Geographical, temporal and racial disparities in late-stage prostate cancer incidence across Florida: A multiscale joinpoint regression analysis

Background

Although prostate cancer-related incidence and mortality have declined recently, striking racial/ethnic differences persist in the United States. Visualizing and modelling temporal trends of prostate cancer late-stage incidence, and how they vary according to geographic locations and race, should help explaining such disparities. Joinpoint regression is increasingly used to identify the timing and extent of changes in time series of health outcomes. Yet, most analyses of temporal trends are aspatial and conducted at the national level or for a single cancer registry.

Methods

Time series (1981-2007) of annual proportions of prostate cancer late-stage cases were analyzed for non-Hispanic Whites and non-Hispanic Blacks in each county of Florida. Noise in the data was first filtered by binomial kriging and results were modelled using joinpoint regression. A similar analysis was also conducted at the state level and for groups of metropolitan and non-metropolitan counties. Significant racial differences were detected using tests of parallelism and coincidence of time trends. A new disparity statistic was introduced to measure spatial and temporal changes in the frequency of racial disparities.

Results

State-level percentage of late-stage diagnosis decreased 50% since 1981; a decline that accelerated in the 90's when Prostate Specific Antigen (PSA) screening was introduced. Analysis at the metropolitan and non-metropolitan levels revealed that the frequency of late-stage diagnosis increased recently in urban areas, and this trend was significant for white males. The annual rate of decrease in late-stage diagnosis and the onset years for significant declines varied greatly among counties and racial groups. Most counties with non-significant average annual percent change (AAPC) were located in the Florida Panhandle for white males, whereas they clustered in South-eastern Florida for black males. The new disparity statistic indicated that the spatial extent of racial disparities reached a peak in 1990 because of an early decline in frequency of late-stage diagnosis observed for black males.

Conclusions

Analyzing temporal trends in cancer incidence and mortality rates outside a spatial framework is unsatisfactory, since it leads one to overlook significant geographical variation which can potentially generate new insights about the impact of various interventions. Differences observed among nested geographies in Florida show how the modifiable areal unit problem (MAUP) also impacts the analysis of temporal changes.     

Inequalities in the use of mammography in Spain: effect of caring for disabled family

Objective

To examine the association between caring for disabled household members and the use of mammography in Spain during the 2?years before the interview.

Methods

Weighted cross-sectional data from the 2006 National Health Survey in Spain. Using logistic regression models we analyzed the association between the dependent variable and caring for disabled family members.

Results

After adjusting for predisposing, enabling and need determinants of the use of mammography screening, having primary caregiving responsibilities for disabled family members-dependent persons, was associated with a lower probability of having received a mammogram (OR 0.56; 95?% CI 0.32–0.98).

Conclusions

Although women in Spain have incorporated mammography into their preventive healthcare regimen, inequalities in its use persist in some groups of women such those with primary caregiving responsibilities for disabled family members.     

The approaches to measuring the potential spatial access to urban health services revisited: distance types and aggregation-error issues

Background

The potential spatial access to urban health services is an important issue in health geography, spatial epidemiology and public health. Computing geographical accessibility measures for residential areas (e.g. census tracts) depends on a type of distance, a method of aggregation, and a measure of accessibility. The aim of this paper is to compare discrepancies in results for the geographical accessibility of health services computed using six distance types (Euclidean and Manhattan distances; shortest network time on foot, by bicycle, by public transit, and by car), four aggregation methods, and fourteen accessibility measures.

Methods

To explore variations in results according to the six types of distance and the aggregation methods, correlation analyses are performed. To measure how the assessment of potential spatial access varies according to three parameters (type of distance, aggregation method, and accessibility measure), sensitivity analysis (SA) and uncertainty analysis (UA) are conducted.

Results

First, independently of the type of distance used except for shortest network time by public transit, the results are globally similar (correlation >0.90). However, important local variations in correlation between Cartesian and the four shortest network time distances are observed, notably in suburban areas where Cartesian distances are less precise. Second, the choice of the aggregation method is also important: compared with the most accurate aggregation method, accessibility measures computed from census tract centroids, though not inaccurate, yield important measurement errors for 10% of census tracts. Third, the SA results show that the evaluation of potential geographic access may vary a great deal depending on the accessibility measure and, to a lesser degree, the type of distance and aggregation method. Fourth, the UA results clearly indicate areas of strong uncertainty in suburban areas, whereas central neighbourhoods show lower levels of uncertainty.

Conclusion

In order to accurately assess potential geographic access to health services in urban areas, it is particularly important to choose a precise type of distance and aggregation method. Then, depending on the research objectives, the choices of the type of network distance (according to the mode of transportation) and of a number of accessibility measures should be carefully considered and adequately justified.

     

Urban sprawl,obesity, and cancer mortality in the United States: cross-sectional analysis and methodological challenges

Background

Urban sprawl has the potential to influence cancer mortality via direct and indirect effects on obesity, access to health services, physical activity, transportation choices and other correlates of sprawl and urbanization.

