A cancer cluster is defined by the Centers for Disease Control and Prevention (CDC) as a "Centers for Disease Control and Prevention, “About Cancer Clusters,” accessed June 21, 2021, https://www.cdc.gov/nceh/clusters/about.htm.greater-than-expected number of cancer cases that occurs within a group of people in a geographic area over a period of time.” Because there are more than 100 different types of cancer with a wide range of causes, it is often difficult to definitively establish the existence, let alone the cause, of a cluster. The investigation process is often prolonged and scientifically, politically, and emotionally fraught for the individuals and communities that form a suspected cancer cluster. National Cancer Institute, "Cancer Clusters," August 29, 2018 [last updated], https://www.cancer.gov/about-cancer/causes-prevention/risk/substances/cancer-clusters-fact-sheet.In 576 cancer cluster investigations in the United States over 20 years, only 72 had a statistically significant increase in cancer rates, and only one had a clear common cause. Even in a well-documented cluster where the source of exposure appears to be self-evident, the presence of other common risk factors in the population,“Research on Adverse Health Effects Related to Rocky Flats,” cdphe.colorado.gov, accessed January 13, 2023, https://cdphe.colorado.gov/rocky-flats-historical-public-exposure-studies.
such as smoking, can lead investigators to declare the cause inconclusive. Because many potentially hazardous exposures occurred in the context of national security—for instance, to workers at the Rocky Flats plutonium production facility—it can be difficult to establish even basic information about exposures endured by civilian workers. For struggling caregivers and grieving communities, an indeterminate investigation not only fails to explain their suffering but also forecloses the possibility of establishing accountability and receiving compensation for debilitating illnesses.
Every diagnosed case of cancer (other than non-melanoma skin cancers) is reported to a central state registry to track cancer incidence, assess prevention results, and evaluate treatment outcomes on a population-scale. In Colorado, the registry is part of the Colorado Department of Public Health and the Environment. However, under CDC guidelines, state public health departments generally do not proactively identify and investigate cancer clusters. Rather, an investigation begins when a concerned or affected citizen reports a suspected cluster—often after compiling significant lay research, on their own time, and on top of the burden of attending to their own or their family’s health needs. An investigator then contacts the reporter and draws on cancer registry data to establish background information such as number of cancer cases, demographics, type of cancer, and potential causes. Analysis of these data determines if a statistically significant difference exists between the number of cancers in the cluster population compared to census data. If there is not a significant difference, then investigation stops.
If the investigation proceeds, the third stage determines the feasibility of an epidemiological study. Investigators confirm cancer cases, undertake an extensive literature review on the given cancer(s), evaluate environmental factors for proven carcinogens, and begin study design. Finally, step four involves a full epidemiological investigation to ascertain the source of the cluster by a process known as causal inference. Douglas L. Weed, “Environmental Epidemiology: Basics and Proof of Cause-effect,” Toxicology 181-182 (2002): 399-403, https://doi.org/10.1016/S0300-483X(02)00476-6.To establish causality, investigators look for five criteria: the relationship between the hypothesized cause and disease, the time of exposure, the amount of exposure, the likelihood that the exposure caused the illness, and the consistency among different populations.
Epidemiological investigations have confirmed relatively few cancer clusters in the United States, linked only three of these to a single exposure, and established causality only once. While cancer clusters do occur by chance, methodological shortcomings in epidemiological research may lead to the infrequent determination of cause. A lack of clear and consistent guidelines governing the literature review process may lead individual researchers to arrive at different conclusions because they consulted different studies. Furthermore, small population numbers inside a given cancer cluster reduce the reliability of the statistical analysis used to establish causation. A. B. Haidich, “Meta-analysis in Medical Research,” Hippokratia 14, no. 1 (2010): 29-37, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3049418/.The meta-analysis methods used to overcome the problem of sample size depends on consulting the same scientific literature for which there are no consistent guidelines. Finally, the exceeding rareness of certain cancers presents a problem even for meta-analysis. In a charged environment where a causal determination has policy implications or may expose a corporation or the government to liability claims, methodological limitations can be hard to differentiate from political motivations and have far-reaching, real-world effects that are born most severely by those who are already sick.
