借助自回归移动平均模型构建儿童口服退烧药消耗量的预测模型
Prediction model of children's oral antipyretic drug consumption by autoregressive moving average model
王卓芸 1于飚 1陶亮亮 2蔡和平1
作者信息
- 1. 安徽省儿童医院临床药学部,合肥 230000
- 2. 安徽省儿童医院信息中心,合肥 230000
- 折叠
摘要
目的 使用自回归移动平均(ARIMA)模型对医院口服退烧药消耗量进行预测,旨在为医院药品供应提供参考.方法 收集某儿童医院2013年1月—2018年12月全院每月口服退烧药的消耗量建立ARIMA模型,预测2019年1-12月口服退烧药的消耗量,以2019年1-12月实际消耗量数据评价预测模型是否合适.结果 2013年1月—2018年12月每月口服退烧药消耗量的时间序列呈现明显的季节性,在每年1月、5-7月及12月出现消耗量峰值.该时间序列经季节性分解后拟合ARIMA(1,1,1)(1,1,1)12模型,拟合效果较好,经过对比发现除2019年1月和2019年12月两个月外,其余月份预测值与实际值的相对误差均在20％以内.结论 依照2013年1月—2018年12月全院每月口服退烧药消耗量建立的ARIMA模型能够较好地预测口服退烧药消耗量,可为医院科学的药品供应提供一定参考.
Abstract
Objective To predict the consumption of oral antipyretic drugs in hospitals by autoregressive integrated moving average(ARIMA)model,and to provide reference for drug supply in hospitals.Methods The monthly consumption of oral antipyretic drugs in the hospital from January 2013 to December 2018 was collected to establish an ARIMA model and predict whether the actual consumption of oral antipyretic drugs from January to December 2019 fit the model.Results The time series of consumption of oral antipyretic drugs from 2013 t0 2018 showed obvious seasonal features,with peaks in January,May-July and December each year.The sequence fit the ARIMA(1,1,1)(1,1,1)12 model well after seasonal decomposition.The relative error between the predicted value and the actual value in other months was less than 20％except for January and December 2019.Conclusion The consumption of oral antipyretic drugs and drug supply could be predicted by the ARIMA model.
关键词
自回归移动平均模型/口服退烧药消耗量/药品供应Key words
autoregressive integrated moving average model/oral antipyretic drug consumption/drug supply引用本文复制引用
出版年
2024
NETL
NSTL
万方数据