For the purpose of this draft guidance, FDA defines a non-interventional study as a study in which patients receive the marketed drug of interest during routine medical practice and are not assigned to an intervention according to a protocol. The guidance describes attributes for design and analysis of a non-interventional study that sponsors should consider when proposing to use such a study to contribute to a demonstration of substantial evidence of effectiveness and/or evidence of safety of a product. Specifically, FDA proposes that sponsors provide information on the following key attributes to support the use of non-interventional studies: (1) summary of the proposed approach, (2) study design, (3) data sources, and (4) analytic approach.

Summary of the proposed approach

FDA advises that sponsors finalize the study protocol, including the research question of interest and rationale for design, before initiating the study. Importantly, FDA expects sponsors to also provide information on alternative study approaches (e.g., randomized trials and single-arm trials), and candidate data sources that they considered before deciding on their proposed approach, and discuss why those alternative approaches were not feasible.

Study design

FDA recommends incorporating key study design elements in the protocol, including, but not limited to: theorized causal relationship; source population; eligibility criteria; relevant covariates (e.g., concomitant treatments) and strategies for addressing bias.

Data sources

FDA requests that sponsors provide information not just on the proposed data source(s) and its appropriateness, but also on potential limitations of the source(s), cautioning that sponsors must determine whether those limitations can be addressed or if another data source should be pursued.

Analytic approach

FDA recommends that sponsors provide a prespecified statistical analysis plan (SAP) which “address[es] the specific study objectives and detail[s] the primary analysis and any secondary analyses.” The plan should include information and justification for the approaches employed regarding: (1) statistical power, (2) confounding factor accounting, (3) overadjustment potential, (4) subgroup analyses, (5) differential surveillance/misclassification, and (6) reverse causality.

Insights

Developing a shared understanding of the collection of RWD and use of RWE to inform regulatory decision making has been an agency-wide strategic priority since the RWE Program was initiated in 2018. The agency’s collection of guidances and publications are online here.

This draft guidance on non-interventional studies provides additional clarity on FDA’s focus when analyzing the adequacy of a RWE program. Notably, unlike previous guidances on the topic, FDA specifies that non-interventional studies can be used to support a demonstration of substantial evidence of effectiveness and/or the safety of a drug. Although this may have been implied in the agency’s use of “regulatory-decision making” in earlier RWD/RWE guidances, specifying the substantial evidence standard may indicate FDA is open to placing more reliance or weight on RWE when determining the approvability of a drug or biologic product, including establishing the risk/benefit profile.

Given FDA’s efforts to provide clarity in this area, FDA is acknowledging the impact that data from electronic health records, medical claims, disease registries, and digital drug development tools can have on approval decisions. Consistent with prior guidance, when seeking to incorporate RWD to support regulatory decisions for product development, sponsors should consult with FDA early in the planning of a study to discuss FDA expectations for design and conduct.

FDA seeks comments on the draft guidance through June 18, 2024. If you wish to submit a comment, or have any questions on utilizing real-world data in drug or biologics applications, or on clinical study design more generally, please contact any of the authors of this alert or the Hogan Lovells attorney with whom you regularly work.

Authored by Lynn Mehler, Sally Gu, Yetunde Fadahunsi, Bryan Walsh, and Katie Kramer