控制图法预警技术在宁波市两种呼吸道传染病中的应用

Early detection for two respiratory infectious diseases in Ningbo City

目的:采用控制图法预警技术确定宁波市流行性感冒和风疹的最佳预警界值,提高呼吸道传染病疫情监测预警能力。方法:利用2005-2008年宁波市流行性感冒和风疹发病数据建立预警数据库,采用控制图法建立预警模型,通过计算比较灵敏度、特异度和阳性预测值及绘制ROC曲线,优选出最佳预警界值。结果:综合平衡预警的灵敏度、特异度和阳性预测值等指标后,将P 90作为宁波市流行性感冒和风疹的预警界值。结论:根据传染病发病特点,结合当地疫情信息优选出适合于当地的传染病预警界值,以提高传染病疫情监测的预警能力。

Objective To determine the alert thresholds for influenza and rubella in Ningbo,so as to improve the efficacy of the early warning system for respiratory infectious disease. Methods Weekly influenza and rubella prevalence data for 2005--2009 in Ningbo were abstracted from the China Information System for Disease Control and Prevention. Control chart was used to detect outbreak of epidemics ,which were defined by consulting specialists. After calculating sensitivity,specificity,positive predictive value and drawing receiver operator characteristic curve （ROC） , the optimal model and thresholds were finally determined. Results By taking into account of sensitivity,specificity, positive predictive value and ROC, 90 percentile（P90） was chosen as the optimal alert threshold for influenza and rubella in Ningbo. Conclusion The characteristics of infectious diseases should be considered together with disease prevalence data to determine the optimal alert threshold for early warning system.