Real-World Data Integration for Causal Inference: Benefits, Costs, and Case Studies

PURPOSE: To guide participants in using integrated real-world data for causal inference, highlighting the advantages and challenges. Several case studies will illustrate practical applications and lessons learned. Participants will interactively share their experiences and participate in open discussion.

DESCRIPTION: Causal inference from observational data is growing in importance, driven by the need for generalizable and rapidly delivered real-world evidence (RWE) to inform regulatory, payer, and patient/provider decision-making. Integrating two or more individual sources of real-world data (e.g., administrative health claims, electronic medical records, registries, and publicly available social determinants of health data) can provide deeper insights into the patient's health journey. Using integrated data can lower methodological and resource barriers to comparative effectiveness and long-term safety assessments. However, integrating data often requires trade-offs regarding variable consistency, available sample size, and selection bias.

Workshop participants with intermediate-level subject knowledge of integrated data or causal inference will be familiarized with crucial considerations when using these data for causal study design. Dr. Grabner will initiate the workshop by discussing the basics of integrated data, causal inference, and their interplay and solicit participant feedback on their experiences using integrated data for causal inference. Afterward, Dr. Yu will present the first case study, which used an external control arm to enable causal inference in a study of patients with relapsed refractory multiple myeloma and resulted in regulatory approval of a new therapy. Next, Dr. Dixon will present the second case study, where directed acyclic graphs were developed to understand causal pathways for treating metastatic colorectal cancer and to illustrate their application to a large, integrated data set. Dr. Lloyd will conclude by providing an FDA case study on generating RWE to study COVID-19 vaccines in the Biologics Effectiveness and Safety Initiative.

There will be real-time polling and time for audience questions to support interactivity.