Optimizing Use of Real‐World Data in Regulatory Decision Making for Medical Devices Used in Surgical Procedures

PURPOSE: In the United States, the Food and Drug Administration has provided guidance on the contexts in which real-world evidence (RWE) can be used to support regulatory decisions for medical devices. METHODS: In this paper, we review different sources of real-world data (RWD) for medical devices used in surgical procedures and how to optimize the design and analysis features of RWD studies to maximize the strength of RWE. The design and analysis of a study are critical to ensuring that the RWE generated from the RWD can ultimately be used to support a regulatory decision or fulfill a commitment for a medical device. We consider several topics relevant to medical device RWE in our review: misclassification and measurement error, causal inference, and incomplete follow-up. RESULTS: Some key challenges include indication and outcome misclassification in non‑registry data, tenuous assumptions of single‑arm designs, device‑ and surgeon‑level confounding in comparative studies, and potentially informative loss to follow‑up from censoring (e.g., insurance termination). CONCLUSION: The review is intended as a guide to researchers and regulators interested in generating or evaluating medical device RWE for regulatory purposes.