摘要
目的探讨自回归积分滑动平均模型(ARIMA模型)在职业性噪声聋(ONID)发病趋势预测中的应用价值。方法以广东省2006—2015年新发ONID病例数建立ARIMA模型,对之进行验证,并采用该模型预测2016—2020年新发ONID的发病趋势。结果广东省2006—2015年新发ONID病例数呈指数式增长趋势。采用2006—2015年新发ONID病例数拟合的最优模型为ARIMA(2,2,2)模型,可较好拟合2008—2015年的新发ONID病例数。根据ARIMA(2,2,2)模型预测,2016—2020年广东省年新发ONID仍呈快速增长趋势。结论基于时间序列的ARIMA模型可较好地拟合ONID发病的时间变化趋势,可用于ONID的发病趋势预测。
Objective To explore the application of the autoregressive integrated moving average model（ ARIMA model）in predicting incidence of occupational noise-induced deafness（ ONID）. Methods The ARIMA model was established and validated based on the number of new onset ONID cases in Guangdong Province from 2006 to 2015. Then the ARIMA model was used to predict the trend of new onset ONID cases from 2016 to 2020. Results The number of new ONID cases in Guangdong Province from 2006 to 2015 showed an exponential growth trend. The optimal model fitted with the number of new onset ONID cases from 2006 to 2015 was the ARIMA（ 2,2,2） model,which better match the number of new onset ONID cases from 2008 to 2015. According to the ARIMA（ 2,2,2） model,the number of new onset ONID cases in Guangdong Province will continue to have a rapidly increasing trend from 2016 to 2020. Conclusion The ARIMA model based on time series matches the time trend of ONID onset,and it can be used for the prediction of ONID incidence trend.
作者
李旭东
瞿红鹰
胡世杰
余宏伟
温贤忠
杨爱初
戚亚洲
陈琳
LI Xudong;QU Hongying;HU Shijie;YU Hongwei;WEN Xianzhong;YANG Aichu;QI Yazhou;CHEN Lin(Guangdong Province Hospital for Occupational Disease Prevention and Treatmen;Guangdong Provincial Key Laboratory of Occupational Disease Prevention and Treatment, Guangzhou, Guangdong 510300, Chin)
出处
《中国职业医学》
CAS
北大核心
2018年第2期164-167,共4页
China Occupational Medicine
基金
国家科技支撑计划项目(2014BAI12B01)
广东省科技计划省国际合作项目(2011B050700001)
广东省科技计划项目(2013B021800176)
广东省医学科研基金(A2015150
C2016014)
广东省职业病防治重点实验室(2017B030314152)
广东省省级工业与信息化发展专项资金互联网+技术应用项目
广州市医药卫生科技项目(20151A010076)
关键词
噪声聋
职业病
ARIMA模型
时间序列
预测
发病趋势
Noise-induced deaihess
Occupational disease
ARIMA model
Time series
Predication
Incidence trend
