Investigating Suspected Cancer Clusters and Responding to Community Concerns: Guidelines from CDC and the Council of State and Territorial Epidemiologists
|Recommendations and Reports |
Volume 62, No. RR-8
September 27, 2013
Investigating Suspected Cancer Clusters and Responding to Community Concerns: Guidelines from CDC and the Council of State and Territorial Epidemiologists
Recommendations and ReportsSeptember 27, 2013 / 62(RR08);1-14
The material in this report originated in the National Center for Environmental Health, Robin Ikeda, MD, Acting Director, and the Division of Environmental Hazards and Health Effects, Judith R. Qualters, PhD, Director.
Corresponding preparer: Kanta Sircar, Division of Environmental Hazards and Health Effects, National Center for Environmental Health, CDC. Telephone: 770-488-3384; E-mail: firstname.lastname@example.org.
This report augments guidelines published in 1990 for investigating clusters of health events (CDC. Guidelines for investigating clusters of health events. MMWR 1990;39[No. RR-11]). The 1990 Guidelines considered any noninfectious disease cluster, injuries, birth defects, and previously unrecognized syndromes or illnesses. These new guidelines focus on cancer clusters. State and local health departments can use these guidelines to develop a systematic approach to responding to community concerns regarding cancer clusters. The guidelines are intended to apply to situations in which a health department responds to an inquiry about a suspected cancer cluster in a residential or community setting only. Occupational or medical treatment-related clusters are not included in this report. Since 1990, many improvements have occurred in data resources, investigative techniques, and analytic/statistical methods, and much has been learned from both large- and small-scale cancer cluster investigations. These improvements and lessons have informed these updated guidelines.
These guidelines utilize a four-step approach (initial response, assessment, major feasibility study, and etiologic investigation) as a tool for managing a reported cluster. Even if a cancer cluster is identified, there is no guarantee that a common cause or an environmental contaminant will be implicated. Identification of a common cause or an implicated contaminant might be an expected outcome for the concerned community. Therefore, during all parts of an inquiry, responders should be transparent, communicate clearly, and explain their decisions to the community.
IntroductionIn 1990, CDC published guidelines for investigating clusters of health events (the 1990 Guidelines) (1). The 1990 Guidelines did not focus on any specific disease and considered any noninfectious diseases, injuries, birth defects, and previously unrecognized syndromes or illnesses. Many state, local, and tribal health departments have used the 1990 Guidelines as a basis for developing and implementing protocols to investigate suspected cancer clusters, employing the four-step approach (initial response, assessment, major feasibility study, and etiologic investigation) identified in the 1990 Guidelines. Since the 1990 Guidelines were published, continued attention has been paid to suspected cancer clusters nationwide, leading CDC to publish additional details on the role of the guidelines in responding specifically to cancer clusters (2). Since 1990, many improvements have been made in the areas of data resources, investigative techniques, and analytic/statistical methods, and much has been learned from both large- and small-scale cancer cluster investigations.
This report augments the 1990 Guidelines by focusing specifically on cancer cluster investigations. The guidance provided in this report addresses additional subject areas that are deemed important by epidemiologists from state and local health departments (3). The additional subject areas include communications and resources for data and use of epidemiologic and spatial statistical methods. Useful websites, a resource not available in 1990, were added. The four-step process was retained, and more details were added.
Public health personnel in state and local health departments can use these guidelines to develop a systematic approach when responding to inquiries about suspected cancer clusters in residential or community settings. In addition, these guidelines might be helpful to a wider community of responders and epidemiologists who are concerned with such inquiries. These types of inquiries often are requested by community members or medical professionals concerned about what appears to be an unusually high number of diagnosed cases of cancer in a particular community, workplace, family, or school. Upon receiving an initial inquiry, health department personnel should respond rapidly to the caller's concerns, gather relevant information about the cancer cases, make a professional judgment on the likelihood that the reported situation could be an actual increase in cancer cases over those expected in a particular population, and determine whether further investigation is warranted. If appropriate, health department personnel then will need to provide resources for investigation of the suspected cluster, working with and involving members of the community as much as possible throughout the process.
