摘要
目的基于季节性指数平滑模型,预测肺癌患者出院人次的变化趋势,探讨季节性指数平滑模型在出院人次预测中的适用性和实际应用价值。方法收集上海市某区2014年1月1日-2020年12月31日间84个月的肺癌出院病案数据,涵盖二三级公立医院、社区卫生服务中心和民办医疗机构290.80万出院人次数据。运用SPSS21.0进行统计分析,应用2014年1月1日-2019年12月31日的数据构建季节性指数平滑模型,用2020年实际数据验证模型预测结果。结果比较多种季节性指数平滑模型拟合结果,发现Winters加法模型预测效果最好。该模型Ljung-Box Q检验P=0.235,模型Stationary R^(2)和R^(2)分别为0.536和0.906,标准化BIC为7.191,RMSE和MAPE分别为33.324和5.49%。结论季节性指数平滑模型能够对肺癌出院人次进行科学合理的预测,可为决策制定、资源配置和工作评估提供参考依据。
Objective This paper predicted the discharge person-time of lung cancer based on seasonal exponential smoothing model,in oder to discuss the application of this model in prediction of discharge person-time and its practical value.Methods Based on the medical record database of main medical institutions,collected the data of the discharge person-time of lung cancer from January 1,2014 to December 31,2020,in a District of Shanghai,covers 2.908 million discharge data from second-and third-tier public hospitals,community health service centers and private medical institutions.Established seasonal exponential smoothing models with the data during January 1,2014 and December 31,2019 by using the SPSS21.0.Predicted the discharge person-time of lung cancer from January to December 2020,and tested its veracity.Results Winters addition model achieved relatively high goodness of fit in various seasonal exponential smoothing models.The P value of the Ljung-Box Q test was 0.235(P>0.05).Stationary R^(2) was 0.536,R^(2) was 0.906,BIC was 7.191,RMSE was 33.324 and MAPE was 5.49%.Conclusions Winters addition model could scientifically and reasonably predict discharge person-time,and provide significant references for making decisions,allocating resources,evaluating work and drafting plans.
作者
王军伟
楚天舒
荆丽梅
Wang Junwei;Chu Tianshu;Jing Limei(Pudong Institute for Health Development,Shanghai 200129,China;不详)
出处
《中国病案》
2021年第12期56-60,共5页
Chinese Medical Record
基金
教育部人文社会科学研究规划基金(20YJAZH045)
上海市哲学社会科学规划课题(2019BGL032)
基于大数据的肺癌患病率动态分析及趋势预测研究-以浦东新区为例(PW2017A-1)。
关键词
季节性指数平滑模型
预测分析
肺癌出院人次
Seasonal exponential smoothing model
Prediction analysis
Discharge person-time of lung cancer
