McKnight R H, Kryscio R J, Mays J R, Rodgers G C
Department of Preventive Medicine, College of Medicine, University of Kentucky, Lexington 40536-0084, USA.
Stat Med. 1996;15(7-9):747-57. doi: 10.1002/(sici)1097-0258(19960415)15:7/9<747::aid-sim246>3.0.co;2-p.
Nicotine poisonings in tobacco farm workers (Green Tobacco Sickness, GTS) reported to the Kentucky Regional Poison Center during 1991 were assessed for spatial, temporal and space-time clustering. Among 62 index cases, 10 had worked with others who also reported GTS that day. Analysis of index cases, using Best and Rayner's V(2), showed significant cyclic clustering by hour of day the report was received (p = 9 x 10(-9)); chi-square revealed significant spatial clustering by agricultural district (p = 0.0002). Space-time clustering was detected using the E-M-M statistic (p = 0.008); however, the space-by-time interaction of Rao's chi-square goodness-of-fit statistic did not detect space-time clustering (p = 0.32).
对1991年向肯塔基地区中毒控制中心报告的烟草农场工人尼古丁中毒事件（绿色烟草病，GTS）进行了空间、时间和时空聚集性评估。在62例索引病例中，有10人与当天也报告患绿色烟草病的其他人一起工作。使用Best和Rayner的V(2)对索引病例进行分析，结果显示按报告接收时间的小时数存在显著的周期性聚集（p = 9×10⁻⁹）；卡方检验显示按农业区存在显著的空间聚集（p = 0.0002）。使用E-M-M统计量检测到时空聚集（p = 0.008）；然而，Rao卡方拟合优度统计量的空间×时间交互作用未检测到时空聚集（p = 0.32）。
