Health & Environmental Research Online (HERO)


Print Feedback Export to File
6030577 
Journal Article 
A potential risk of cancer in a central laboratory 
Blumstein, Z; Luxenburg, O; Modan, B; Novikov, I; Shemer, J 
1996 
1 
Statistics in Medicine
ISSN: 0277-6715
EISSN: 1097-0258 
15 
7-9 
759-763 
English 
9132903 
WOS:A1996UG35700009 
Clusters of cancer cases in a workplace usually raise academic curiousity with regard to potential aetiology, coupled with public outcry for investigation. The nature of such clusters makes it extremely difficult to delineate a chance occurrence from the true event. Following are some methodological problems we have encountered in the course of an investigation of a recent 'outbreak of cancer' in a central governmental laboratory in Jerusalem. The laboratory under investigation controls production and distribution of drugs and chemicals. It employs approximately 50 workers who are exposed to toxic substances. Among the list of materials used in the laboratory the potential carcinogens are asbestos, organic solvents, benzene, formaldehyde, hormones and radiomimetic drugs. In early 1994, a short newspaper item with bold headlines, read '11 cancer cases among 50 workers'. This led to considerable stress among the workers and their families. Consequently, union representatives began applying pressure on the Ministry of Health who agreed to temporarily reduce the workers' hours by one-third, without a parallel pay-cut, until a full assessment of the apparent cluster was undertaken. Two specific questions needed to be answered: (a) is the number of cancer cases indeed higher than expected, considering all potential confounders; (b) if there is a true excess, is it causal. 
Conference on Statistics and Computing in Disease Clustering 
Vancouver, Canada 
July 21-22, 1994