目的探索甘肃省天水市建立可即时探测的传染病监测预警系统,将预警空间拓展至乡镇(街道)层级,增强预警的准确性和及时性,指导暴发疫情和处置突发公共卫生事件。方法用Excel 2010函数、VBA(visual basic application)编程和Active X控件构建天水市传染病监测预警信息系统;采用在市、县(区)、乡镇(街道)3个层级,逐步缩小空间范围的方法,最终定位病例聚集的乡镇(街道),获取突发公共卫生事件预警信息;时空聚集性预警采用控制图法,建立移动百分位数法模型,参照不同病种预警阈值,得到预警信息。结果 2014年1月─2016年10月,天水市共报告聚集性疫情14起,其中突发公共卫生事件4起,传染病监测预警信息系统探测到10起、灵敏度为71.43%,获取预警信息287条、阳性预测值3.48%;预警信息以流行性感冒、手足口病和水痘居多,占全部信息的89.2%,报告病例数与预警信息数之间呈正相关(r=0.771,P<0.05)。结论该系统能早期即时探测传染病的时空聚集性和突发公共卫生事件的发生,灵敏度和阳性预测值较高,但系统需要手工操作和记录较多,今后需优化和改进。 Objective To explore the establishment of a real-time monitoring and early warning system for infectious diseases in Tianshui City, Gansu Province. The early warning space was expanded to the township (street) level to enhance the accuracy and timeliness of early warning and guide the outbreak and public health emergencies. Methods The information system of monitoring and early warning of infectious diseases in Tianshui was constructed by Excel 2010 function, VBA (visual basic application) programming and Active X control. The method of gradually reducing the spatial range was adopted in three levels of cities, counties (districts) and towns (streets) , Finally locate the villages and towns (streets) where the cases were collected, and obtain the early warning information of public health emergencies. The spatial and temporal aggregation early warning adopts the control chart method to establish the moving percentile method model and get the early warning information with reference to the warning thresholds of different disease types. Results From January 2014 to October 2016, a total of 14 epidemic cases were reported in Tianshui City, including 4 public health emergencies, 10 infectious disease surveillance and early warning information systems, 71.43% sensitivity and 287 early warning information The positive predictive value was 3.48%. The information of early warning was mostly influenza, hand, foot and mouth disease and chickenpox, accounting for 89.2% of the total information. There was a positive correlation between the number of reported cases and the number of early warning information (r = 0.771, P <0.05). Conclusion The system can detect spatiotemporal aggregation and emergent public health emergencies of infectious diseases in real time with high sensitivity and positive predictive value. However, the system needs more manual operation and more records, and needs to be optimized and improved in the future.