MethodsIn March 2010, the Council of State and Territorial Epidemiologists (CSTE) and CDC convened a workgroup (the authors of this report) to revise the 1990 Guidelines. The group comprised public health professionals selected by the leadership of CSTE's Environmental Epidemiology Subcommittee and by CDC's National Center for Environmental Health's (NCEH) Division of Environmental Hazards and Health Effects (EHHE). CSTE and CDC selected workgroup members with experience in responding to cancer cluster inquiries from communities and managing of cancer cluster investigations. Representatives included epidemiologists from state health departments who were selected in order to have input from states that represent a range of approaches to and capacities for cancer cluster investigations. In addition, CDC workgroup members included representatives from CDC organizations typically called upon to assist in cancer cluster investigations: NCEH/EHHE, the Agency for Toxic Substances and Disease Registry (ATSDR), and the National Center for Chronic Disease Prevention and Health Promotion's Division of Cancer Prevention and Control. CDC risk communications and statistical specialists, as well as epidemiologists at academic institutions experienced in cancer cluster investigations, participated in the workgroup.
The intent of the workgroup was to ensure a practical approach to the assessment, analysis, and investigation of response to cancer cluster concerns. Through regularly scheduled conference calls and meetings from March 2010 to May 2011, the workgroup identified areas that warranted change from the 1990 Guidelines and sources of new information to incorporate in the revision of the guidelines. For these topics, the medical librarians at the CDC Public Health Library and Information Center conducted a comprehensive review of the published, peer-reviewed literature. To identify articles related to community cancer clusters, librarians conducted a structured literature search using multiple databases including PubMed (National Library of Medicine, National Institute of Health, Bethesda, Maryland, available at http://www.ncbi.nlm.nih.gov/PubMed), MEDLINE (available at http://www.nlm.nih.gov/bsd/pmresources.html), and CAB (available at http://cabdirect.org). English language peer-reviewed articles published between 1969 and 2010 were searched by using the following medical subject heading (MeSH) terms: "cluster analysis," "cancer cluster," "neoplasm," "environmental illness," and "not occupational diseases." Through this process, 166 articles were identified. In addition, members of the workgroup recommended 26 publications, including publications on communications and statistical analysis as well as nonscientific publications related to cancer clusters, and three unpublished cancer cluster investigation reports that were relevant to topics addressed in the guidelines. All articles and reports were reviewed by the workgroup members. Regarding topics on which no new published evidence was available, expert opinion was sought from workgroup members. In October 2010, an in-person meeting of the workgroup was held to begin writing these guidelines.
In addition to convening a technical workgroup, CSTE sent a survey to all state and territorial epidemiologists to assess the needs of public health professionals when responding to cancer cluster concerns in order to direct the focus and content of the guidelines (3). The survey included questions about the most common activities which states engage in when addressing a cancer cluster inquiry and what type of information would be useful. This survey identified areas (e.g., communications, resources for data, and epidemiologic methods) in which more details would be useful. After discussion, review, and incorporation of the findings from the survey, the workgroup decided to retain and update the four-step approach first described in the 1990 Guidelines. Updates included incorporating new technological advances (e.g., use of the Internet and websites) for information on relevant data resources, statistical tests, and mapping techniques as well as lessons learned from recent cancer cluster investigations. One important update is the emphasis on the importance of developing a robust working relationship with the community as soon as possible, including clear two-way communication and transparency in all aspects of the response process, while maintaining scientific rigor.
The revised guidelines address questions about the availability of data, limitations associated with understanding cancer clusters, and decision-making about the extent to which inquiries can be followed up. For specificity, the revised guidelines are limited in scope to include only those situations in which a health department responds to an inquiry about a suspected cancer cluster in a residential or community setting. These guidelines do not address workplace cancer clusters or those related to medical treatment (e.g., cancers associated with pharmaceuticals). Workplace or occupational clusters and medically related clusters each present unique sets of circumstances, have unique and clearly defined populations at risk, and generally call for specific investigative methods, agencies, and partnerships (4,5). Similarly, these guidelines do not discuss diseases other than cancer that persons might suspect have occurred in clusters in their communities. However, some of the principles of risk communication, data analysis, and community involvement discussed in this report might be applicable to noncancer cluster investigations as well. Finally, the revised guidelines do not address routine surveillance conducted by cancer registries and programs to assess trends.
