Prediction of high incidence of dengue in the Philippines.
Prediction of high incidence of dengue in the Philippines.
复制标题
DOI：
10.1371/journal.pntd.0002771
复制
发表时间：
2014-04
影响因子：
3.8
通讯作者：
Lewis SH
中科院分区：
文献类型：
--
作者：
Buczak AL;Baugher B;Babin SM;Ramac-Thomas LC;Guven E;Elbert Y;Koshute PT;Velasco JM;Roque VG Jr;Tayag EA;Yoon IK;Lewis SH
在爆发前几周准确预测登革热发病水平可能会降低与这种被忽视的疾病相关的发病率和死亡率。因此，建立了预测登革热高、低发病率的模型，以便在菲律宾提供及时的预警。模型输入的选择是基于表明可能影响登革热发病率的变量的研究。该方法首先使用模糊关联规则挖掘技术提取关联规则，从这些历史流行病学，环境和社会经济数据，以及气候数据表明未来的天气模式。选择标准用于为分类器选择这些规则的子集，从而生成预测模型。这些模型提前四周预测了菲律宾一个省登革热的高或低发病率。根据历史发生率数据确定高和低之间的阈值。模型准确度通过阳性预测值（PPV）、阴性预测值（NPV）、灵敏度和特异性来描述，这些数据是根据先前未用于开发模型的测试数据计算的。选择使用F0.5测量的模型，其使PPV比灵敏度更重要，得到这些结果：PPV = 0.780，NPV = 0.938，灵敏度= 0.547，特异性= 0.978。 使用F3测量值（灵敏度比PPV更重要），所选模型的PPV = 0.778，NPV = 0.948，灵敏度= 0.627，特异性= 0.974。 至于哪种模型具有更大的效用，取决于在特定情况下如何使用预测。这种方法建立了菲律宾未来登革热发病率的预测模型，并能够进行修改，以用于不同的情况;登革热以外的疾病;以及菲律宾以外的地区。菲律宾登革热预测模型预测高或低发病率的爆发前四周的高准确性，如PPV，NPV，灵敏度和特异性。描述了一种用于创建模型的基本自动化的方法，该模型使用过去和最近的数据来预测登革热发病率水平，提前几周预测特定时间段和可以是次国家的地理区域。输入数据包括历史和最近的登革热发病率、社会经济因素以及与天气、气候和环境有关的遥感变量。在气候变量中，有一些是已知的，可以指示未来的天气模式，这些模式可能是季节性的，也可能不是季节性的。最终的预测模型遵循以下原则：1）所使用的数据必须在做出预测时可用（避免使用最新数据的研究所造成的陷阱，在实际操作中，这些数据直到做出预测的日期之后才可用）; 2）模型在开发过程中没有使用的数据上进行测试（从而避免对预测准确性过于乐观的测量）。考虑了当地公共卫生部门对低数量假阳性和假阴性的偏好。这些模型似乎是强大的，即使应用到附近的地理区域，没有使用模型的发展。该方法可应用于其他媒介传播和环境影响的疾病。
Accurate prediction of dengue incidence levels weeks in advance of an outbreak may reduce the morbidity and mortality associated with this neglected disease. Therefore, models were developed to predict high and low dengue incidence in order to provide timely forewarnings in the Philippines. Model inputs were chosen based on studies indicating variables that may impact dengue incidence. The method first uses Fuzzy Association Rule Mining techniques to extract association rules from these historical epidemiological, environmental, and socio-economic data, as well as climate data indicating future weather patterns. Selection criteria were used to choose a subset of these rules for a classifier, thereby generating a Prediction Model. The models predicted high or low incidence of dengue in a Philippines province four weeks in advance. The threshold between high and low was determined relative to historical incidence data. Model accuracy is described by Positive Predictive Value (PPV), Negative Predictive Value (NPV), Sensitivity, and Specificity computed on test data not previously used to develop the model. Selecting a model using the F0.5 measure, which gives PPV more importance than Sensitivity, gave these results: PPV = 0.780, NPV = 0.938, Sensitivity = 0.547, Specificity = 0.978. Using the F3 measure, which gives Sensitivity more importance than PPV, the selected model had PPV = 0.778, NPV = 0.948, Sensitivity = 0.627, Specificity = 0.974. The decision as to which model has greater utility depends on how the predictions will be used in a particular situation. This method builds prediction models for future dengue incidence in the Philippines and is capable of being modified for use in different situations; for diseases other than dengue; and for regions beyond the Philippines. The Philippines dengue prediction models predicted high or low incidence of dengue four weeks in advance of an outbreak with high accuracy, as measured by PPV, NPV, Sensitivity, and Specificity. A largely automated methodology is described for creating models that use past and recent data to predict dengue incidence levels several weeks in advance for a specific time period and a geographic region that can be sub-national. The input data include historical and recent dengue incidence, socioeconomic factors, and remotely sensed variables related to weather, climate, and the environment. Among the climate variables are those known to indicate future weather patterns that may or may not be seasonal. The final prediction models adhere to these principles: 1) the data used must be available at the time the prediction is made (avoiding pitfalls made by studies that use recent data that, in actual practice, would not be available until after the date the prediction was made); and 2) the models are tested on data not used in their development (thereby avoiding overly optimistic measures of accuracy of the prediction). Local public health preferences for low numbers of false positives and negatives are taken into account. These models appear to be robust even when applied to nearby geographic regions that were not used in model development. The method may be applied to other vector borne and environmentally affected diseases.
登录
查看更多内容
DOI：
10.1038/nrg2579
发表时间：
2009-06
期刊：
Nature reviews. Genetics
影响因子：
--
作者：
Cordell HJ
通讯作者：
Cordell HJ
影响因子：
10.4
作者：
Hales, S;Weinstein, P;Woodward, A
通讯作者：
Woodward, A
DOI：
10.1088/1748-9326/4/1/014011
发表时间：
2009-03-04
期刊：
Environmental research letters : ERL [Web site]
影响因子：
--
作者：
Fuller DO;Troyo A;Beier JC
通讯作者：
Beier JC
影响因子：
2
作者：
Aguiar, Maira;Ballesteros, Sebastien;Stollenwerk, Nico
通讯作者：
Stollenwerk, Nico
DOI：
10.4269/ajtmh.1995.53.489
发表时间：
1995-11-01
影响因子：
3.3
作者：
FOCKS, DA;DANIELS, E;KEESLING, JE
通讯作者：
KEESLING, JE