Cancer Cluster Investigations Research Articles

Cancer cluster among small village residents near the fertilizer plant in Korea.

In Jang-jeom, a small village in Hamra-myeon, Iksan-si, Jeollabuk-do, South Korea, residents raised concerns about a suspected cancer cluster that they attributed to a fertilizer plant near the village. We aimed to investigate whether the cancer incidence in the village was higher than that in the general Korean population when the factory was in operation (2001-2017) and whether living in the village was associated with a higher risk of cancer. Using national population data and cancer registration data of South Korea, we estimated the standardized incidence ratios (SIRs) in the village to investigate whether more cancer cases occurred in the village compared to other regions. The SIRs were standardized by age groups of 5 years and sex. In order to investigate whether residence in the village increased the risk of cancer, a retrospective cohort was constructed using National Health Insurance Service (NHIS) databases. We estimated the cancer hazard ratios (HRs) using the Cox proportional hazard model, and defined the exposed area as the village of Jang-jeom, and the unexposed or control area as the village neighborhood in Hamra-myeon. We considered potential confounding variables such as age, sex, and income index in the models. Additionally, we measured polycyclic aromatic hydrocarbons (PAHs) and tobacco-specific nitrosamines (TSNAs), suspected carcinogens that may have caused the cancer cluster, in samples collected from the plant and the village. Twenty-three cancer cases occurred in Jang-jeom from 2001 to 2017. Between 2010 and 2016, the incidence rates of all cancers (SIR: 2.05, except thyroid cancer: 2.22), non-melanoma skin cancer (SIR: 21.14, female: 25.41), and gallbladder (GB) and biliary tract cancer in men (SIR: 16.01) in the village were higher than those in the national population in a way that was statistically significant. In our cohort analysis that included only Hamra-myeon residents who have lived there for more than 7 years, we found a statistically significant increase in the risk of all cancers (HR: 1.99, except thyroid cancer: 2.20), non-melanoma skin cancer (HR: 11.60), GB and biliary tract cancer (HR: 15.24), liver cancer (HR: 6.63), and gastric cancer (HR: 3.29) for Jang-jeom residents compared to other Hamra area residents. We identified PAHs and TSNAs in samples of deposited dust and residual fertilizer from the plant and TSNAs in dust samples from village houses. The results of the SIR calculation and cancer risk analyses of Jang-jeom village residents from the retrospective cohort design showed consistency in the effect size and direction, suggesting that there was a cancer cluster in Jang-jeom. This study would be a good precedent for cancer cluster investigation.

A causal inference framework for cancer cluster investigations using publicly available data.

Often, a community becomes alarmed when high rates of cancer are noticed, and residents suspect that the cancer cases could be caused by a known source of hazard. In response, the US Centers for Disease Control and Prevention recommend that departments of health perform a standardized incidence ratio (SIR) analysis to determine whether the observed cancer incidence is higher than expected. This approach has several limitations that are well documented in the existing literature. In this paper we propose a novel causal inference framework for cancer cluster investigations, rooted in the potential outcomes framework. Assuming that a source of hazard representing a potential cause of increased cancer rates in the community is identified a priori, we focus our approach on a causal inference estimand which we call the causal SIR (cSIR). The cSIR is a ratio defined as the expected cancer incidence in the exposed population divided by the expected cancer incidence for the same population under the (counterfactual) scenario of no exposure. To estimate the cSIR we need to overcome two main challenges: 1) identify unexposed populations that are as similar as possible to the exposed one to inform estimation of the expected cancer incidence under the counterfactual scenario of no exposure, and 2) publicly available data on cancer incidence for these unexposed populations are often available at a much higher level of spatial aggregation (e.g. county) than what is desired (e.g. census block group). We overcome the first challenge by relying on matching. We overcome the second challenge by building a Bayesian hierarchical model that borrows information from other sources to impute cancer incidence at the desired level of spatial aggregation. In simulations, our statistical approach was shown to provide dramatically improved results, i.e., less bias and better coverage, than the current approach to SIR analyses. We apply our proposed approach to investigate whether trichloroethylene vapor exposure has caused increased cancer incidence in Endicott, New York.

