基于人群特征的药源性心律失常自动监测模块优化及应用研究

Optimization and application of an automatic monitoring module for drug-induced arrhythmias based on population characteristics

目的对“临床药物不良事件(adverse drug event,ADE)主动监测与智能评估警示系统-Ⅱ”的心律失常自动监测模块进行功能开发与优化,持续完善性能,提升监测效率,探索模块优化方式。方法拓展优化模块功能,增加模块自定义配置内容,并确定最优设置条件;对比优化后测试数据与前期大样本用药人群中药源性心律失常自动监测评价研究结果,验证模块优化程度和准确性。结果根据监测人群特征优化后的新模块,增加了“强制医嘱关键词”功能,并确定纳入6个心电图检查相关的医嘱关键词、且出现频率不少于2次为模块优化最佳配置条件;结合前期自动监测评价研究结果,开展全药和2种单药条件下系统功能验证和比较。在真阳性病例零丢失的同时,系统报警的病例数较模块优化前分别下降了30.75%、80.13%和90.82%,阳性预测值得到了显著提升。结论经过功能拓展与优化完善的心律失常自动监测模块显著降低了病例评价人工成本,同时保持监测结果准确度不变;新模块能够更好地适应自动监测不同模式需求且运行稳定,更具普适性和灵活性,为自动监测模块研发提供了新的思维途径。

Objective To develop the functions and optimize the automatic monitoring module for arrhythmias of the adverse drug event active surveillance and assessment system-Ⅱ,in order to continuously improve the performance,enhance the monitoring efficiency,and explore the ways to optimize the module.Methods Expand and optimize the functions of the module,increase the customized configuration,and determine the optimal setting conditions;compare the optimized test data with the results of the evaluation studies on the automatic monitoring of drug-induced arrhythmias in large samples of medicated population previously,and verify the optimization extent as well as the accuracy of the module.Results In the new module optimized according to the characteristics of the monitoring population,the function of“mandatory medical order keywords”was added,and it was determined that the inclusion of 6 electrocardiogram examination-related medical order keywords with a frequency of not less than 2 occurrences was the optimal configuration condition for the optimization of the module;combining the results of the previous automatic monitoring and evaluation researches,the system functions were verified and compared under the conditions of using the whole drugs and 2 kinds of single drug.While there was no loss of true positive cases,the number of cases with system alarms decreased by 30.75%,80.13%and 90.82%,respectively,compared with that before the optimization of the module,and the positive predictive value was significantly improved.Conclusion After the function expansion and optimization,the automatic monitoring module of drug-induced arrhythmias significantly reduces the labor cost of case evaluation and keeps the accuracy of monitoring results constant;the new module can better adapt to the demands of different automatic monitoring modes and operates stably,which is more generalizable and flexible,and provides a new way of considering for the research and development of automatic monitoring modules.