摘要
目的探讨组合预测模型在低值医用耗材管理中的应用,预测未来一段时间内低值医用耗材的库存需求。方法以杭州某三甲医院低值医用耗材2017年1月至2018年12月的历史领用数据为初始建模数据,采用基于二次指数平滑模型和灰色马尔科夫模型,分别利用变异系数法、简单加权法和方差倒数法来确定组合权重的组合预测模型,对低值医用耗材2019年1—4月的库存需求建立预测模型。结果利用三种组合预测模型分别计算得到的预测值的相对误差均值都远小于5%。其中使用方差倒数法确定权重的组合预测的预测准确度最高,其预测值的相对误差均值为1.39%,接近实际值。结论由方差倒数法确定权重的基于二次指数平滑模型和灰色马尔科夫模型的组合模型能够准确地对低值医用耗材进行短期预测,根据组合模型的预测值,可以为低值医用耗材的库存准备提供可靠依据。
Objective To discusses the application of combined forecasting model in the management of low value medical consumables,and predicts the inventory demand of low value medical consumables in the future.Methods Based on the historical acquisition data of low value medical consumables from January 2017 to December 2018 in a first-class hospital at grade 3 in Hangzhou,a combined forecasting model based on quadratic exponential smoothing model and grey Markov model was used to determine the combined weight by using the variation coefficient method,simple weighting method and the inverse variance method,respectively.A forecast model was established for the inventory demand of low-value medical consumables from January to April 2019.Results The mean value of relative error calculated by the three combined forecasting models was far less than 5%.Among them,the prediction accuracy of the combination prediction was the highest,using the reciprocal variance method to determine the weight.The relative error mean of the prediction value was 1.39%,which was close to the actual value.Conclusion The combination model based on quadratic exponential smoothing model and grey Markov model,which is determined by the reciprocal variance method,can accurately predict the short-term value of low-value medical consumables.According to the prediction value of the combination model,it can provide a reliable basis for the inventory preparation of low-value medical consumables.
作者
张林灵
郑焜
ZHANG Linling;ZHENG Kun(Department of Medical Equipment,The Children’s Hospital of Zhejiang University School of Medicine,Hangzhou Zhejiang 310006,China)
出处
《中国医疗设备》
2021年第8期120-123,共4页
China Medical Devices
关键词
库存需求预测
二次指数平滑预测
灰色马尔科夫预测
组合预测模型
inventory demand forecast
quadratic smoothing prediction
grey Markov prediction
combined forecasting model
