论文部分内容阅读
目的通过负二项回归模型探讨气象因素与猩红热发病的关系。方法对1985-2005年安徽省某市猩红热月平均发病率和月平均降水量、月平均气压、月平均气温、月平均相对湿度、月平均最低气温5项气象资料的数据进行描述性分析,然后拟合负二项回归模型,并且对2006年每个月份的发病率做一个预测。结果模型的超离散度κ=0.41(95%CI:0.32~0.53),进行似然比2检验,2=306.42,P<0.001,认为发现负二项回归是适合的模型。猩红热的发生与月平均气压、月平均相对湿度和月平均最低气温有统计学意义(均有P<0.05)。对2006年各个月份的月发病率预测的结果表明(Wilcoxon符号秩和检验,Z=0.24,P=0.814),预测值与实际值之间差异无统计学意义,提示预测效果比较理想。结论通过拟合负二项回归模型发现,对猩红热的发生和预测,月平均气压、月平均相对湿度和月平均最低气温是不可忽略的气象因素。
Objective To explore the relationship between meteorological factors and the incidence of scarlet fever by the negative binomial regression model. Methods Descriptive analysis was made on the data of five meteorological data of the average monthly incidence of scarlet fever, monthly average precipitation, monthly mean air pressure, monthly average temperature, monthly average relative humidity and monthly mean minimum temperature from 1985 to 2005 in Anhui Province, and then Fit a negative binomial regression model and make a prediction of the morbidity in each month of 2006. Results The hyperdispersion of the model was 0.41 (95% CI: 0.32-0.53). The likelihood ratio 2 test was used. 2 = 306.42, P <0.001. It was found that the negative binomial regression was a suitable model. Scarlet fever occurred with the monthly average pressure, monthly average relative humidity and monthly mean minimum temperature were statistically significant (P <0.05). The results of monthly incidence forecast in each month of 2006 showed that there was no significant difference between predicted and actual values (Wilcoxon signed rank sum test, Z = 0.24, P = 0.814), suggesting that the prediction effect was satisfactory. Conclusion By fitting the negative binomial regression model, it is found that the occurrence and prediction of scarlet fever, monthly average pressure, monthly average relative humidity and monthly mean minimum temperature are not negligible meteorological factors.