目的 采用偏最小二乘-判别分析(PLS-DA)算法分析参麦注射剂药物不良反应(ADR)关键影响因素.方法 监测患者使用参麦注射剂后的临床表现,提取患者个体信息和临床用药信息,构建多变量数据矩阵.建立PLS-DA模型,通过构建得分图比较发生ADR和未发生ADR患者的整体差异,构建载荷图筛选ADR关键影响因素.结果 当患者同时具有外伤手术史、药物过敏史、低体质量、高龄等特点时,ADR发生的风险增高;使用葡萄糖注射剂或葡萄糖氯化钠注射剂作溶媒且溶媒体积较大时,ADR发生的风险降低.结论 本研究证明了PLS-DA算法在参麦注射剂ADR关键影响因素研究中的有效性,可为中药注射剂ADR研究提供新的方法.“,”AIM To investigate the critical effect factors of adverse drug reaction (ADR) induced by Shenmai injection by partial least squares discriminant analysis (PLS-DA) algorithm.METHODS The clinical manifestation of patients after using Shenmai injection was monitored.The information of patients and medication was also extracted.Based on the information,the multivariate data matrix was built and the PLS-DA model was then established.The scores plot was applied to investigate the overall differences between the groups experienced ADR and the other without experienced ADR.The variables with most significant variation between 2 groups were screened according to the loading plot.RESULTS The results showed that there was higher risk of ADR for patients who had such characteristics as history of trauma surgery,history of drug allergy,low body weight and advanced age.It was conducive to reducing the risk of ADR by subjecting Shenmai injection with large volume of glucose injection or glucose and sodium chloride injection as dissolvant.CONCLUSION This case study demonstrates that it is effective to employ PLS-DA algorithm for analyzing the critical effect factors of ADR induced by Shenmai injection.This research provides a new scientific tool for the investigation of ADR induced by traditional Chinese medicine injections.