目的 探讨时间序列分析方法 在医院医用低值耗材管理中的应用,分析和预测未来一段时间内医用低值耗材的使用需求.方法采用自回归积分滑动平均模型(Autoregressive Integrated Moving Average,ARIMA)对上海某三甲医院医用低值耗材2008年1月~2016年7月的逐月使用量进行数据挖掘及预测.结果 ARIMA(6,1,3)(1,1,1)模型的绝对误差为8.58%,在实际业务可接受范围之内,因此模型拟合效果较好,预测结果接近实际产生值.结论 ARIMA(6,1,3)(1,1,1)模型能够准确的进行医用低值耗材的短期预测,并将其应用于医院物资管理信息系统中,系统根据预测值,合理生成申领、采购计划,实现对医院耗材的合理管控,并为科室医用耗材的资金预算申请提供可靠依据.%Objective Through application of time series analysis in hospital low-cost medical consumables management, to analyze and forecast the demand of low-cost medical consumables in the future. Methods Autoregressive integrated moving average (ARIMA) model was used for data mining and forecasting for monthly data of low-cost medical consumables from January 2008 to December 2016 in Shanghai Ruijin Hospital. Results The absolute error of the model was 8.58%,which controlled in a certain range. ARIMA (6,1,3) (1,1,1) had a good fitting effect, the predicted result was close to the actual value. Conclusion ARIMA (6,1,3) (1,1,1) model could accurately predict low-cost medical consumables in the short-term, and apply it to the hospital consumables management information system. The system can reasonably generate the requisition and purchase plan according to the predicted value, realize the reasonable control of the hospital consumables, and provide a reliable basis for funding budget applications.
展开▼
