Can the FDA track cannabis safety? Researchers reveal the challenges

The applicability of spontaneous reporting systems such as the US Food and Drug Administration Adverse Event Reporting System (FAERS) to detect cannabis-related safety signals remains unclear due to the potential for discrepant reporting patterns between pharmaceutical and non-pharmaceutical cannabis-derived products (CDPs). We conducted a descriptive analysis of seven groups of CDP reports submitted to FAERS between 1999 and 2023 to investigate product definitions and reporting patterns. We then performed hypothesis-free disproportionality analyses using reporting odds ratio, proportional reporting ratio, and information component for pharmaceutical cannabidiol (CBD) and non-pharmaceutical CBD reports to assess differences in signal detection profiles, potential exposure misclassification, and the influence of reporting context. We identified 42 530 reports related to CDPs, characterized by highly heterogeneous terminology and variable reporting patterns by product type, reflecting the real-world CDP usage. Epidiolex reports often involved pediatric patients, whereas non-pharmaceutical CBD reports were more frequently associated with older adults and concomitant product use. Disproportionality analysis showed divergent signal profiles, with strong seizure-related events predominating for Epidiolex and a broader range of signals, including neoplasm-related and neurological events, observed for non-pharmaceutical CBD. These differences likely reflected variations in CDP indication and utilization and reporting behaviors. This study showed that signal detection using FAERS has potential feasibility for CDP safety surveillance. However, unique challenges related to exposure definitions, reporting patterns, motivation for utilization, and the need for a robust study design must be addressed to ensure reliable safety signal detection. Self‐reports of adverse health events suspected to be related to medications are collected in the US FDA Adverse Event Reporting System (FAERS) and screened for early signals of potential side effects. The feasibility of using FAERS to monitor potential adverse drug reactions to cannabis‐derived products (CDPs) is unclear, as individual cannabinoids are utilized in different ways for varying reasons, described with different names, and include both prescription and non‐prescription forms, unlike other medications in the database. Reports concerning seven groups of CDPs submitted to FAERS from 1999 to 2023 were summarized to explore product descriptions and reporting patterns. Potential safety signals were then screened for prescription cannabidiol and non‐prescription cannabidiol using standard methods. Significant variation in product description was observed, and reporting patterns varied by product type. Signal detection profiles also differed between prescribed and non‐prescribed cannabidiol. These differences may reflect patterns of use, indications, and reporting context. We conclude that while FAERS can contribute to CDP safety surveillance, reliable interpretation requires clearer product definitions, methods of classification, and study designs that account for the unique reporting patterns, which pose challenges beyond the usual limitations of drug safety surveillance in this database.