Finally, the nature of cancer itself makes definitively establishing causation exceedingly difficult. If an investigator begins to study one suspected environmental cause of the cancer, the inability to isolate a given exposure from other risk factors is a serious obstacle. Epidemiological studies cannot control variables since they involve real people with complex lives. Due to this lack of control, patients may exhibit other risk factors that confound the relationship between a hypothetical cause and their cancers. Moreover, individual people react differently to carcinogens. One person could be exposed to a carcinogen for decades and not develop cancer, yet another person could be exposed for a much shorter period and still develop cancer. This translates to a low level of consistency and weak relationship between dose (or exposure) and response (cancer)—both criteria for causal inference. There are disparities in cancer incidence and outcomes by race, ethnicity, and gender from the interplay of socioeconomic, socioenvironmental, and cultural-behavioral factors. Linear epidemiological models make it difficult if not impossible to incorporate multiple models of exposure, and evidence of pre-existing vulnerabilities frequently weighs heavier than exposure causality for explaining negative health outcomes. Given this inherent complexity and high degree of uncertainty, it is no surprise that epidemiological studies into cancer clusters have uncovered so few conclusive causes.
Scientific advances—particularly around meta-analysis and cancer research—may eventually improve existing processes of causal inference. However, many scholars of science, law, and policy have called for a transformation in the way scientific evidence is used in cases of environmental health where inevitably David Kriebel, "How Much Evidence is Enough? Conventions of Causal Inference," Law and Contemporary Problems 72, no. 1 (2009): 122, https://www.jstor.org/stable/40647168."facts are uncertain, social values are in dispute, the stakes are high, and decisions are urgent," as epidemiologist David Kriebel has written. Public and environmental health research will never be conducted like empirical science, and investigators have an ethical obligation to prevent or mitigate harm even in the absence of absolute certainty. New methods of analyzing scientific evidence that bear in mind degrees of uncertainty, such as sensitivity analysis and Bayesian statistics, may facilitate the incorporation of research into policy. However, social values fundamentally shape the work at the interface of science and public policy, and American culture has long favored expediting corporate and institutional action while shifting the burden of proof onto those harmed. The adoption of legal frameworks such as the Precautionary Principle (formalized in German law as Vorsorgeprinzip) or the Rights of Nature (enshrined in the constitution of Ecuador) may shape both policy and scientific practice around environmental health in ways that will help us better understand existing cancer clusters and perhaps prevent them in the future.
Abrams, Beth, Henry Anderson, Carina Blackmore, et al. “Investigating Suspected Cancer Clusters and Responding to Community Concerns: Guidelines from CDC and the Council of State and Territorial Epidemiologists.” CDC, September 27, 2013. Accessed April 4, 2021.
Community Environmental Legal Defense Fund. “Champion the Rights of Nature.” Accessed February 28, 2021.
Goodman, Michael, Joshua S. Naiman, Dina Goodman, and Judy S. LaKind. “Cancer Clusters in the USA: What Do the Last Twenty Years of State and Federal Investigations Tell Us?” Critical Reviews in Toxicology 42, no. 6 (2012): 474-490.
Haidich, A.B. “Meta-analysis in Medical Research.” Hippokratia 14, no. 1 (2010): 29-37. Accessed December 13, 2020.
Kriebel, David. "How Much Evidence is Enough? Conventions of Causal Inference.” Law and Contemporary Problems 72, no. 1 (2009): 121-136.
Krupar, Shiloh. Hot Spotter’s Report: Military Fables of Toxic Waste. Minneapolis, MN: University of Minnesota Press, 2013.
National Cancer Institute. "Cancer Clusters." Last updated August 29, 2018. Accessed June 22, 2021.
“Research on Adverse Health Effects Related to Rocky Flats.” cdphe.colorado.gov. Accessed January 13, 2023.
Umwelt Bundesamt. “The Precautionary Principle," January 21, 2021. Accessed February 28, 2021.
Weed, Douglas L. “Environmental Epidemiology: Basics and Proof of Cause-effect.” Toxicology 181-182 (2002): 399-403. Accessed December 2, 2020.