Barium exposure of an invasive breast cancer cluster investigation – quantitative drinking water chemistry for carcinogen search

Abstract Quantitative review of oral barium exposure was conducted using drinking water metals data from the confirmed sixfold (18-fold15–39 years) Brisbane Australia invasive breast cancer cluster (IBC) investigation, which also included fivefold duct carcinoma in situ (DCIS). The new approach here was to assess sub-guideline on-site water system barium fluctuations and use cancer slope factor (CSF) firstly to profile match barium as a carcinogen, then correlate risk. Dissolved barium was converting, up to 71%, into non-dissolved forms in the cluster location; with only up to 3% in a comparison location. A CSF of approximately 330 (mg L−1)−1 gave a chemistry-based number of women of 12, consistent with 14 originally (mostly IBC), less 2 per general population biostatistics. The comparison location gave null women numbers. One-woman IBC correlated to 0.0018 mg/L barium, comparable to the US Environmental Protection Agency's (USEPA) 0.002 mg/L, 10−4 cancer risk for arsenic. The two- to fivefold San Diego USA IBC cluster had barium as the only metal in common with Brisbane; and, breast cancer alcohol risk factor fingered barium qualitatively. It is put that barium causes IBC and DCIS; proof submitted hereby to wider scrutiny. The relevant barium exposure was likely calcium coprecipitated.

Brain Cancer Cluster Investigation Around a Factory Emitting Dichloromethane

Background The health risks associated with dichloromethane (DCM) for the general population living near industrial activities have not yet been quantified, primarily due to lack of epidemiological datasets. In the absence of such human data, we undertook a cancer cluster investigation in Cyprus around a historically using DCM plant producing shoe soles that were globally exported. We designed the methodology to investigate the possible existence of a cancer cluster in the area around the factory (point zero) and within a radius of 500 meters. Methods A retrospective comparative population study was designed using a group of cancer patients living or working in the chosen geographical area around the factory. Results Mean stack emissions of DCM of 88 mg/Nm3 and flow rates of 850 g/h exceeded the permissible DCM limits established for industrial zones. Brain and central nervous system (CNS) cancer incidence rates showed significant (P < 0.001) increase in the study area around the plant when compared with those observed in other areas of Cyprus. Calculated standardized incidence ratios for brain/CNS after adjusting for the age at diagnosis ranged from 11.3-25.7 [mean 6.5 (3.02 : 12.3)] for the study area. Conclusions We showed the association between chronic, unintentional DCM exposures and brain/CNS cancer cases for the general population located in a residential area being in close proximity with a plant historically emitting DCM.

Brain cancer cluster investigation around a factory emitting dichloromethane.

The health risks associated with dichloromethane (DCM) for the general population living near industrial activities have not yet been quantified, primarily due to lack of epidemiological datasets. In the absence of such human data, we undertook a cancer cluster investigation in Cyprus around a historically using DCM plant producing shoe soles that were globally exported. We designed the methodology to investigate the possible existence of a cancer cluster in the area around the factory (point zero) and within a radius of 500 meters. A retrospective comparative population study was designed using a group of cancer patients living or working in the chosen geographical area around the factory. Mean stack emissions of DCM of 88 mg/Nm3 and flow rates of 850 g/h exceeded the permissible DCM limits established for industrial zones. Brain and central nervous system (CNS) cancer incidence rates showed significant (P < 0.001) increase in the study area around the plant when compared with those observed in other areas of Cyprus. Calculated standardized incidence ratios for brain/CNS after adjusting for the age at diagnosis ranged from 11.3-25.7 [mean 6.5 (3.02 : 12.3)] for the study area. We showed the association between chronic, unintentional DCM exposures and brain/CNS cancer cases for the general population located in a residential area being in close proximity with a plant historically emitting DCM.

The completeness and timeliness of cancer registration and the implications for measuring cancer burden

BackgroundPopulation based cancer registration provides a critical role in disease surveillance in terms of incidence, survival, cancer cluster investigations and prevalence trends, and therefore high levels of completeness and timeliness are required. This study estimates completeness and variation between early and late registrations in the N. Ireland Cancer Registry (NICR) and assesses the implications for reporting cancer incidence and for registry-based research. MethodsTwo main approaches assessed completeness. For the period 2010–2012, incidence reported in the first year of data publication was compared to incidence reported in subsequent years until 2015. Demographic characteristics and survival of incident cases ascertained before the first publication year were compared to those ascertained in subsequent years. The flow method approach was used to estimate completeness annually after the incident year. ResultsOverall incidence for all cancers increased between the first year of data publication and subsequent years up to 2015, irrespective of year of diagnosis. Late registrations had poorer survival. The flow method approach estimated the completeness of case ascertainment of NICR data to be 96% complete at five years for all cancers combined. ConclusionThe estimated completeness levels for the NICR are comparable to other high quality cancer registries internationally. While data timeliness has little impact on incidence estimates, delays in registration may have implications for specific research studies into incidence and survival. This means that improvements in the timeliness of reporting should be a target for all registries but not at the expense of completeness.

