Skip to main content
         

Vol. 13 Issue 3, Summer 2008

Approach for Responding to Cancer and Environmental Concerns
Interview with Elizabeth Lewis-Michl, Ph.D.
The Ribbon 

Community Exposure Research Section
Bureau of Environmental and Occupational Epidemiology
Center for Environmental Health, NYS Department of Health

The Ribbon: We at BCERF receive many questions regarding “cancer clusters.” The term seems to carry various meanings, among the public, the media, epidemiologists, and others. At times, it is used to refer to the perception of an excess of cancer in a certain time and place, but the situation may not have been formally evaluated. Epidemiologists reserve the term for when a group of cancer cases has been statistically established to be in excess of the expected rate. How do you define cancer cluster?

Dr. Elizabeth Lewis-Michl: Detecting and responding to elevations of disease rates above expected rates, in space, time, or among specific groups of people, is a central focus of epidemiology, but the term “cluster” is rarely seen in an epidemiology textbook. In my experience, the term “cluster” is often used when there is also an assumption of a link between cause and effect. So I avoid using the term.

When we are asked to look into perceived excesses of cancer in space or time, we are quite limited in the conclusions we can draw using available data and traditional statistical tests. My program responds to concerns about cancer related to potential environmental exposures, and the lack of information about past exposures in most situations creates even greater challenges for responding to people’s concerns.

We attempt to avoid the confusion associated with the term “cluster”. Rather, in situations where we see elevated cancer incidence, we describe our methods, the data used, the statistical tests, and conclusions as precisely as possible, to help put the findings in the context of the limitations of the data and methods. Epidemiological studies only very rarely can link cause and effect, and the types of data reviews we often conduct, assessing whether cancer elevations are evident among the people diagnosed while living in a specific place, and over a specific time period, are unable to link cause and effect.

The Ribbon: What are the criteria that the Center for Environmental Health (CEH) uses to determine whether to investigate a potential excess of cancer in a particular place?

Dr. Lewis-Michl: We developed a flow chart to summarize our step-wise approach, which includes consideration primarily of health outcome and exposure characteristics (second box, then third box), but also of data quality and study feasibility issues (third box). When a concern is first brought to us in a letter, phone call, e-mail, or at a public event, the first step is critical and may include several conversations with an individual or a series of meetings with the community. We have to make sure we understand the concern, and we often need to gather additional information to respond. Through this process of information sharing, the concern is very often addressed, and there is no need to move forward with a request for an investigation.

In other cases, if available data confirm an unusual pattern or clearly elevated level of health outcomes, or if there is information suggesting an unusual exposure, we move to the third box. We gather additional information, if possible, and continue to interact with interested individuals, community groups, and stakeholders as we consider the severity of the health and/or exposure problems, the quality of available data, and the likelihood that we have data, methods, and resources for addressing scientific questions and/or answering the community’s questions.

The fourth box describes a spectrum of options, including no further action or requesting more information from the community. The fourth box names a few types of follow-up investigations that may be appropriate. We are most often able to follow up with health statistics reviews, which use high quality, statewide comprehensive data available from birth certificates, the NYS Congenital Malformations Registry and the NYS Cancer Registry to see if these outcomes are occurring at a higher, lower, or the same level in the community of concern as in the general NYS population.

The Ribbon: How many requests for investigations does the New York State Department of Health (NYS DOH) receive, and how many are pursued? Aside from reviewing requests from the public, how else does NYS DOH determine what situations may require evaluation?

Dr. Lewis-Michl: Cancer concerns are handled by the Health Department’s Cancer Surveillance Program most often, with the CEH handling a subset of these concerns that are more specifically focused on environmental issues. Similarly, most cancer incidence investigations are conducted by the Cancer Surveillance Program, with the CEH conducting some of the more in-depth and smaller area studies involving well-defined exposures. The Health Department receives inquiries about cancer concerns through many programs, totaling to at least 100 per year. The number of inquiries that then lead to specific requests, and then to the conduct of cancer studies is approximately 4-5 per year in recent years. Since 1981, the Health Department has conducted more than 350 cancer-incidence investigations.

We move through the decision-making process on the flow chart for determining appropriate follow-up as part of our routine program activities if an unusual exposure has been documented. In other words, we do not wait for a request from the community if we are already aware of a documented, unusual exposure for which follow-up may be warranted and feasible. We learn of such exposures from sampling investigations conducted routinely by our agency or other agencies. For example, we are currently following up on cancer, low birth weight and birth defects in populations at several sites in NYS with documented exposures to volatile organic compounds (VOCs). This is a multi-site health statistics review (fourth box).

