Real-world patterns of herbal medicine-related adverse event reports and surveillance gaps in national pharmacovigilance: a 15-year retrospective analysis from a single integrative medical center in Korea

Herbal medicine (HM) is widely used in traditional Korean medicine (TKM); however, national databases primarily capture licensed HM products and may not fully reflect diverse formulations used in real-world practice. This study analyzed HM-related adverse event (AE) reports from an integrative medical center in Korea, focusing on reporting patterns, reported HMs and AEs, and differences between institutional reports and those submitted to the national pharmacovigilance agency. We conducted a retrospective descriptive study of HM-related AE reports collected between January 2010 and December 2024 from the shared AE reporting databases of Dongguk University Ilsan Hospital and Dongguk University Ilsan Oriental Hospital. Reports in which HMs were recorded as suspected agents were included. We analyzed reporting characteristics, patient demographics, suspected HM types, AE profiles, and national pharmacovigilance submission status, and compared submitted and non-submitted reports. After duplicate removal, 454 HM-related AE reports were analyzed. Among the 38,367 AE reports handled during the study period, HM-related reports accounted for 1.2%. Of the reported HMs, non-licensed oral HMs accounted for the largest proportion (48.7%), followed closely by MFDS-licensed oral HMPs (47.0%). At the report level, 51.3% involved at least one non-licensed HM formulation. Gastrointestinal disorders were the most frequently reported system organ class, followed by skin and subcutaneous tissue disorders. Among the 572 drug–event pairs with evaluable causality, most were assessed as possible (87.7%), whereas certain cases were rare (0.2%). Overall, 24.0% of the reports were not submitted nationally. Compared with submitted reports, non-submitted reports more frequently involved non-licensed HMs (83.5% vs. 41.2%, p < 0.001) and less often included documented clinical severity grading (7.3% vs. 59.1%, p < 0.001) and causality assessments (26.6% vs. 98.8%, p < 0.001). The current pharmacovigilance framework does not fully capture the range of HMs used in real-world Korean practice, particularly non-licensed formulations. Broader surveillance coverage, improved clinical assessment documentation, and stronger reporting support for TKM professionals are needed to build a more comprehensive HM pharmacovigilance system in Korea.