差分—指数平滑预测模型在门诊人次预测中的应用  

The Application of Difference-Index Smoothing Predictive Model in Predicting the Person-time of the Outpatients

作　　者：辛珏[1] 季江川[2] 

机构地区：[1]四川省成都市第九人民医院.妇产科医院 [2]四川省成都市第三人民医院

出　　处：《中国病案》2009年第5期32-33,共2页Chinese Medical Record

摘　　要：目的探讨差分—指数平滑预测模型在医院管理中的应用价值。方法建立差分—指数平滑预测模型对某院2008年门诊诊次进行定量预测,并对预测模型的预测精度进行评价。结果差分—指数平滑预测模型预测门诊诊次的结果满意。通过计算机反复试算,可获得最佳的平滑常数(α),使预测模型的预测误差最小。本例预测平均绝对百分误差(MAPE)4.4040%<10%,预测精度较高。结论差分—指数平滑预测模型预测精度较高,计算简便且具有递推性质,在医院管理中有应用价值。Objective To discuss the application of difference - index smoothing predictive model in hospital management. Methods We quantitatively predict the person - time of the outpatients in 2008 by the establishment of difference - index smoothing predictive model. Results The result of prediction with difference- index smoothing predictive model is satisfactory, and we can get optimal smoothing constant （α） and minimize the errors in prediction by repeated trial calculation. The mean absolute percent error （MAPE） was - 3.9080% 〈 10%, which means that the predictive accuracy is high. Conclusions Difference - index smoothing predictive model is of high predictive accuracy, easy calculation, recurrence properties and application value in hospital management.

关 键 词：差分-指数平滑模型 预测精度 预测误差 门诊诊次 卫生服务需求 

分 类 号：R197.3[医药卫生—卫生事业管理]