2019—2021年重庆市急救医疗中心急性酒精中毒病例流行特征及时间序列分析

Epidemiology and time series analysis of patients with acute alcoholism in Chongqing Emergency Medical Center from 2019 to 2021

目的回顾性分析重庆市急救医疗中心2019—2021年急性酒精中毒病例的流行特征和发生趋势,为建立急救机制提供理论依据。方法从重庆市急救医疗中心医疗急救办公管理系统中导出2019年1月1日至2021年12月31日期间的急性酒精中毒患者4142例的病历资料,通过Excel软件进行描述性分析,SPSS 26.0软件进行时间序列分析,建立预测模型。结果2019—2021年重庆市急救医疗中心共报告急性酒精中毒患者4142例,环比增长率逐年增高,男女性别比为3.63:1,中青年为高发群体。一天中22~23时发生的病例数最多。急性酒精中毒发病具有明显的周期性和季节性,夏季有小幅度上升,冬季达到高峰。通过自动拟合确定的最佳模型为温特斯加法指数平滑模型,模型的决定系数(R2)为0.654,正态化的贝叶斯信息准则(BIC)值为5.778,均方根误差(RMSE)、平均绝对误差(MAE)和平均绝对百分比误差(MAPE)分别为15.60、10.60和12.81%,残差序列为白噪声(P>0.05)。利用该模型对2022年1月至12月急性酒精中毒的发生情况进行预测,实际发生例数较往年升高,实际发病数大部分在预测值95%CI内。结论重庆市急性酒精中毒病例逐年增加,在夏季和冬季要加强建立对急性酒精中毒病例的急救机制。温特斯加法指数平滑模型对急性酒精中毒病例的发生趋势具有较好的拟合效果。

Objective To analyze the epidemiological characteristics and trends of acute alcoholism in Chongqing Emergency Medical Center from 2019 to 2021 and provide an evidence-based basis for the emergency mechanism.Methods The medical records of 4142 patients with acute alcoholism from January 1,2019 to December 31,2021 were exported from the medical emergency office management system of Chongqing Emergency Medical Center.Excel software was used for descriptive analysis,and SPSS 26.0 software was used for time series analysis to establish a prediction model.Results A total of 4142 records of patients with alcoholism were entered into the Chongqing Emergency Medical Center during 2019—2021.The year-on-year growth rate was increasing year by year.The male-to-female ratio was 3.63:1.Young and middle-aged people were prone to alcoholism.The majority of cases occurred from 22 and 23 o'clock in the day.The incidence of acute alcoholism showed significant periodicity and seasonality,with a slow rise in the summer and a seasonal peak in the winter.The best model identified by auto-fitting was the additive Holt-winters exponential smoothing model,with model coefficient of determination(R2)was 0.654,normal Bayesian information criterion(BIC)was 5.778,root mean squared error(RMSE),mean absolute error(MAE),and mean absolute percentage error(MAPE)were 15.60,10.60,and 12.81%,respectively,and residuals were white noise sequence(P>0.05).Using the model to predict the occurrence of alcoholism from January to December 2022,the actual number of cases was elevated compared to previous years and mostly within the 95%confidence interval.Conclusion The incidence of acute alcoholism in Chongqing shows an upward tendency over time.More attention should be paid to strengthening the first aid mechanism for alcoholism cases in the summer and winter.The additive Holt-winters exponential smoothing model could well fit the incidents of acute alcoholism.