Cancer cluster detection in British Columbia school districts, 1983-1989

• Main Author: Rosychuk, Rhonda Jean
• Language: English
• Published: 2009
• Online Access: http://hdl.handle.net/2429/5008

A disease cluster is an aggregation of occurrences of a disease. The observa tion of a perceived excess number of similar illnesses is termed disease clustering. Statistical tests for disease clustering investigate if the observed pattern of cases in at least one geographical area could possibly have happened by chance alone. This pattern may be spatial, temporal, or both. Investigating possible cancer clusters in British Columbia for the period of 1983—1989 inclusive is the objective of this thesis. Whether or not cancer clustering appears near pulp and paper mills within the province is of specific interest. The geographical units upon which our investigation will be based are the B.C. school districts. The variation in size and population demographics among districts requires a cluster detection method which is considerate of the underlying population distribution within the study region. School district population size diversity requires a modification to the Besag and Newell (1991) method. This modification is implemented with B.C. school district data and several possible clusters are detected for various types of cancer.