论文部分内容阅读
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 investigate the feasibility of ARIMA in predicting the clinical use of apheresis platelets and provide a reference for the development of apheresis platelet collection plan and the recruitment of blood donors. Methods The clinical data of blood collected from center of Foshan city from 2007 to 2015 were collected. The ARIMA model was established with clinical data of 2007-2014 by using SPSS software. The monthly apheresis platelet clinical dose was predicted in 2015 . Results The model was established as ARIMA (1,1,1). The average relative error of prediction was 5.56% and the prediction effect was good. Conclusion The ARIMA model can well fit the clinical usage sequence of apheresis plateau and make short-term effective prediction. It can provide data support for the development of apheresis platelet collection program and recruitment of blood donors.