基于ARIMA模型的产房分娩情况预测研究

Prediction research on delivery conditions in delivery rooms based on ARIMA model

目的：建立医院分娩情况统计的预测模型,验证模型的可行性,为医院决策提供依据。方法：应用时间序列ARIMA模型对2005-2012年湖南省妇幼保健院分娩情况进行预测评价。建立ARIMA（1,1,1）（0,1,1）预测模型,预测湖南省妇幼保健院2013-2014年的分娩情况。结果：ARIMA（1,1,1）（0,1,1）模型具有很强的趋势性、周期性、季节性和规律性,下半年的分娩数明显高于上半年,每年的分娩数峰值一般出现在8月和9月,第1、2季度向上的趋势性较弱,第3、4季度向上的趋势性较强。结论：ARIMA（1,1,1）（0,1,1）模型为医院决策者对产房分娩量及工作情况进行合理安排提供了科学依据。

Objective： To establish a statistical predictive model of hospital childbirth,validate the feasibility of the model,and provide a basis for decision- making for the hospital. Methods： ARIMA time series model was used to predict and evaluate the delivery conditions in the hospital from 2005 to 2012. ARIMA（ 1,1,1）（ 0,1,1） predictive model was established to predict the delivery conditions in the hospital from 2013 to 2014. Results： ARIMA（ 1,1,1）（ 0,1,1） model had a strong trend,periodicity,seasonality and regularity,the number of delivery in the second half of this year was significantly higher than that in the first half of this year,the peak appeared in August and September; the trends in the first quarter and the second quarter were relatively weak,while the trends in the third quarter and the fourth quarter were relatively strong. Conclusion： ARIMA（ 1,1,1）（ 0,1,1） model provide scientific basis for hospital policymakers to take reasonable arrangement for delivery quantity and work.