Kernel density analysis reveals a halo pattern of breast cancer incidence in Connecticut

Breast cancer (BC) incidence rates in Connecticut are among the highest in the United States, and are unevenly distributed within the state. Our goal was to determine whether artificial light at night (ALAN) played a role. Using BC records obtained from the Connecticut Tumor Registry, we applied the double kernel density (DKD) estimator to produce a continuous relative risk surface of a disease throughout the State. A multi-variate analysis compared DKD and census track estimates with population density, fertility rate, percent of non-white population, population below poverty level, and ALAN levels. The analysis identified a "halo" geographic pattern of BC incidence, with the highest rates of the disease observed at distances 5-15 km from the state's major cities. The "halo" was of high-income communities, with high ALAN, located in suburban fringes of the state's main cities.

Light at night and breast cancer incidence in Connecticut: An ecological study of age group effects

The aim of this study was to test the prediction that within the state of Connecticut, USA, communities with high nighttime outdoor light level would have higher breast cancer incidence rates. Breast cancer cases were identified from the Connecticut Tumor Registry, the oldest within the United States, for years 2005 and 2009 and geocoded to the 829 census tracts in the state. Nighttime light level (LAN) was obtained from the Defense Meteorological Satellite Program (DMSP), 1996/97 satellite image, providing a 10-year lag. Regression models were used incorporating the LAN levels and census level data on potential confounders for the whole female population of the state, and for separate age groups. Light level emerged as a significant predictor of breast cancer incidence. After taking account of several potential confounders, the excess risk in the highest LAN level census tracts compared to the lowest was about 63% (RR=1.63; 95% CI=1.41, 1.89). The association of LAN with breast cancer incidence weakened with age; the association was strongest among premenopausal women.

Secular trends, race, and geographic disparity of early-stage breast cancer incidence: 25 years of surveillance in Connecticut

OBJECTIVES

We considered changes in the geographic distribution of early stage breast cancer among White and non-White women while secular trends in lifestyle and health care were under way.

METHODS

We aggregated tumor registry and census data by age, race, place of residence, and year of diagnosis to evaluate rate variation across Connecticut census tracts between 1985 and 2009. Global and local cluster detection tests were completed.

RESULTS

Age-adjusted incidence rates increased by 2.71% and 0.44% per year for White and non-White women, respectively. Significant global clustering was identified during surveillance of these populations, but the elements of clustering differed between groups. Among White women, fewer local clusters were detected after 1985 to 1989, whereas clustering increased over time among non-White women.

CONCLUSIONS

Small-area variation of breast cancer incidence rates across time periods proved to be dynamic and race-specific. Incidence rates might have been affected by secular trends in lifestyle or health care. Single cross-sectional analyses might have confused our understanding of disease occurrence by not accounting for the social context in which patient preferences or provider capacity influence the numbers and locations of diagnosed cases. Serial analyses are recommended to identify "hot spots" where persistent geographic disparities in incidence occur.

Who's assessing tobacco use in cancer clinical trials?

INTRODUCTION

Clinical trials that do not collect data on tobacco use/exposure may not adequately assess the efficacy and effectiveness of experimental treatments.

METHODS

A cross-sectional study of interventional trials cited on ClinicalTrials.gov was undertaken that inquired of Local Project Directors from Connecticut guiding studies of breast, prostate, or colorectal cancer chemotherapy (N = 68) whether their protocols measured tobacco use by trial participants. Information pertaining to 46 trials (68%) is reported here. All but 1 were multicentered trials enrolling patients around the country.

RESULTS

Only 3 trials (7%) reported routine collection of tobacco use information at baseline and no trial reported monitoring tobacco use during treatment follow-up. None of the 3 trials collecting tobacco data reported using exposure information in analysis of treatment effects. Survey respondents suggested that uncertainty about the relevance of tobacco exposure to therapeutic efficacy, ambivalence about how to incorporate such data into analyses, insufficient resources for collecting such information, and uncertainty about the validity of assessment methods might be reasons why tobacco use is not routinely assessed.

DISCUSSION

Additional studies that address a fuller range of cancers, therapies, disease states, and clinical environments are needed to fully define the extent of this data lapse. Providing clinicians and trialists with appropriate tools for tobacco use assessment and encouraging them to collect such information about patients during treatment and follow-up may offer a simple cost-effective way to improve the quality and consequences of cancer care for every patient.

Place of residence effect on likelihood of surviving prostate cancer

PURPOSE

To examine geographic variation in survival time of men diagnosed with prostate cancer, adjusted for patient and disease characteristics.

