基于贝叶斯网络的医疗设备临床应用风险评估研究  

Risk Evaluation Method for the Clinical Application Risk of Medical Devices Based on the Bayesian Network

作　　者：程菊[1] 刘胜林[1] 魏建新[1] 江明尹 冯庆敏[1] 张建洋[1] 高佳琦[1] 张强[1] 

机构地区：[1]华中科技大学同济医学院附属协和医院,湖北武汉430022

出　　处：《中国医院管理》2016年第5期78-80,共3页Chinese Hospital Management

基　　金：国家卫生和计划生育委员会医政医管局基金项目

摘　　要：为提高医疗设备临床应用安全水平,在没有充足样本数据情况下,提出了一种基于贝叶斯网络的医疗设备临床应用风险评估方法。风险评估方法首先基于人、机、环境和管理系统理论,使用德尔菲法设计医疗设备临床应用风险评估指标体系;然后根据构建的指标体系手动建立贝叶斯网络拓扑结构;在网络节点变量满足Ranked nodes方法的前提下,利用专家判断确定网络结构中的概率参数;选用Agena Risk概率推理软件建立贝叶斯网络风险评估模型。将所建模型应用于某三级甲等医院ICU在用呼吸机的风险评估,可获得其不同风险等级发生概率,有效提高医疗设备风险管理效率。In order to improve the safety level of medical devices, a risk evaluation method based on Bayesian network was presented in the absence of sufficient sample data. Firstly, according to the human, machine, environment and management system theory and Delphi method, an index system about the clinical application risk of medical devices was set up. The Bayesian network topology was built by using the index system, and the parameters of the network structure were obtained based on expert judgment when the network nodes can meet the requirements of ranked nodes method. Then, bayesian network risk evaluation model was established by using Agenarisk software. Finally, the model was applied to the risk analysis of the in-use ventilators in the ICU of a hospital, which can quantitatively evaluate the probability of different risk levels. It can effectively improve the efficiency of medical devices risk management.

关 键 词：医疗设备 贝叶斯网络 指标体系 风险评估 

分 类 号：R197.39[医药卫生—卫生事业管理]