Geospatial cryptography: enabling researchers to access private, spatially referenced, human subjects data for cancer control and prevention.

As the volume, accuracy and precision of digital geographic information have increased, concerns regarding individual privacy and confidentiality have come to the forefront. Not only do these challenge a basic tenet underlying the advancement of science by posing substantial obstacles to the sharing of data to validate research results, but they are obstacles to conducting certain research projects in the first place. Geospatial cryptography involves the specification, design, implementation and application of cryptographic techniques to address privacy, confidentiality and security concerns for geographically referenced data. This article defines geospatial cryptography and demonstrates its application in cancer control and surveillance. Four use cases are considered: (1) national-level de-duplication among state or province-based cancer registries; (2) sharing of confidential data across cancer registries to support case aggregation across administrative geographies; (3) secure data linkage; and (4) cancer cluster investigation and surveillance. A secure multi-party system for geospatial cryptography is developed. Solutions under geospatial cryptography are presented and computation time is calculated. As services provided by cancer registries to the research community, de-duplication, case aggregation across administrative geographies and secure data linkage are often time-consuming and in some instances precluded by confidentiality and security concerns. Geospatial cryptography provides secure solutions that hold significant promise for addressing these concerns and for accelerating the pace of research with human subjects data residing in our nation's cancer registries. Pursuit of the research directions posed herein conceivably would lead to a geospatially encrypted geographic information system (GEGIS) designed specifically to promote the sharing and spatial analysis of confidential data. Geospatial cryptography holds substantial promise for accelerating the pace of research with spatially referenced human subjects data.

Residential cancer cluster investigation nearby a Superfund Study Area with trichloroethylene contamination.

Trichloroethylene (TCE) is an industrial solvent associated with liver cancer, kidney cancer, and non-Hodgkin's lymphoma (NHL). It is unclear whether an excess of TCE-associated cancers have occurred surrounding the Middlefield-Ellis-Whisman Superfund site in Mountain View, California. We conducted a population-based cancer cluster investigation comparing the incidence of NHL, liver, and kidney cancers in the neighborhood of interest to the incidence among residents in the surrounding four-county region. Case counts and address information were obtained using routinely collected data from the Greater Bay Area Cancer Registry, part of the Surveillance, Epidemiology, and End Results program. Population denominators were obtained from the 1990, 2000, and 2010 US censuses. Standardized incidence ratios (SIRs) with two-sided 99 % confidence intervals (CIs) were calculated for time intervals surrounding the US Censuses. There were no statistically significant differences between the neighborhood of interest and the larger region for cancers of the liver or kidney. A statistically significant elevation was observed for NHL during one of the three time periods evaluated (1996-2005: SIR = 1.8, 99 % CI 1.1-2.8). No statistically significant NHL elevation existed in the earlier 1988-1995 (SIR = 1.3, 99 % CI 0.5-2.6) or later 2006-2011 (SIR = 1.3, 99 % CI 0.6-2.4) periods. There is no evidence of an increased incidence of liver or kidney cancer, and there is a lack of evidence of a consistent, sustained, or more recent elevation in NHL occurrence in this neighborhood. This evaluation included existing cancer registry data, which cannot speak to specific exposures incurred by past or current residents of this neighborhood.

Investigation of spatio-temporal cancer clusters using residential histories in a case-control study of non-Hodgkin lymphoma in the United States.

BackgroundNon-Hodgkin lymphoma (NHL) is an enigmatic disease with few known risk factors. Spatio-temporal epidemiologic analyses have the potential to reveal patterns that may give clues to new risk factors worthy of investigation. We sought to investigate clusters of NHL through space and time based on life course residential histories.MethodsWe used residential histories from a population-based NHL case–control study of 1300 cases and 1044 controls with recruitment centers in Iowa, Detroit, Seattle, and Los Angeles, and diagnosed in 1998–2000. Novel methods for cluster detection allowing for residential mobility, called Q-statistics, were used to quantify nearest neighbor relationships through space and time over the life course to identify cancer clusters. Analyses were performed on all cases together and on two subgroups of NHL: Diffuse large B-cell lymphoma and follicular lymphoma. These more homogenous subgroups of cases might have a more common etiology that could potentially be detected in cluster analysis. Based on simulation studies designed to help account for multiple testing across space and through time, we required at least four significant cases nearby one another to declare a region a potential cluster, along with confirmatory analyses using spatial-only scanning windows (SaTScan).ResultsEvidence of a small cluster in southeastern Oakland County, MI was suggested using residences 10–18 years prior to diagnosis, and confirmed by SaTScan in a time-slice analysis 20 years prior to diagnosis, when all cases were included in the analysis. Consistent evidence of clusters was not seen in the two histologic subgroups.ConclusionsSuggestive evidence of a small space-time cluster in southeastern Oakland County, MI was detected in this NHL case–control study in the USA.

