Li Xiaoning, Luo Lei, Xiao Xincai, Jing Qinlong, Wei Yuehong, Li Yilan, Cao Qing, Yang Zhicong, Xu Ya
Department of Epidemiology and Health Statistics, School of Public Health, Guangdong Pharmaceutical University, Guangzhou 510310, China.
Guangzhou Center for Disease Control and Prevention.
Zhonghua Liu Xing Bing Xue Za Zhi. 2014 Jul;35(7):821-4.
To understand the threshold of Breteau Index (BI) on Dengue fever outbreak in Guangzhou.
Dengue cases from Guangzhou during 2006 to 2012 in the National Infectious Disease Report and Management System were collected and described. Receiver-operating characteristic (ROC) curve was used to judge the threshold of BI on the outbreaks of Dengue fever.
A total of 1 038 local cases were reported from 2006 to 2012 in Guangzhou city, with a total of 71 outbreaks and 259 sporadic episodes. Data from the ROC curve analysis showed that the highest Youden index that BI predicting Dengue fever outbreaks or sporadic were 6.4 and 9.5, respectively. When using BI = 6.4 in predicting the sporadic case of Dengue, sensitivity and specificity were 67.8% , 79.1% , respectively. When using BI = 9.5 in predicting the outbreaks of Dengue, sensitivity and specificity were 81.7%, 90.9%, respectively.
Both BI = 5.0 and BI = 9.5 showed effects on predicting the nature of sporadic or outbreak on Dengue, suggesting that the threshold need to be monitored, according to the purpose of control and availability of manpower, in order to get better sensitivity and specificity.
了解广州登革热疫情中布雷图指数（BI）的阈值。
收集2006年至2012年广州在国家传染病报告管理系统中的登革热病例并进行描述。采用受试者工作特征（ROC）曲线判断登革热疫情中BI的阈值。
2006年至2012年广州市共报告本地病例1038例，共71起疫情和259例散发病例。ROC曲线分析数据显示，BI预测登革热疫情或散发病例的最高约登指数分别为6.4和9.5。以BI = 6.4预测登革热散发病例时，灵敏度和特异度分别为67.8%、79.1%。以BI = 9.5预测登革热疫情时，灵敏度和特异度分别为81.7%、90.9%。
BI = 5.0和BI = 9.5对预测登革热散发病例或疫情性质均有作用，提示需根据防控目的及人力情况监测阈值，以获得更好的灵敏度和特异度。
