目的 建立消化道和呼吸道传染病患者首诊医疗机构选择预测模型并分析其影响因素,为医疗机构精准防治传染病提供参考.方法 通过整群抽样方法,抽取北京市通州区某医联体覆盖区域内1 524例消化道和呼吸道传染病患者.将患者数据按照8∶2随机分成训练集和测试集,采用二元logistic回归模型分析患者首诊医疗机构选择的影响因素并建立预测模型,通过测试集中受试者工作特征(receiver operating characteristic,ROC)曲线下的面积(area under the curve,AUC)评价模型预测效果.结果 1 524例患者中,有417例选择社区首诊.文化水平、医保类型、职业、相对距离、是否签约家庭医生是传染病患者选择首诊医疗机构的影响因素.测试集的AUC=0.851,模型拟合效果较好.结论 通过建立传染病患者首诊医疗机构预测模型可用来预测患者流向,帮助区域合理配置医疗资源,提高区域内传染病患者基层首诊率.
Influencing Factors and Predictive Modeling of the Selection of Infectious Disease Patients'First Choosing Institution
Objective To establish a predictive model for digestive and respiratory infectious disease patients'selection of first-entry healthcare institutions and analyze the influencing factors,providing a reference for regional healthcare institution development.Methods The cluster sampling was used to select 1524 target patients.The data was randomly split 8∶2 into training and testing sets.The binary logistic regression method was employed to establish a predictive model for the selection of infectious disease patients'first choosing institution and analyze the influencing factors.The model's predictive performance was evaluated using the AUC of the testing set.Results It analyzed 1 524 infectious disease cases;417 chose community clinics first.Influences on their initial choice included education,insurance type,occupation,relative distance,and having a family doctor.The AUC of the test set was 0.851.Conclusion The predictive model of the selection of infectious disease patients'first choosing institution established in this study can be used to predict patient flow and help allocate medical resources reasonably.Increasing the rate of primary care for patients with infectious diseases in the region.
infectious diseasefirst choosing institutioninfluencing factorspredictive model
NETL
NSTL
万方数据
