基于前瞻性时空重排扫描的蚌埠市新冠肺炎疫情预警模型

为研究地级市市域使用前瞻性时空重排扫描统计量分析模型对探测新冠肺炎疫情潜在的暴发区进行预警的可行性,利用前瞻性时空扫描模型分析软件SaTScan 9.6及地理信息可视化表达软件ArcGIS 10.8对安徽省蚌埠市2020年1月23日至2月23日新冠肺炎疫情进行逐日模拟监测预警;通过时空重排扫描分析,探测出4个集簇,其中非随机性显著的集簇2个,包含怀远县全域20个乡镇、固镇县3个乡镇、蚌埠市区17个街道及开发区,其中第1集簇为低聚集区,涵盖1个街道,复现间隔15天;第2集簇为高聚集区,涵盖34个乡镇街道及开发区,复现间隔11天;蚌埠市禹会区、淮上区、蚌山区若干街道及怀远县全域为下一阶段新冠疫情空间聚集强预警的重点区域,前瞻性时空重排扫描分析可应用于地级市疫情防控应急响应检测预警。

Analysis of COVID-19 Epidemic in Bengbu Areas using Prospective Space-time Permutation Scan Statistics

To study COVID-19 epidemic potential in urban areas using prospective space-time permutation scan statistics.The prospective space-time scanning model analysis software SaTScan 9.6 and geographic information software ArcGIS 10.8 were used to carry out the daily simulation monitoring and early warning for the COVID-19 epidemic in Bengbu from January 23rd,2020 to February 23rd was simulated and monitored daily.Four cluster areas were detected by the model,including two clusters with high significant non randomness,including 20 towns in Huaiyuan County,3 towns in Guzhen county and 17 streets or development zones in Bengbu.The first cluster area was the low-rate cluster area,covering one street,with a recurrence interval of 15 days;the second cluster area was high-rate cluster area,covering 34 streets,The recurrence interval was 11 days.Several streets of Yuhui District,Huaishang District,Bangshan District and the whole Huaiyuan County of Bengbu are key areas for strong spatial clustering early warning of COVID-19 in the next stage.Prospective space-time permutation model can be applied to detection and early warning of the epidemic.