This report is divided into two sections and three appendices:
- The first section explains cancer cluster definitions, characteristics and lessons learned from recent investigations;
- The second section outlines a systematic, four-step process for evaluating potential cancer clusters;
- Appendix A provides an overview of sources of data and other resources useful for cancer cluster investigations;
- Appendix B describes considerations for developing effective communication strategies; and
- Appendix C highlights some useful statistical and epidemiologic approaches for investigating suspected cancer clusters.
Cancer Cluster Definitions, Characteristics, and Recent Investigations
Definition of a ClusterCDC defines a cancer cluster as a greater than expected number of cancer cases that occurs within a group of people in a geographic area over a defined period of time (6). This definition can be broken down as follows:
- a greater than expected number: Whether the number of observed cases is greater than one typically would observe in a similar setting (e.g., in a cohort of a similar population size and within demographic characteristics) depends on a comparison with the incidence of cancer cases seen normally in the population at issue or in a similar community.
- of cancer cases: The cancer cases are all of the same type. In rare situations, multiple cancer types may be considered when a known exposure (e.g., radiation or a specific chemical) is linked to more than one cancer type or when more than one contaminant or exposure type has been identified.
- that occurs within a group of people: The population in which the cancer cases are occurring is defined by its demographic factors (e.g., race/ethnicity, age, and sex).
- in a geographic area: The geographic boundaries drawn for inclusion of cancer cases and for calculating the expected rate of cancer diagnoses from available data are defined carefully. It is possible to "create" or "obscure" a cluster inadvertently by selection of a specific area.
- over a period of time: The time period chosen for analysis will affect both the total cases observed and the calculation of the expected incidence of cancer in the population.
Characteristics of Cancer and ClustersThe National Cancer Institute (NCI) of the National Institutes of Health (NIH) defines cancer as a term for a group of diseases in which abnormal cells divide without control and can invade nearby tissues (7). As a group, cancers are very common. Cancers are the second leading cause of death in the United States, exceeded only by diseases of the heart and circulatory system (8). One of every four deaths in the United States is attributable to some form of cancer. In 2009, approximately 1.47 million persons in the United States received a cancer diagnosis, and approximately 568,000 persons died from cancer (9).
Because cancer is common, cases might appear to occur with alarming frequency within a community even when the number of cases is within the expected rate for the population. As the U.S. population ages, and as cancer survival rates continue to improve, in any given community, many residents will have had some type of cancer, thus adding to the perception of an excess of cancer cases in a community. Multiple factors affect the likelihood of developing cancer, including age, genetic factors, and such lifestyle behaviors as diet and smoking. Also, a statistically significant excess of cancer cases can occur within a given population without a discernible cause and might be a chance occurrence (10,11).
Three considerations are important for suspected cancer cluster investigations. First, types of cancers vary in etiologies, predisposing factors, target organs, and rates of occurrence. Second, cancers often are caused by a combination of factors that interact in ways that are not fully understood. Finally, for the majority of cancers, the long latency period (i.e., the time between exposure to a causal agent and the first appearance of symptoms and signs) complicates any attempt to associate cancers occurring at a given time in a community with local environmental contamination. Often decades intervene between the exposures that initiate and promote the cancer process and the development of clinically detectable disease (12).
Communicating effectively about the frequency and nature of cancer in explaining suspected cancer clusters can be difficult for public health agencies, and many of the scientific concepts involved (e.g., random fluctuation, statistical significance and latency period) might not be easy to explain to the community (13). Any number of community members, friends, or relatives with cancer is alarming and is too many from a personal perspective (11). When persons are affected personally by a case of cancer, they naturally seek an explanation of the cause of the cancer (13).