(A-335) Leveraging Real-World Data for Regulatory Commitment: Methodological Considerations in a Post-Approval Safety Study Evaluating the Risk of Atrial Fibrillation (AF) Following Respiratory Syncytial Virus (RSV) Vaccination

Background: Post-marketing requirements and commitments using real-world data are mandated by regulators to further assess vaccine safety after approval. Given the limitations of observational data, these studies require robust methodology to mitigate biases (eg, confounding) and accurately estimate safety risks. ABRYSVO® is a newly approved vaccine for prevention of RSV. A numerical imbalance of AF was observed in ABRYSVO’s Phase 3 clinical trial, though no cases were considered related to ABRYSVO by investigators. Further evaluation of the potential risk of AF associated with ABRYSVO is designated as a post-marketing commitment to the FDA.

Objectives: To describe methodological considerations for a post-authorization safety study to assess the risk of AF following ABRYSVO vaccination in adults 60 years and older.

Methods: This is a longitudinal cohort study utilizing electronic health record (EHR) data from the US Veterans Health Administration (VHA). Exposure to ABRYSVO will be identified by procedure codes and immunization records. Pre-identified methodological challenges include: 1) lack of established post-vaccination risk intervals for AF; 2) potential for outcome misclassification given lack of validated case definition; 3) potential for incomplete vaccination status and outcome data as veterans may receive care outside of VHA and affiliated health care systems; 4) impact of confounding by indication and healthy user bias.

Results: In the final study protocol, several post-vaccination risk intervals are used to identify AF. Days 0-1 and 0-3 (Day 0 is the vaccination date) were selected based on FDA’s feedback on early post-marketing case reports indicating acute onset, while Days 0-30 is also used as a sensitivity analysis based on broader vaccine literature for cardiac events. To increase outcome specificity, new onset AF is defined by an initial diagnosis code for AF in any setting without a prior diagnosis in the 2-year pre-vaccination washout period followed by a confirmatory diagnosis code within the next 30 days. Validation via medical chart review is planned to evaluate the accuracy of the proposed case definition. AF risk is assessed through internal comparator cohort and self-controlled risk interval designs considering their complementary strengths with respect to confounding control, interpretability, and efficiency. To enhance data completeness, EHR will be linked with Medicare and community care.

Conclusions: This abstract illustrates approaches to address common methodological challenges in real-world vaccine safety studies, offering useful insights to pharmacoepidemiologists looking to leverage real-world data to fulfill regulatory requirements.