Stop and listen to the people: an enhanced approach to cancer cluster investigations.

Cancer cluster investigations need to address the disconnect between traditional public health approaches and human needs. Cancer cluster investigations often magnify fear and uncertainty because they rarely find a definitive environmental cause. Traditional approaches emphasize population-level data analysis and undervalue active listening. Because few studies have explored active listening in cancer cluster investigations, we conducted a descriptive oral history case study of a Frederick, Maryland, investigation. We interviewed 12 community members and 9 public health professionals about the investigation of a perceived cancer cluster. Many believed it was linked to environmental contamination at Fort Detrick, a local US Army base. We propose enhanced active listening that seeks out peoples' perspectives, validates their concerns, and engages them in the investigative process.

Cancer cluster investigations: review of the past and proposals for the future.

Residential clusters of non-communicable diseases are a source of enduring public concern, and at times, controversy. Many clusters reported to public health agencies by concerned citizens are accompanied by expectations that investigations will uncover a cause of disease. While goals, methods and conclusions of cluster studies are debated in the scientific literature and popular press, investigations of reported residential clusters rarely provide definitive answers about disease etiology. Further, it is inherently difficult to study a cluster for diseases with complex etiology and long latency (e.g., most cancers). Regardless, cluster investigations remain an important function of local, state and federal public health agencies. Challenges limiting the ability of cluster investigations to uncover causes for disease include the need to consider long latency, low statistical power of most analyses, uncertain definitions of cluster boundaries and population of interest, and in- and out-migration. A multi-disciplinary Workshop was held to discuss innovative and/or under-explored approaches to investigate cancer clusters. Several potentially fruitful paths forward are described, including modern methods of reconstructing residential history, improved approaches to analyzing spatial data, improved utilization of electronic data sources, advances using biomarkers of carcinogenesis, novel concepts for grouping cases, investigations of infectious etiology of cancer, and “omics” approaches.

Inflammatory breast cancer clusters: A hypothesis.

Cancer clusters have long been a focus of interest because of the possibility of identifying etiologic agents. Only on rare occasions, however, have such cluster investigations been successful. One major difficulty in cluster investigations, particularly in the area of breast cancer, is the long latent period. There have been a number of publications providing a discouraging picture regarding cancer cluster investigations. The possibility of learning from a cluster investigation, however, is greatly increased if the cancer involved is relatively rare and if it has a short latent period. Inflammatory breast cancer (IBC) fits these criteria and is worth pursuing because of the strong evidence that environmental factors play a major role. In this report we describe our experience with several clusters and the lessons learned which are now being utilized to improve investigation of future IBC clusters. The first IBC cluster that we evaluated was in 2000, when we were asked to investigate an apparent cluster of IBC in Castro Valley, California where three women in an office setting of 24 people were diagnosed with IBC in a ten month period from May 1999 to March 2000. Our investigation of this striking cluster did not yield a specific trigger for this cluster but it did indicate that the women involved all had at least two IBC risk factors that may well have made them susceptible to getting IBC. We are now investigating another apparent cluster in Texas and are aware of several others requiring careful consideration. We see a need for a consistent protocol for the evaluation of IBC clusters focusing on the laboratory investigation of environmental triggers, primarily infectious agents and chemical carcinogens.

An extensive epidemiological investigation of a kidney cancer cluster in a chemical plant: what have we learned?

ObjectivesIn 2003, a cluster of renal cell carcinoma (RCC) cases was reported among men working at a French chemical plant using a proprietary process to produce vitamin A. The 10...

