*Constructing a Continuous, Registry-independent, AI-enhanced RWE Infrastructure

Post-market surveillance is a core FDA mandate to ensure the continued safety and effectiveness of medical devices throughout their total product life cycle. Device-specific post-approval studies often have a limited reach both horizontally and longitudinally. Registries, administrative claims data, and the linkages between them have provided invaluable real-world data for post-market surveillance. However, registries and claims data also have limitations. For example, national registries often have decreasing data completeness over time, and CMS claims data lack granularity and do not include non-Medicare beneficiaries. The proposed project aims to overcome some of the limitations of the registry/claims-based post-market surveillance paradigm by drawing directly from the primary sources of real-world clinical data (electronic health records, EHRs) and to construct a continuous, registry-independent, AI-enhanced real-world evidence (RWE) infrastructure for medical device surveillance. This infrastructure has the following core characteristics: Registry-independent AI-native Could support a federated or aggregated architecture Longitudinal and lifelong Prospectively pre-specified