The Ribbon: In 1990, the Centers for Disease Control and Prevention (CDC) issued "Guidelines for Investigating Clusters of Health Events," providing state health departments the framework of a four-stage process for responding to clusters appropriately and efficiently. A monograph published last year, "An Update on Cancer Cluster Activities at the CDC" (reference below) says that “[the Guidelines] retain their original usefulness and validity," while offering the 2007 article as "an addendum for use with the original document." Does the NYS DOH use the CDC Guidelines, and what is the current relationship between DOH cancer-cluster-related activities and those of the CDC?

Dr. Lewis-Michl: While we did not use the CDC Guidelines specifically when we developed the CEH flow chart, the CDC Guidelines describe a process that matches ours in general concepts and many details. The CDC’s four general stages are (1) Initial Contact & Response; (2) Assessment, which emphasizes determining, using existing data, whether an excess has actually occurred and whether the excess can be linked to some exposure; (3) Feasibility study; and (4) Conduct an epidemiological study. The CEH works routinely with the CDC via the Agency for Toxic Substances and Disease Registry (ATSDR), our federal partner for health outcome investigations. Our partner here in NYS DOH, the Cancer Surveillance Program, is in routine contact with the CDC programs that respond to cancer inquiries.

The Ribbon: What type of public involvement or public education is typically involved in an investigation?

Dr. Lewis-Michl: CEH staff work closely with communities and stakeholder groups when such groups come together and request our assistance to address health outcome and environmental issues. Public availability sessions and meetings and other communication tools are used depending on the level of community interest. In some situations, where there is less community interest, CEH may complete a study with very little community input. Ideally, we like to have at least some opportunity for communication with an affected community so that when an investigation is completed, people are already aware of the strengths, limitations, and possible outcomes from such studies.

The Ribbon: At what point in an investigation do possible environmental or occupational exposures come into play?

Dr. Lewis-Michl: As the flow chart shows, exposure issues, along with evidence for a disease elevation, are foremost among the criteria we use when determining appropriate follow-up. We seek information about potential exposures and exposure data from the beginning to understand the nature of the issue. We also have to determine if there are environmental data available, or that could be gathered, of sufficient quality to make an environmental epidemiological follow-up investigation feasible.

The Ribbon: There are several methodological issues that cancer cluster analysis has traditionally not been well equipped to deal with, such as the long latency period of cancer, the multitude of possible exposures, transient populations, and low numbers of cases. Do you see improvements in addressing these challenges?

Dr. Lewis-Michl: For the surveillance or incidence studies conducted for specific geographic areas, the typical studies we do, there is no easy solution for these problems. For resource-intensive studies, more likely to be conducted at medical and academic research centers, there are expanding options for improving cancer studies. There is a growing body of knowledge about biological markers of genetic damage that may be associated with environmental exposures that occurred in the past. If such biological markers can be improved to become more chemical, exposure, or time-frame specific, our methods for studying possible causes of cancer will benefit. Similarly, as scientists learn more about markers of susceptibility for certain cancer types, studies can take account of differences among individuals and better detect the effects of any environmental exposures.

The Ribbon: Many investigations do not yield the explanations or answers that concerned people – who justifiably feel that there is too much cancer around them – are anxious to have. What message do you have for those hoping to learn more than an investigation might reveal?

Dr. Lewis-Michl: I share this frustration about the level of cancer we see around us. The statistics now tell us that in our current US population, one in two men and one in three women will be diagnosed with cancer at some point in their lifetime. Our risk for cancer increases with age, so as the elderly have grown as a proportion of our population, cancer rates have increased. People’s concerns about the numbers of cases of cancer in their neighborhood often arise because they do not realize how frequently cancer occurs. People can take preventive steps, however. Estimates based on studies of cancer risk factors suggest that more than half of cancer deaths are preventable by not smoking, reduced alcohol consumption, healthful nutrition and regular physical activity, in addition to routine cancer screening for early detection.

My group at the CEH focuses on evaluating levels of cancer and adverse birth outcomes, particularly for areas where we have specific information about unusual exposures. While these types of studies do not provide the answers about cause and effect that people may be seeking, they do provide people with information about their community’s health status, how the levels of health problems in their community compare to levels for the general NYS population. These types of health data reviews are especially useful when they focus on an area with documented unusual exposures. The findings from such studies may suggest hypotheses for more in-depth research that can draw stronger conclusions about whether the exposure is causally related to a specific adverse health effect.

Reference
Kingsley, B.S., Schmeichel, K.L., and Rubin, C.H. (2007). An Update on Cancer Cluster Activities at the Centers for Disease Control and Prevention. EHP 115, 165-171.

Back to the top

Last Update 09.09.08