南京市呼吸道疾病时间序列模型的建立与分析

Establishment and Analysis of Time Serial Models for Respiratory Diseases

利用1990～2000年南京市支气管哮喘与感冒逐月就诊人数病例资料和同期国家气象中心资料室的月平均压、温、湿等资料,讨论了2种疾病发病人数的月际和年际变化规律,分别建立了包含气象因子的预报2种疾病逐月就诊人数的条件异方差和动态回归模型.南京地区支气管哮喘就诊人数高峰期出现在春末与秋季,逐年的就诊人数呈现上升趋势,其发病与当月月平均气压、上一个月月平均温度有关.感冒就诊人数高峰期出现在夏秋季,其发病与月平均最低温度和月极端最低温度有关.将模型应用于2种疾病逐月就诊人数的等级预报中,其预报等级准确率分别达75.0%和66.7%,表明此建模方案是可行的,此类模型有一定的使用价值.

By use of the month-by.month hospital admissions caused by asthma and the influenza and monthmean pressure, temperature, and humidity data in Nanjing, the occurrence regularities of the two diseases were discussed. Two time serial models including meteorological factors were established to predict the number of patients on a monthly basis. Results show that in the recent 10 years, hospital admissions caused by asthma increased year after year and asthma is prone to occur at autumn and the end of spring. The occurrence of asthma is correlated with the mean pressure of the same month and the mean temperature of previous month. However, the influenza is more likely to occur in summer and autumn and its occurrence is related to the month mean minimum temperature and the month＇s lowest temperature. The conditional heteroskedasticity model and ARIMAX model were established to predict the number of patients caused by asthma and influenza, respectively, The prediction veracities for asthma and influenza are 75.5 percent and 66.7 percent, respectively, which indicates that the methods are feasible and applicable in application.