| アブストラクト | This study aims to systematically evaluate the adverse reactions of flumazenil by analyzing data from the U.S. Food and Drug Administration's Adverse Event Reporting System (FAERS), identifying and quantifying its potential risks across different system organ classifications. The research focuses on patients with long-term benzodiazepine use, aiming to provide more comprehensive safety guidance for clinical application and to support future pharmacovigilance policy development. This study utilized adverse event report data from the FAERS database, covering the period from 2004 to 2023. Multiple signal detection algorithms, including the reporting odds ratio (ROR), proportional reporting ratio (PRR), Bayesian confidence propagation neural network, and multi-item gamma Poisson shrinker, were employed to analyze and classify the adverse reaction signals of flumazenil. The study enhanced the sensitivity and accuracy of signal detection using advanced machine learning tools, and the findings were compared and validated against existing literature data. The results showed a significant association between flumazenil and neurological adverse reactions, such as "withdrawal seizures" (ROR 1737.25, PRR 1645.3) and "psychogenic epilepsy" (ROR 774.32, PRR 764.68), with higher risk particularly observed in patients with long-term benzodiazepine use. In addition, procedural complications, such as delayed emergence from anesthesia and intentional overdose, also exhibited notable signal strength. Although cardiovascular and respiratory adverse reactions were rare, rare risks such as stress-induced cardiomyopathy and respiratory depression still warrant attention. By integrating FAERS data and multiple signal detection algorithms, this study confirmed the potential high risk of flumazenil use in clinical practice, particularly concerning neurological adverse reactions. Clinicians should increase monitoring in high-risk patients, such as those with long-term benzodiazepine use, and adjust dosing appropriately to reduce the incidence of adverse reactions. Future research should further incorporate electronic health records and prospective clinical trials to develop more sensitive real-time monitoring tools, optimize high-risk patient management, and improve drug safety. |
|---|
