ARIMA模型在单采血小板临床用量预测中的应用  

Application of autoregressive integrated moving average model in predicting the clinical usage of apheresis platelet

作　　者：马春会[1] 余晋林[1] 温丽玲[1] 郭如华[1] 余卓丽[1] 罗益红[1] 

机构地区：[1]佛山市中心血站,广东佛山528000

出　　处：《中国输血杂志》2016年第12期1394-1396,共3页Chinese Journal of Blood Transfusion

摘　　要：目的探讨自回归移动平均模型(ARIMA)预测单采血小板临床用量的可行性,为单采血小板采集计划的制定和献血者的招募提供参考。方法采集佛山市中心血站2007-2015年单采血小板每月的临床用量数据,用SPSS软件对2007-2014年的临床用量数据建立ARIMA模型,对2015年每月的单采血小板临床用量进行预测。结果建立的模型为ARIMA(1,1,1),预测平均相对误差为5.56%,预测效果较好。结论 ARIMA模型能较好地拟合单采血小板临床用量序列并进行短期的有效预测,可为血站单采血小板采集计划的制定和献血者的招募提供数据支持。Objective To assess the feasibility of the autoregressive moving average （ARIMA） model in predicting the clinical usage of apheresis platelet, and provide a reference for making platelet collecting plans and donor recruitment. Meth- ods The monthly usage dates of apheresis platelets provided by Foshan Central Blood Station from 2007 to 2015 were taken, and SPSS software was used to set up a ARIMA model with dates from 2007 to 2014. Then, the model was used to predict the clinical usage of apheresis platelet of 2015. Results The established apheresis platelet forecasting model is ARIMA （ 1,1, 1 ）. The predicted values calculated by the ARIMA model and the actual values fit well, and the mean relative error was 5. 56%. Conclusion The ARIMA model can be used to predict the clinical usage of apheresis platelets in a short time, and to provide information for platelet collecting plans and donor recruitment.

关 键 词：自回归移动平均模型(ARIMA) 单采血小板 预测 

分 类 号：R457.1[医药卫生—治疗学] R331.143[医药卫生—临床医学]