Commentary: Assessment of chance should be central in investigation of cancer clusters

As a biostatistician who has considerable experience investigating both fixed putative source clusters and general cluster alarms, and developing methodology for these analyses, I feel I should respond to the various points raised by Coory and Jordan. I believe the authors provide a very narrow view of cluster assessment, characterizing it as testing for an excess of incident cases in a given area. A much broader view has developed over the last two decades, as described below.

Commentary: Statistical assessment of cancer cluster evidence--in search of a middle ground

As Coory and Jordan point out, probabilities linked to cancer cluster reports must take into account the cherry-picking nature of the reported clusters in order to avoid a large number of false-positive results. These authors must be congratulated for bringing this difficult issue to debate and for summarizing several of the challenges facing researchers. In fact, the challenges are so large that the authors conclude that statistical assessments are impossible to interpret and, rather than assisting, they impair decisionmaking. Their final recommendation is to remove the statistical assessment of chance from protocols for investigating clusters in favour of emphasizing exposure assessment. A principal feature of cancer cluster reports in most countries has been the quick dismissal of the large majority of cases after initial contact. Even for the minority in which an investigation is conducted, neither a plausible biological cause nor an environmental exposure is commonly identified. Goodman et al. evaluated more than 400 reports from cancer cluster investigations carried out since 1990 in the USA, evaluating more than 500 cancers of concern. A statistical increase in incidence was confirmed in 72 of these, but only three were linked to hypothetical exposures, and only one identified a clear cause. The main statistical concern of Coory and Jordan is what they call ‘silent multiple comparisons’ when calculating P-values and confidence intervals associated with the significance of the cancer clusters. Their criticism is valid, since most of these statistical summaries are calculated as if the reported cancer cluster were not picked from among numerous other potential clusters. These unaccounted potential clusters compose the silent comparisons. The authors discuss the use of Bonferroni-type corrections to calculate an approximation for the true P-value associated with given evidence. Coory and Jordan mention also the harder problem that a reference population can be defined arbitrarily. Considering an example of a cluster at the city of Brisbane, they mention that a putative cluster could be evaluated with respect to the population of Brisbane, or the state of Queensland, or the entire Australian population. They show that the P-value can be quite different depending on this choice, if using Bonferroni corrections. There are two problems with this argument. The first is that the public health officials in charge of replying to the cancer reports typically have a population to whom they are accountable. This is the natural reference population. The second problem is that, to carry out the authors’ recommendations, at some point some reference population must be selected, and the conclusions will depend on this selection. The common definition for a disease cluster is the occurrence of a greater than expected number of cases of a particular disease within a group of people, a geographical area or a period of time. It will be virtually impossible to start any cluster investigation, including probing for exposure assessment as the authors suggest, if we do not have a measure of how extreme the occurrence is. One common measure, suggested by the authors, is the standardized incidence ratio, SIR1⁄4O/E, where O is the observed number of cases reported for the cluster, and E is the expected number of cases that would have occurred inside the cluster by taking into account age-specific rates of the entire reference population. Therefore, the SIR itself needs a reference population. Its calculation would also produce different results if the Brisbane, Queensland or Australian population is taken as the reference population, although its impact on E is likely to be much smaller than in Bonferroni correction calculations. Public health officials should have this reference population predetermined before any cluster reports are made. Concerning the multiple comparison problem, statisticians have been working hard as this problem appears in many other application areas such as, for Published by Oxford University Press on behalf of the International Epidemiological Association

Race and colorectal cancer survival

Anthony S Robbins speaks to Ruth Williamson, Commissioning Editor: Dr Robbins received a BS in Biochemistry from the University of California, a MD from the University of Southern California School of Medicine, a MPH in Epidemiology from Loma Linda University and a PhD in Epidemiology from Stanford University. While at Stanford he also completed a fellowship in Health Services Research. He served for 13 years on active duty as a medical officer in the US Air Force, spending much of that time directing and conducting both observational and interventional epidemiologic research studies. In 2005, Dr Robbins separated from active military service and joined the California Cancer Registry, where he spent 2 years as a regional cancer epidemiologist responsible for cancer surveillance and cluster investigation for a 16-county region covering much of northern California. During this time Dr Robbins also used statewide California Cancer Registry data to conduct numerous epidemiologic studies examining racial/ethnic differences in cancer treatment and survival, as well as the relationship between county-level changes in breast cancer incidence and the use of hormone therapy in California. He joined the Research Department at the American Cancer Society in 2008 as Director of Health Services Research. Dr Robbins’ research in the areas of cancer epidemiology and health services research has focused on cancer health disparities; specifically, geographic and racial/ethnic differences in cancer incidence and mortality, and racial/ethnic differences in cancer survival and cancer treatment. The goal of his research is to move from why cancer health disparities occur to how they can be prevented. Dr Robbins also serves on a number of committees and workgroups of the American College of Surgeons Commission on Cancer.

