The Food and Drug Administration (FDA) continues to advance its RWE Program, this time with a new Draft Guidance titled “Real-World Evidence: Considerations Regarding Non-Interventional Studies for Drug and Biological Products” (the Draft Guidance). The Draft Guidance, released on March 20, clearly outlines what is expected of sponsors when proposing non-interventional (observational) studies that intend to demonstrate substantial evidence of effectiveness and/or generate evidence on the safety of a drug. This could include a use similar to the Prograf scenario, where the sponsor used observational methods for a label expansion (to prevent organ rejection in adult and pediatric patients receiving lung transplantation). The Agency cites three types of non-interventional (fully observational) study designs for the purposes of the draft guidance – cohort studies, case-control studies, and self-controlled studies – and reviews key methods issues that should be considered during protocol development.
Throughout the Draft Guidance, FDA reiterates key points emphasized in previous draft and final guidances. These include the importance of prespecification of key study design aspects and early engagement with the Agency. The importance of early engagement is emphasized by the intentional flexibility of some portions of the Draft Guidance. The decision to conduct a non-interventional study and its design involves many situational considerations, making it critical to engage the relevant FDA review division and the RWE Subcommittee to determine the most appropriate course of action. This will help produce the most interpretable study and allow the sponsor to bring FDA along on its approach, ensuring an efficient and productive engagement. FDA also reminds sponsors of the importance of data quality, stating that data need to have sufficient fitness to support a study proposal. If they do not, the sponsor should consider an alternative (such as primary data collection).
It is worth noting that much of the Draft Guidance’s content is aligned with what we at Aetion call the “show your work” concept, reflected in the SURF and SPACE/SPIFD/SPIFD2 publications that were led and co-authored by Aetion colleagues. The Agency notes that sponsors should include explanations of why alternative approaches like randomized controlled trials (RCTs) or single-arm trials (SATs) are not feasible to answer the study questions and provide the rationale for the proposed non-interventional study design and data source, including why other options were discarded. The first point about explaining why a randomized trial is infeasible is explicitly addressed by our SURF framework, which sponsors can use to determine and communicate the regulatory feasibility of an RWE approach for providing substantial evidence of effectiveness. In the same spirit, the SPIFD2 framework (built on our SPACE and SPIFD predecessor frameworks) facilitates clear communication of non-interventional study design and data selection decisions, including their rationale.
One other notable aspect of the Draft Guidance is that the Agency distances itself from using the terms “prospective” and retrospective.” FDA states, “The terms prospective and retrospective are commonly but variably used to indicate whether timing of the cause-effect association occurs prior to or concurrent with the investigation that is examining it, whether inferential reasoning is from cause-to-effect or vice versa, whether sample selection is based on exposure or outcome status, or whether a study hypothesis is established prior to or after the corresponding data were collected. These terms are used sparingly in this guidance.” Aetion appreciates this approach because we believe the most important consideration for evaluating data sources is how much control a researcher has over the reliability and relevance of the data for the research question of interest, regardless of the data source. Sponsors should not anchor on a prospective/retrospective distinction when assessing data fitness because it is potentially misleading. Instead, they should use a primary/secondary data collection paradigm for quality and objectivity needs.
Watch the on-demand webinar to learn more about Aetion’s SURF and SPIFD2 frameworks. Contact us today for more information on how we can partner with you to address your regulatory RWE needs.