基于FAERS数据库的泼尼松不良事件信号挖掘与分析OA

Objective To conduct data mining on post-marketing adverse events of prednisone based on the data from the FDA Adverse Event Reporting System(FAERS),to provide evidence for its rational clinical use.Methods Adverse event reports(ADEs)covering 81 quarters from the first quarter of 2004 to the first quarter of 2024 were retrieved from the FAE-RS database.Disproportionality analysis was performed to detect safety signals.The mainly involved organ systems and high-risk signals were also analyzed.Results A total of 107 109 ADEs related to prednisone were collected,and 1,312 consist-ent positive signals were identified by three data mining methods across 26 system organ classes(SOCs).The major SOCs included infections and infestations(34.38%),benign,malignant and unspecified neoplasms(8.84%,including cysts and polyps),and respiratory,thoracic and mediastinal disorders(5.34%).The top three adverse drug reactions by signal strength were histoplasma endocarditis,stage Ⅳ extranodal marginal zone B-cell lymphoma(MALT type)and parapoxvir-us infection.Conclusion No positive signal of prednisone-related death was found in this study.Prednisone can exacerbate infections as stated in its package insert,and also increase the risk of developing unlisted benign,malignant and unspeci-fied neoplasms.These findings warrant heightened clinical vigilance,necessitating enhanced therapeutic monitoring and fol-low-up.