METHOD

Survival times for a geographically referenced database of 27,189 incident prostate cancer cases (ICD-O-2: C61.9) from Connecticut, 1984-1998, were evaluated using a newly developed extension of the spatial scan statistic for survival data.

RESULTS

Statewide, median survival time was 4.6 years following diagnosis. Age-adjusted survival times across most locales around Connecticut did not differ markedly from the statewide pattern, but our analysis revealed 3 zones with noteworthy differences. Analysis of survival times adjusted for age as well as tumor grade and stage produced only two locations with significant results, and further adjustment for racial composition of cases yielded only one location with significant distinct (lower) survival times. Among cases within that place, the likelihood of dying was estimated to be 1.39-times greater than that of cases different from those diagnosed elsewhere around the state (p = 0.009).

CONCLUSION

The prognosis for men with prostate cancer may differ, in part, by virtue of where they live when diagnosed. Measuring geographic differences in survival time should facilitate the targeting of clinical and ancillary services to persons at high risk of poor outcomes.

Race-specific geography of prostate cancer incidence

Background

This study evaluated geographic distribution of race-specific prostate cancer incidence in Connecticut and Massachusetts. This cross-sectional analysis of census and cancer registry data included records of 29,040 Whites and 1,647 African Americans diagnosed with incident prostate cancer between 1994 and 1998. A spatial scan statistic was used to detect and test significance of the geographic variation in race-specific incidence rates within the two-state area.

Results

Significant geographic variation in age-adjusted incidence rates among both White and African American men was observed, with little overlap noted between distributions. Identified locations reflected patterns of residential segregation and socio-economic conditions. Among Whites, places with higher than expected incidence had higher socioeconomic status than places with lower than expected incidence. No discernable relationship between social indicators and rate variation among African Americans was evident.

Conclusion

Differences in race-specific geographic distribution of prostate cancer incidence do not suggest a shared environmental etiology. Furtherstudyof genetic, behavioral and health carefactors affecting the occurrence and/or reporting of the disease is warranted. This study highlights the need for race- and geographic-specific interventions to better control disease within at-risk communities and for on-going analysis into social and contextual factors that contribute to observed disparities between African Americans and Whites in the occurrence of cancer.

Effects of study area size on geographic characterizations of health events: prostate cancer incidence in Southern New England, USA, 1994-1998

BACKGROUND

We consider how representations of geographic variation in prostate cancer incidence across Southern New England, USA may be affected by selection of study area and/or properties of the statistical analysis.

METHOD

A spatial scan statistic was used to monitor geographic variation among 35,167 incident prostate cancer cases diagnosed in Massachusetts, Connecticut and Rhode Island from 1994 to 1998, in relation to the 1990 populations of men 20+ years of age living in that region. Results from the combined-states analysis were compared to those from single-states. Impact of scanning procedures set to examine up to 50% or no more than 10% of at-risk populations also was evaluated.

RESULTS

With scanning set to 50%, 5 locations in the combined-states analysis were identified with markedly distinct incidence rates. Fewer than expected cases were estimated for nearly all Connecticut, Rhode Island and West Central Massachusetts, whereas census tracts on and around Cape Cod, and areas of Southwestern Connecticut and adjacent to greater Boston were estimated to have yielded more than expected incidence. Results of single-state analyses exhibited several discrepancies from the combined-states analysis. More conservative scanning found many more locations with varying incidence, but discrepancies between the combined- and single-state analysis were fewer.

CONCLUSION

It is important to acknowledge the conditional nature of spatial analyses and carefully consider whether a true cluster of events is identified or artifact stemming from selection of study area size and/or scanning properties.

Cancer map patterns: are they random or not?

BACKGROUND

Maps depicting the geographic variation in cancer incidence, mortality or treatment can be useful tools for developing cancer control and prevention programs, as well as for generating etiologic hypotheses. An important question with every cancer map is whether the geographic pattern seen is due to random fluctuations, as by pure chance there are always some areas with more cases than expected, or whether the map reflects true underlying geographic variation in screening, treatment practices, or etiologic risk factors.

METHODS

Nine different tests for spatial randomness are evaluated in very practical settings by applying them to cancer maps for different types of data at different scales of spatial resolution: breast, prostate, and thyroid cancer incidence; breast cancer treatment and prostate cancer stage in Connecticut; and nasopharynx and prostate cancer mortality in the U.S.

RESULTS

Tango's MEET, Oden's Ipop, and the spatial scan statistic performed well across all the data sets. Besag-Newell's R, Cuzick-Edwards k-NN, and Turnbull's CEPP often perform well, but the results are highly dependent on the parameter chosen. Moran's I performs poorly for most data sets, whereas Swartz Entropy Test and Whittemore's Test perform well for some data sets but not for other.

