Signature-forecasting and early outbreak detection system
Signature-forecasting and early outbreak detection system
复制标题
DOI：
10.1002/env.734
复制
发表时间：
2005-11-01
期刊：
影响因子：
1.7
通讯作者：
MacNeill, IB
中科院分区：
文献类型：
--
作者：
Naumova, EN;MacNeill, IB
通过公共卫生监测系统进行的日常疾病监测提供了有关人口风险的宝贵信息。对于现代监测来说，早期发现疾病发病率快速变化的有效统计工具是必不可少的。显然，需要统计工具来及早发现不基于历史信息的暴发。讨论了一种监测感染病例的系统，以期及早发现疫情并预测检测到的疫情的程度。我们提出了一套不依赖于大量历史记录的早期疫情检测的自适应算法。我们还将传染病流行病学知识纳入预测。为了证明这个系统，我们使用了1993年在密尔沃基发生的最大规模的水媒隐孢子虫病暴发的数据。历史数据使用黄土类型的平滑器进行平滑。在接收到新的数据时，更新平滑，并对平滑曲线的前两个导数进行估计，这些估计用于近期预测。最近的数据和近期预测被用来计算颜色编码的警告指数，该指数量化了担忧的程度。计算预警指数的算法被设计成平衡第一类错误(对流行病的错误预测)和第二类错误(未能正确预测流行病)。如果警告指数表示发生疫情的可能性足够高，则对疫情的可能规模进行预测。这一较长期的预测是通过对现有数据进行“签名”曲线拟合而做出的。预测的有效性取决于签名曲线在多大程度上捕捉到了正在考虑的感染暴发的形状。版权所有(C)2005 John Wiley&Sons，Ltd.
Daily disease monitoring via a public health surveillance system provides valuable information on population risks. Efficient statistical tools for early detection of rapid changes in the disease incidence are a must for modem surveillance. The need for statistical tools for early detection of outbreaks that are not based on historical information is apparent. A system is discussed for monitoring cases of infections with a view to early detection of outbreaks and to forecasting the extent of detected outbreaks. We propose a set of adaptive algorithms for early outbreak detection that does not rely on extensive historical recording. We also include knowledge of infection disease epidemiology into forecasts. To demonstrate this system we use data from the largest water-borne outbreak of cryptosporidiosis, which occurred in Milwaukee in 1993. Historical data are smoothed using a loess-type smoother. Upon receipt of a new datum, the smoothing is updated and estimates are made of the first two derivatives of the smooth curve, and these are used for near-term forecasting. Recent data and the near-term forecasts are used to compute a color-coded warning index, which quantify the level of concern. The algorithms for computing the warning index have been designed to balance Type I errors (false prediction of an epidemic) and Type II errors (failure to correctly predict an epidemic). If the warning index signals a sufficiently high probability of an epidemic, then a forecast of the possible size of the outbreak is made. This longer term forecast is made by fitting a 'signature' curve to the available data. The effectiveness of the forecast depends upon the extent to which the signature curve captures the shape of outbreaks of the infection under consideration. Copyright (c) 2005 John Wiley & Sons, Ltd.