ARIMA季节性模型在海口市院外心脏骤停发病预测中的应用  

Application of ARIMA seasonal model in predicting the incidence of out-of-hospital cardiac arrest in Haikou City

作　　者：孙佳乐 梅媛[2] 徐百超 周凡 张华[2] SUN Jiale;MEI Yuan;XU Baichao;ZHOU Fan;ZHANG Hua(International School of Public Health and One Health,Hainan Medical University,Haikou 571199,Hainan Province,China;International Nursing College of Hainan Medical University,Haikou 571199,Hainan Province,China;Hainan Key Laboratory of Sports and Health Promotion,Haikou 571199,Hainan Province,China)

机构地区：[1]海南医学院公共卫生与全健康国际学院,海口571199 [2]海南医学院国际护理学院,海口571199 [3]海南省运动与健康促进重点实验室,海口571199

出　　处：《预防医学情报杂志》2024年第12期1479-1484,1492,共7页Journal of Preventive Medicine Information

基　　金：国家自然科学基金项目(项目编号:81960420);海南省重点研发项目(项目编号:ZDYF2022SHFZ306);海南省自然科学基金项目(项目编号:822MS071);海南省高等学校教育教学改革研究项目(项目编号:Hnjgs2022-8);海南医学院校级教学成果奖培育入库项目(项目编号:HYjcpx202202);海南省普通高等学校研究生创新科研课题(项目编号:HYYS2022B10)。

摘　　要：目的建立海口市院外心脏骤停发病的自回归移动平均模型(autoregressive integrated moving average model,ARIMA),探讨该模型在预测海口市院外心脏骤停发病趋势中应用的可行性,为其医疗资源的配置提供依据。方法收集海口市某急救中心2015年1月至2021年12月院外心脏骤停的病例,以每月发生人次进行整理,采用统计软件R 4.2.3以2015—2021年海口市院外心脏骤停病例数据为基础建立季节性ARIMA模型,并用2022年1至12月的实际发病数进行验证,利用最优模型预测2023年至2024年8月海口市院外心脏骤停发病人次。检验水准为α=0.05。结果最优模型为ARIMA(0,0,2)(0,1,1)_(12),参数均有统计学意义(P均<0.05),AIC=514.58,AICc=515.49,BIC=525.96,残差序列为白噪声(Ljung-Box=0.030,P=0.861),拟合效果较好。2015—2022年院外心脏骤停人次逐年呈上升趋势,每年的1、2、3、11和12月人次均高于其余月份;经预测,2023年至2024年8月海口市院外心脏骤停人次与以往变化趋势一致。结论ARIMA(0,0,2)(0,1,1)12模型拟合海口市院外心脏骤停的发病人次效果较好,可用于海口市院外心脏骤停发病趋势的短期预测。Objective To establish an autoregressive integrated moving average model(ARIMA)for the incidence of out-of-hospital cardiac arrest(OHCA)in Haikou City,and to explore the feasibility of the model in predicting the incidence trend of out-of-hospital cardiac arrest in Haikou City,so as to provide a basis for the allocation of medical resources.Methods From January 2015 to December 2021,the cases of out-of-hospital cardiac arrest in an emergency center in Haikou City were collected,and the number of cases per month was sorted out.The statistical software R4.2.3 was used to establish a seasonal ARIMA model based on the data of out-of-hospital cardiac arrest in Haikou City from 2015 to 2021,α=0.05.The results were verified by the actual number of OHCA cases from January to December 2022,and the optimal model was used to predict the number of OHCA cases in Haikou from 2023 to August 2024,α=0.05.Results The optimal model was ARIMA(0,0,2)(0,1,1)_(12),the parameters were statistically significant(P<0.05),AIC=514.58,AICC=515.49,BIC=525.96,and the residual sequence was white noise(Ljung-Box=0.030,P=0.861),the fitting effect was good.From 2015 to 2022,the number of people with out of hospital cardiac arrest increased year by year,and the number of people in January,February,March,November and December was higher than that in other months;It was predicted that the number of people with cardiac arrest outside the hospital in Haikou in 2023 was consistent with the previous trend.Conclusions ARIMA(0,0,2)(0,1,1)12 model has a good effect in fitting the number of people with out of hospital cardiac arrest in Haikou,which can be used for the shortterm prediction of the incidence trend of out of hospital cardiac arrest in Haikou.

关 键 词：ARIMA模型 院外心脏骤停 发病预测 模型应用 

分 类 号：R181.32[医药卫生—流行病学]