河北省基于广义相加模型的肾综合征出血热发病预测研究  

Study of prediction of hemorrhagic fever with renal syndrome incidence in Hebei Province based on generalized additive model

作　　者：岳中航 韩旭[1] 魏亚梅 蔡亚男[1] 韩占英[1] 张艳波[1] 许永刚[1] 李琦[2] Yue Zhonghang;Han Xu;Wei Yamei;Cai Yanan;Han Zhanying;Zhang Yanbo;Xu Yonggang;Li Qi(Virial Disease Prevention and Control Institute,Hebei Provincial Center for Disease Control and Prevention,Shijiazhuang 050021,China;Hebei Provincial Center for Disease Control and Prevention,Shijiazhuang 050021,China)

机构地区：[1]河北省疾病预防控制中心病毒病防治所,石家庄050021 [2]河北省疾病预防控制中心,石家庄050021

出　　处：《中华流行病学杂志》2025年第3期418-422,共5页Chinese Journal of Epidemiology

基　　金：河北地区病毒性传染病病原谱和病毒基因变异变迁规律研究(2018ZX10713002-001-005)。

摘　　要：目的利用广义相加模型(GAM)对河北省肾综合征出血热(HFRS)月发病数进行预测。方法收集河北省2006-2020年HFRS的发病资料,通过Spearman相关分析气象因素与河北省HFRS月发病之间的相关系数,并将气象因素滞后0~6阶,筛选相关系数绝对值最大的因素纳入多因素GAM中,探讨气象因素的效应。结果HFRS月发病数与滞后2阶的月平均气温、滞后0阶的月平均风速、滞后4阶的月平均日照、滞后2阶的月平均降水、滞后1阶的月平均湿度相关性最强,经过方差膨胀因子诊断后将其纳入多因素GAM中,各因素差异有统计学意义(均P<0.001),且与HFRS月发病数均呈明显的非线性关系。月平均气温是影响发病的重要因素。月平均气温、月平均日照和月平均风速与HFRS的发病呈负相关,月平均降水和月平均湿度与HFRS的发病呈正相关。结论各气象因素与HFRS的发病之间存在复杂的非线性关系。纳入滞后气象因素的GAM可用于河北省HFRS的发病预测。Objective To predict the monthly incidence of hemorrhagic fever with renal syndrome(HFRS)in Hebei Province by using the generalized additive model(GAM).Methods The incidence data of HFRS in Hebei from 2006 to 2020 were collected,and the correlation coefficients between meteorological factors and the monthly incidence of HFRS in Hebei were analyzed by Spearman's correlation,and the meteorological factors were lagged by 0-6 orders,and those with the largest absolute values of the correlation coefficients were screened to be included in the multifactorial GAM to evaluate the effects of meteorological factors.Results The monthly incidence of HFRS had the strongest correlation with monthly mean air temperature at lag order 2,monthly mean wind speed at lag order 0,monthly mean sunshine at lag order 4,monthly mean precipitation at lag order 2 and monthly mean humidity at lag order 1,which were diagnosed by the variance inflation factor and included in the multifactorial GAM,and the results showed significant differences among the factors(all P<0.001),and they showed non-linear relationships with the monthly incidence of HFRS.Mean monthly temperature was an important factor influencing HFRS incidence.Mean monthly air temperature,mean monthly sunshine and mean monthly wind speed were negatively associated with HFRS incidence,whereas mean monthly precipitation and mean monthly humidity were positively associated with HFRS incidence.Conclusions There was a complex non-linear relationship between meteorological factors and the incidence of HFRS.GAM incorporated with lagged meteorological factors can be used to predict the incidence of HFRS in Hebei.

关 键 词：肾综合征出血热 气象因素 广义相加模型 发病预测 

分 类 号：R51[医药卫生—内科学]