基于气象因素的手足口病发病风险预测模型

目的 探讨我国手足口病发病与气象因素的关系,建立基于气象因素的手足口病发病风险预测模型.方法 收集2010-2014年全国31个省市手足口病月发病率资料和当地同期气象资料;利用Spearman相关分析、Logistic回归分析筛选影响手足口病发病的气象危险因素;应用哈佛疾病风险指数模型建立预测模型;采用ROC曲线下面积评价模型优劣.结果 影响手足口病高发的气象危险因素分别为月平均气温12 ～22℃(OR =13.922)、月平均气温大于22℃(0R =25.932)、标准大气压地区(OR=7.953)、月降水量大于100 mm(OR=1.439)及月平均湿度大于70％ (OR=1.597).根据以上气象因素建立手足口病发病风险预测模型,并绘制风险评估图,模型的ROC曲线下面积为0.810.结论 气象因素对手足口病的发生、发展过程存在影响;成功建立了基于气象因素的手足口病发病风险预测模型.

Development of a risk prediction model for hand-foot-mouth disease based on meteorological factors

Objective To explore the relationship between the outbreak of hand-foot-mouth disease (HFMD) and meteorological factors in China and develop a risk prediction model for HFMD.Methods The data of monthly HFMD incidences in 31 provinces(cities) and the local meteorological data from 2010 to 2014 were collected.Spearman correlation analysis and logistic regression were used to identify the meteorological risk factors for HFMD,and the risk prediction model incorporating these meteorological variables was developed based on Harvard disease risk index model.The performance of this model in predicting the risk of HFMD was assessed using the area under the ROC curve.Restults The meteorological risk factors for HFMD includes the monthly average temperature of 12 to 22℃ (OR =13.922),monthly average temperature above 22 ℃ (OR =25.932),a region with a normal atmospheric pressure (OR =7.953),monthly rainfall above 100 mm(OR =1.439) and monthly average humidity above 70％ (OR =1.597).The risk prediction model for HFMD established using these meteorological variables had an area under the ROC curve of 0.810.Corcltusior Meteorological factors can influence the risk of HFMD,and this risk prediction model may provide assistance in establishing an early warning system for HFMD outbreak and can have practical values in HFMD prevention and control.