(1) Greater than expected number of case. The observed or actual number of cases is greater than one would typically observe (e.g., the mean number of cases) in a similar setting. Similar setting refers to the parameters of the group, such as among an age group, a gender group, or a racial or ethnic group.
(2) Greater than expected cases involves identification of differences for the same type of cancer across geographically situated groups.
(3) Period of time. The parameters of the period should be similar for the suspected cluster and the comparison population(s).
General cancer data is available from the National Institute of Cancer for the US and may be generated for different cancers and locations (Cancer Data US).
A number of environmental factors may cause cancer. As the World Life Expectancy website notes, if cancers were caused largely by individual level life choices, when examining cancer maps, one would except that the distribution of cancer would appear random. As an example of the potentially non-random distribution of cancer and perhaps as an indicator that there are rather large, significant cancer cluster locations within the US, World Life Expectancy has published a map of cancer rates by county using data from the Center for Disease Control and Prevention (Cancer Map US). A cursory visual inspection of that map indicates that there are locations with both high, moderate and low rates of cancer, and at issue is whether this distribution approximates a pattern that could indicate large geographic areas that might be considered to approximate cancer clusters.
Many chemicals are known or suspected to cause cancer (see entry in this dictionary on Cancer Causing Chemicals). Pesticide exposure has been widely linked to cancer (see this website for studies linking Pesticides and Cancer), and thus one might suspect that the distribution of heavy pesticides use or manufacturer may be one of the factors that produces cancer clusters geographically.
There is little agreement in the scientific literature about the existence of cancer clusters. This may be the result of how cancer clusters have been defined, and the level of analysis at which such assessments ought to take place. Using county based assessments for thyroid cancer in Florida for persons aged 15-39, Amin and Burns (2014) discovered what they believe is evidence of large area cancer clusters (for additional discussion of county level analysis of cancer clusters see Bender et al., 2012; Amin et al., 2010; for a national level study of breast cancer clusters see, Tian, Wilson and Zhan, 2010; Chien, Yu and Schootman, 2013; on county level evidence of cancer clusters in Kentucky see, Christian et al., 2011; on prostate cancer clusters see, Altekruse et al., 2010; on possible cancer clusters in Chinese villages see, Liu, 2011; on pediatric cancer clusters in Florida see, Amin, Hendryx, Shull and Bohnert, 2014).
Altekruse, Sean F., Lan Huang, James E. Cucinelli, Timothy S. McNeel, Kristen M. Wells, and M. Norman Oliver. 2010. “Spatial patterns of localized-stage prostate cancer incidence among white and black men in the southeastern United States, 1999-2001.” Cancer Epidemiology Biomarkers & Prevention 19, 6: 1460-1467.
Amin, Raid, and James J. Burns. 2014. “Clusters of Adolescent and Young Adult Thyroid Cancer in Florida Counties.” BioMed Research International Thyroid Cancer Cluster Study.
Amin, Raid, Alexander Bohnert, Laurens Holmes, Ayyappan Rajasekaran, and Chatchawin Assanasen. “Epidemiologic mapping of Florida childhood cancer clusters.” Pediatric blood & cancer 54, no. 4 (2010): 511-518.
Amin, Raid W., Michael Hendryx, Matthew Shull, and Alexander Bohnert. (2014). “A cluster analysis of pediatric cancer incidence rates in Florida: 2000-2010.” Statistics and Public Policy (accepted).
Bender, Alan P., Allan N. Williams, John Soler, and Margee Brown. “A nonparametric approach for determining significance of county cancer rates compared to the overall state rate: illustrated with Minnesota data.” Cancer Causes & Control 23, no. 6 (2012): 791-805.
Chien, Lung-Chang, Hwa-Lung Yu, and Mario Schootman. 2013. “Efficient mapping and geographic disparities in breast cancer mortality at the county-level by race and age in the US.” Spatial and spatio-temporal epidemiology 5 :27-37.
Christian, W. Jay, Bin Huang, John Rinehart, and Claudia Hopenhayn. 2011. “Exploring geographic variation in lung cancer incidence in Kentucky using a spatial scan statistic: elevated risk in the Appalachian coal-mining region.” Public Health Reports 126, 6: 789-796.
Liu, Lee. 2010. “Made in China: cancer villages.” Environment: Science and Policy for Sustainable Development 52, 2: 8-21.
Tian, Nancy, J. Gaines Wilson, and F. Benjamin Zhan. 2010. “Female breast cancer mortality clusters within racial groups in the United States.” Health & place 16, 2: 209-218.