasets or claims data. But registries also have limitations for regulatory use. For instance, registries may focus on a single disease and provide limited information on comorbid conditions. Additionally, the enrolled patients may not be representative of the target population of interest.
 
“In general, registries are better suited as a data source for regulatory purposes when sponsors aim to capture objective endpoints, such as death or hospitalization. Subjective endpoints, such as pain, can be collected in a registry, but additional challenges are involved to standardize such measurements,” FDA wrote. “In addition, a registry that is designed to collect data to answer a specific research question can have advantages over an existing registry designed for another purpose, which is subsequently repurposed for that same question.”
 
Data relevance and reliability
When considering whether RWD is suitable for use in a regulatory context, the FDA advises considering the relevance and reliability of the data.
 
The assessment of the relevance of registry data should consider whether the registry’s data elements are adequate for evaluating the scientific objectives, FDA advised. The specific data elements depend on the intended use of the registry, but registry data used for an external control arm in an externally controlled trial may require more comprehensive data than a registry used to enroll participants in an interventional study, according to the draft guidance.
 
FDA noted that the registry should retain documentation on data elements that are no longer being collected or new data elements that begin to be collected. Sponsors should also develop a plan to reduce loss to follow up of registry participants.
 
When it comes to establishing reliability of a registry, sponsors should ensure that there are processes and procedures governing the registry operation, education and training of staff, resource planning, and general practices dedicated to ensuring the quality of data. For instance, good governance includes the maintenance of access control and audit trails on the provenance and traceability of the registry data.
 
“Indicators of data consistency, accuracy, and completeness should be assessed periodically, with the frequency dependent on the purposes of the registry data (e.g., for the sole purpose of facilitating recruitment in a randomized controlled trial versus using the registry data in an interventional or non-interventional study analysis). Routine descriptive statistical analyses should be performed to detect the extent of any missing data, inconsistent data, outliers, and losses to follow-up,” FDA wrote in the draft guidance.
 
Linking to another data source
The draft guidance also outlines considerations when linking a registry to another registry or a different type of RWD source, such as an EHR. If a registry does not capture all the necessary information, sponsors may consider obtaining supplemental data from an EHR or medical claims database or through digital health technologies, such as sensors that allow for health monitoring. 
 
If linking to another data source, sponsors should correct for redundant data and resolve any data inconsistencies, FDA advised. Sponsors should also establish a way to protect patient privacy while transferring data.
 
Regulatory review
Sponsors interested in using registry data to support a regulatory decision should meet with the agency before conducting a study that includes registry data, according to the guidance. Sponsors should be able to provide protocols and statistical analysis plans that include essential elements of a registry study’s design, analysis, and conduct. Additionally, if the registry data are owned and controlled by a third party, the sponsor should have an agreement in place to ensure that all relevant patient-level data can be provided to the FDA and that source records are made available for inspection.
 
Draft Guidance: Using Registries as RWE