作者：杨帆 秦银河 刘丽华 Yin-he Li-hua时间序列模型门诊人次建立预测模型应用decisionmakingarimamodel数据identificationhospitalmanagement相对误差predictionenvironmentanddatacollectionsupportsystemrelativeerrortimesequence验证模型决策支持系统患者满意度医院管理
摘要：目的 探讨ARlMA模型在门诊人次预测中的应用,阐述建模过程,建立预测模型,验证模型的适用性,为医院管理决策服务.方法 数据源于HIS集成统计与管理决策支持系统门诊报表,采集范围选自1999年～2005年逐月门诊人次数据,其中1999年～2004年各月数据用于建立时间序列模型,2005年数据用于验证所建立的模型,统计软件用SPSS13.0完成.结果 通过模型识别、参数估计、检验诊断、模型评价,建立ARIMA(1,0,1)(0,1,1)12模型,具有较高地拟和精度,全年门诊人次相对误差是6.84%,各月相对误差在-3.15%～9.80%之间.实际值都在预测的95%上下限范围之内.讨论 本研究验证了ARIMA模型适用于门诊人次预测,同时在预测门诊人次时也要考虑到数据量、就医环境、患者满意度等因素. Abstract： Objective of the study is to explore how to apply ARIMA model in prediction of outpatients headcount,describe the modeling process,build the prediction model,and verify its applicability to serve decision making for hospital management.Methods Data originates from the outpatient statements of the HIS integrated statistics and management decision making support system.The data collection ranges from the monthly outpatients headcount from 1999 to 2005,in which the monthly data from 1999 to 2004 were used to build the time sequence model,and those of 2005 to verify the model so built.The statistic software was programmed with SPSS13.0.Results The ARIMA(1,0.1)(0,1,1)12 model was built by megns of model identification,parameter estimate.inspection/diagnosis,and model appraisal.This model features high fitting precision,as the relative error or outpatients headcount for the year is 6.85%,and that for the months ranges from-3.15%to-9.80%,with the actual values falling within the 95%upper and lower thresholds of the prediction results.Conclusion This study proves that ARIMA model fits the purpose of outpatients headcount prediction.In the meantime,prediction of such headcount
注：因版权方要求，不能公开全文，如需全文，请咨询杂志社
