摘要
目的采用控制图法预警技术确定上海市松江区流行性感冒最佳预警界值,提高流行性感冒疫情监测预警能力。方法利用2006—2009年松江区流行性感冒报告数据(ILI就诊百分比)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、阳性预测值和阴性预测值及绘制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、阳性预测值和阴性预测值等指标后,将P95作为松江区流行性感冒的预警界值。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。
OBJECTIVES To select the premium alert threshold for influenza by using the control graph alert technique and to upgrade the early dection capability of influenza. METHODS Alert data was based on the report of influenza information（ILI doctor seeking rate） from 2006 to 2009, control graph was used to detect outbreak or epidemic, after calculating sensitivity, specifillty, positive predictive value and negative predictive value and ROC, the premium alert threshoud for influenza in Songjiang Distirct was chosed. RESULTS After the comprehensive consideration of sensitivity, speeifility, positive predictive value and negative predictive value and ROC, P95 was ehosed as the premium alert threshold for influenza in Songjiang Distiret. CONCLUSION Premium alert threshold selection should consider the local disease report and the characteristics of infectious diseases to upgrade the early detection capability.
出处
《中国初级卫生保健》
2010年第7期77-78,共2页
Chinese Primary Health Care
关键词
流行性感冒
控制图法
预警
influenza
control graph
early detection
