April 9, 2019
For decades, epidemiologists have studied the incidence and distribution of cancer with the aim of preventing the disease and identifying its causes. A cancer cluster is the occurrence of a greater than expected number of cancer cases over a period of time in a defined place or population subgroup. Notable historical examples of cancer clusters include occupational exposure to vinyl chloride and angiosarcoma of the liver, and having HIV/AIDS and a rare cancer known as Kaposi sarcoma.
In recent years, the process of cancer cluster investigation has been facilitated by the expansion of the National Program of Cancer Registries and the increased use of electronic medical records for cancer surveillance. At the same time, ongoing environmental policy discussions and improvements in digital information sharing have heightened public concern about potential health risks that may be associated with exposure to chemicals from consumer products, industrial processes, pesticides, and pollution. Employers, school administrators, community leaders, and the general public should understand how epidemiologists evaluate potential clusters and why specific environmental causes usually cannot be identified.
Unlike the clinical practice of medicine, which focuses on individual patients, epidemiologists take a population-level approach to investigating potential clusters of diseases or other health-related events. To identify a cluster, the primary question we seek to answer is whether there is an increase in the occurrence of cancer beyond the norm.
" ... even if there is a statistically significant excess of cases in a certain place and time, it might be due to chance rather than a single identifiable cause."
For example, let's imagine that an employer has five employees diagnosed with a particular type of cancer in a single year, and asks an epidemiologist to investigate whether the cases represent a potential cancer cluster. Our first step is to confirm the number of people affected, their demographics, the specific type(s) of cancer that they have, and the dates of diagnosis. Once that information is collected, we use statistics to compare the number of identified cancer cases of a given type with the number expected in that population. To make this comparison we use cancer registry data, accounting for age, gender, race/ethnicity, calendar period, geography, and possibly other factors. If the number of observed cancer cases is statistically greater than the number expected, we can conduct further investigation to uncover relevant occupational, medical, lifestyle, and family exposure histories, and to attempt to identify shared risk factors among the cases. Whether or not the number of observed cases is beyond expectation, we provide the employer with information about cancer risk factors, the process of cancer cluster investigation, our statistical findings, and their interpretation.
As is well documented in the epidemiologic literature, statistical testing shows that the number of observed cases is not statistically higher than expected for the majority of suspected cancer clusters. Furthermore, even if there is a statistically significant excess of cases in a certain place and time, it might be due to chance rather than a single identifiable cause. That is, diseases often do not occur in an evenly distributed manner, but instead may randomly occur in apparent clusters. Therefore, the process of cancer cluster investigation is often a task of educating others about how cancer occurs, the known causes of a given cancer type, and the role of chance.
Our team at Exponent has investigated many potential cancer clusters. Recent investigations have involved potential clusters at schools, universities, and places of employment.
In each case, our team of epidemiologists partners with the client to obtain demographic and disease-specific information about the affected individuals, and we can perform thorough statistical analyses as well as literature reviews to summarize known risk factors for the types of cancer at issue. In some instances, the number of observed cancer cases may be statistically significantly greater than expected. In these cases, additional investigation, conducted in partnership with Exponent experts in toxicology, occupational medicine, and environmental health, may include research such as collecting extensive survey information from cases, evaluating environmental site assessment reports, and reviewing lists of agricultural chemicals used nearby. These types of analyses are needed to determine if a cluster is simply due to random chance or if there could be an identifiable underlying cause. Exponent scientists can perform cancer cluster analyses with speed and multidisciplinary expertise that public entities may be challenged to replicate.
How Exponent Can Help
Exponent's team of epidemiologists, toxicologists, physicians, environmental health scientists, exposure scientists, and statisticians can conduct comprehensive investigations of potential disease clusters and provide timely communication of our findings to clients and their constituents.
Authors