Nuclear power plants cleared of leukaemia link

Investigation of cancer clusters should turn to non-radiation causes, say British researchers.

Regional variations in testicular cancer rates in Ireland: clusters and communities

We welcome Alsinnawi et al.’s [1] paper on rates of testicular cancer. However, the authors imply an association with persistent organic pollutants (POPs), particularly polychlorinated biphenyls (PCBs) and higher socio-economic groups that does not stand up to scrutiny. No new evidence of a link between testicular cancer and POPs is presented in this paper. The authors state that PCBs and testicular cancer are ‘directly linked’. The authors cite two small studies which show a weak positive association between testicular cancer and PCBs. A recent, large and well-designed case–control study shows an inverse association between testicular cancer and several individual and grouped PCBs [2]. A more comprehensive synthesis of the total body of evidence is needed. They state ‘there does certainly seem to be a link with higher socio-economic status’. An aggregate of electoral districts in Cork with higher socio-economic status (SES) may have a greater incidence of testicular cancer, but this does not mean that testicular cancer is occurring at a greater frequency in men in higher socio-economic groups. SES was not recorded for individual men who had testicular cancer in this study. This is potential ecological bias [3]. The socio-economic status of cases is not collected by the National Cancer Registry of Ireland (NCRI). This is important when cancer rates for a small area are published. The socio-economic composition of a small area may be different from that of the comparison areas. Many specific cancers are linked to both socio-economic group and other known causes of cancer (smoking for instance). The confounding effect of social class on cancer, particularly in small areas, should always be considered before hypothesising a novel carcinogen. Worldwide, there is a strong public perception that environmental exposures are directly responsible for cancer. This is seldom the case. The epidemiological criteria for successful investigation of cancer clusters are a high relative risk (between 10 and 20); a rare cancer; the purported causative agent can be reliably measured in the environment; bio-monitoring is possible; there is heterogeneity of exposure in an area and a plausible route of exposure can be assessed [4]. These conditions are often not met, and most of them are not present in this study. However, the epidemiology may not matter as much as we think [5]. The strategic learning from the vast literature on cancer clusters is that regardless of the quality and validity of epidemiological and environmental investigation, if there has not been early and effective communication with the public, the scientific approach will not satisfy [6]. Publishing cancer rates without a comprehensive communication and investigation plan is not advised. This paper may cause anxiety, while it offers no plan of action and ignores the necessity of dealing with the community’s risk perception. Ireland needs a dedicated multi-disciplinary environment and health unit, with epidemiological, environmental and Public Health expertise. This unit could then respond to possible cancer clusters and other community environment and health concerns in a timely, coordinated, U. B. Fallon (&) Department of Public Health Medicine, HSE Dublin Mid-Leinster, Tullamore, Co. Offaly, Ireland e-mail: unab.fallon@hse.ie

A Multidisciplinary Investigation of a Polycythemia Vera Cancer Cluster of Unknown Origin

Cancer cluster investigations rarely receive significant public health resource allocations due to numerous inherent challenges and the limited success of past efforts. In 2008, a cluster of polycythemia vera, a rare blood cancer with unknown etiology, was identified in northeast Pennsylvania. A multidisciplinary group of federal and state agencies, academic institutions, and local healthcare providers subsequently developed a multifaceted research portfolio designed to better understand the cause of the cluster. This research agenda represents a unique and important opportunity to demonstrate that cancer cluster investigations can produce desirable public health and scientific outcomes when necessary resources are available.

Local Newspaper Coverage of Health Authority Fairness During Cancer Cluster Investigations

This study explores local newspaper coverage of official cancer cluster investigations in seven different communities. The authors developed the underlying coding schema based on social-psychological research about how individuals assess the fairness of decision makers. The analysis therefore focuses on the degree to which news coverage describes science authorities as being procedurally just, which includes relational fairness and voice. The findings suggest that a significant amount of cancer cluster news coverage includes content consistent with procedural aspects of fairness. The authors also found substantial content suggesting that decision makers should “do more” to address community concerns. The authors hypothesize that such content may also be relevant to procedural justice. The discussion addresses the theoretical and practical use of additional research examining the media portrayal of scientific authorities as fair or unfair.