Spatial and temporal clustering of an occupational poisoning: the example of green tobacco sickness.
McKnight-RH; Kryscio-RJ; Mays-JR; Rodgers-GC Jr.
Stat Med 1996 Apr; 15(7):747-757
Statistical analyses for spatial and temporal clustering of green tobacco sickness (GTS) cases reported to the Kentucky Regional Poison Center in 1991 were presented. Caused by skin absorption of nicotine (54115) through contact with tobacco plants, GTS was classified as an occupational health hazard of seasonal nature. Cases reported between the 10 week period of tobacco harvesting were classified by geographical area, and week and time of day of report. A total of 76 cases were reported during this time period. Case demographics showed that the average age of subjects was 22 years with 25% being minors and 78.9% male. Slightly over half the cases were treated at health care facilities. Clusters were identified for 31.6% of the cases. Analyses were performed on the 62 index cases. Ederer Meyers Mantel (EMM) and chi square statistics identified significant temporal clustering by harvest week, with most cases occurring in the third through eighth weeks of harvest. Significant clustering of reports between 6:00PM and 1:59AM was detected using Best and Rayner's V2. Spatial clustering was also seen across the state of Kentucky. EMM, but not chi square goodness of fit, statistics detected significant relationships between time and place of cases. The authors conclude that the clustering observed is consistent with the mechanism of tobacco harvesting and disease progression
NIOSH-Publication; NIOSH-Cooperative-Agreement; Farmers; Occupational-hazards; Tobacco-industry; Agricultural-workers; Epidemiology; Occupational-exposure
Statistics in Medicine
University of Kentucky, Department of Preventive Medicine, Lexington, KY