CONCLUSIONS

When publishing cancer maps we recommend evaluating the spatial patterns observed using Tango's MEET, a global clustering test, and the spatial scan statistic, a cluster detection test.

Lumping or splitting: seeking the preferred areal unit for health geography studies

Background

Findings are compared on geographic variation of incident and late-stage cancers across Connecticut using different areal units for analysis.

Results

Few differences in results were found for analyses across areal units. Global clustering of incident prostate and breast cancer cases was apparent regardless of the level of geography used. The test for local clustering found approximately the same locales, populations at risk and estimated effects. However, some discrepancies were uncovered.

Conclusion

In the absence of conditions calling for surveillance of small area cancer clusters ('hot spots'), the rationale for accepting the burdens of preparing data at levels of geography finer than the census tract may not be compelling.

Geographic distribution of prostate cancer incidence in the era of PSA testing, Connecticut, 1984 to 1998

OBJECTIVES

To evaluate the geographic variation of prostate cancer incidence in Connecticut during a 15-year period: before the introduction of prostate-specific antigen (PSA) testing (1984 to 1988), during its introduction (1989 to 1993), and after its widespread adoption (1994 to 1998).

METHODS

A spatial scan statistic was used to detect and test the significance of the geographic rate variation among 29,944 incident cases.

RESULTS

During 1984 to 1988, eight locations had incidence rates that differed significantly from the statewide level; for 1989 to 1993, only two locations were identified. By 1994 to 1998, a significant rate variation was noted for 10 locations around Connecticut.

CONCLUSIONS

The observed geographic variation of incidence rates may reflect differing opportunities to uncover latent cases at given locales.

Breast cancer surveillance using gridded population units, Connecticut, 1992 to 1995

PURPOSE

To assess geographic variation in invasive breast cancer across Connecticut using gridded population areas to enumerate cases and the population at-risk.

METHODS

The state's land mass was divided into 5168, 1-by-1 square mile areas and the population of women, 20+ years of age, within each location was estimated by areal interpolation of the 1990 US Census Block Group STF-3A data file. Using information on breast cancer incidence, 1992 to 1995, from the Connecticut Tumor Registry, latitude-longitude coordinates for place of residence at the time of breast cancer diagnosis were determined for 8530 records and assigned to appropriate grid locations. A spatial scan statistic was used to detect variation in incidence and test the significance of observed differences across the state. Standardized Incidence Ratios (SIRs) described the proportional change in the age-adjusted breast cancer incidence rate across gridded locations.

RESULTS

The statewide age-adjusted invasive cancer incidence rate was 163.6/100,000 women/year. The spatial scan statistic identified three locations around Connecticut with significantly low incidence rates and four places where rates were significantly high. The most probable place of low incidence was rural Northeastern Connecticut where risk of disease, relative to elsewhere around the state, was 0.73 (p = 0.001). The most probable location of elevated incidence was a suburban location in Southwestern Connecticut with a relative risk of 2.02 (p = 0.001).

CONCLUSIONS

Visual representation of disease incidence and underlying populations at-risk according to gridded units provides a useful tool for assessing small area variation in disease patterns.

Geographic differences in invasive and in situ breast cancer incidence according to precise geographic coordinates, Connecticut, 1991-95

To evaluate geographical variation of invasive and in situ breast cancer incidence rates using precise geographical coordinates for place of residence at diagnosis, latitude-longitude coordinates pertaining to 10,601 invasive and 1,814 in situ breast cancers for Connecticut women, 1991-95, were linked to US Census information on the 2,905 State census block groups. A spatial scan statistic was used to detect geographic excess or deficits in incidence and test the statistical significance of results, without prior assumptions about the size or location of such areas. The age adjusted invasive cancer incidence rate was 165.3/100,000 women/year. The spatial scan statistic identified 3 places with significantly low incidence rates and 4 places where rates were significantly high. The most probable location of low incidence was rural northeastern Connecticut where risk of disease, relative to elsewhere around the state, was 0.70 (p = 0.0001); the most probable place of elevated incidence was north central Connecticut where a relative risk of 1.34 (p = 0.002) was observed. Incidence of in situ disease was estimated to be significantly high for north central Connecticut (RR = 1.84; p = 0.0001). Geographic differences of invasive and in situ breast cancer incidence were observed. Examining cancer events at the lowest available level of data aggregation is beneficial in highlighting localized rate variations. Such information may enable public health officials to target additional resources for promoting breast cancer screening to specific locations.