Methods

This paper presents a cross-sectional analysis of associations between urban sprawl and cancer mortality in urban and suburban counties of the United States. This ecological analysis was designed to examine whether urban sprawl is associated with total and obesity-related cancer mortality and to what extent these associations differed in different regions of the US. A major focus of our analyses was to adequately account for spatial heterogeneity in mortality. Therefore, we fit a series of regression models, stratified by gender, successively testing for the presence of spatial heterogeneity. Our resulting models included county level variables related to race, smoking, obesity, access to health services, insurance status, socioeconomic position, and broad geographic region as well as a measure of urban sprawl and several interactions. Our most complex models also included random effects to account for any county-level spatial autocorrelation that remained unexplained by these variables.

Results

Total cancer mortality rates were higher in less sprawling areas and contrary to our initial hypothesis; this was also true of obesity related cancers in six of seven U.S. regions (census divisions) where there were statistically significant associations between the sprawl index and mortality. We also found significant interactions (p?<?0.05) between region and urban sprawl for total and obesity related cancer mortality in both sexes. Thus, the association between urban sprawl and cancer mortality differs in different regions of the US.

Conclusions

Despite higher levels of obesity in more sprawling counties in the US, mortality from obesity related cancer was not greater in such counties. Identification of disparities in cancer mortality within and between geographic regions is an ongoing public health challenge and an opportunity for further analytical work identifying potential causes of these disparities. Future analyses of urban sprawl and health outcomes should consider exploring regional and international variation in associations between sprawl and health.     

Aboriginal Health Workers experience multilevel barriers to quitting smoking: a qualitative study

Background

Breast cancer screening rates have increased over time in the United States. However actual screening rates appear to be lower among black women compared with white women.

Purpose

To assess determinants of breast cancer screening among women in Michigan USA, focusing on individual and neighborhood socio-economic status and healthcare access.

Methods

Data from 1163 women ages 50-74 years who participated in the 2008 Michigan Special Cancer Behavioral Risk Factor Survey were analyzed. County-level SES and healthcare access were obtained from the Area Resource File. Multilevel logistic regression models were fit using SAS Proc Glimmix to account for clustering of individual observations by county. Separate models were fit for each of the two outcomes of interest; mammography screening and clinical breast examination. For each outcome, two sequential models were fit; a model including individual level covariates and a model including county level covariates.

Results

After adjusting for misclassification bias, overall cancer screening rates were lower than reported by survey respondents; black women had lower mammography screening rates but higher clinical breast examination rates than white women. However, after adjusting for other individual level variables, race was not a significant predictor of screening. Having health insurance or a usual healthcare provider were the most important predictors of cancer screening.

Discussion

Access to healthcare is important to ensuring appropriate cancer screening among women in Michigan.     

Residential Segregation and the Availability of Primary Care Physicians

Objective

To examine the association between residential segregation and geographic access to primary care physicians (PCPs) in metropolitan statistical areas (MSAs).

Data Sources

We combined zip code level data on primary care physicians from the 2006 American Medical Association master file with demographic, socioeconomic, and segregation measures from the 2000 U.S. Census. Our sample consisted of 15,465 zip codes located completely or partially in an MSA.

Methods

We defined PCP shortage areas as those zip codes with no PCP or a population to PCP ratio of >3,500. Using logistic regressions, we estimated the association between a zip code''s odds of being a PCP shortage area and its minority composition and degree of segregation in its MSA.

Principal Findings

We found that odds of being a PCP shortage area were 67 percent higher for majority African American zip codes but 27 percent lower for majority Hispanic zip codes. The association varied with the degree of segregation. As the degree of segregation increased, the odds of being a PCP shortage area increased for majority African American zip codes; however, the converse was true for majority Hispanic and Asian zip codes.

Conclusions

Efforts to address PCP shortages should target African American communities especially in segregated MSAs.     

Heterogeneity in mammography use across the nation: separating evidence of disparities from the disproportionate effects of geography

Background

Mammography is essential for early detection of breast cancer and both reduced morbidity and increased survival among breast cancer victims. Utilization is lower than national guidelines, and evidence of a recent decline in mammography use has sparked concern. We demonstrate that regression models estimated over pooled samples of heterogeneous states may provide misleading information regarding predictors of health care utilization and that comprehensive cancer control efforts should focus on understanding these differences and underlying causal factors. Our study population includes all women over age 64 with breast cancer in the Surveillance Epidemiology and End Results (SEER) cancer registries, linked to a nationally representative 5% reference sample of Medicare-eligible women located in 11 states that span all census regions and are heterogeneous in racial and ethnic mix. Combining women with and without cancer in the sample allows assessment of previous cancer diagnosis on propensity to use mammography. Our conceptual model recognizes the interplay between individual, social, cultural, and physical environments along the pathways to health care utilization, while delineating local and more distant levels of influence among contextual variables. In regression modeling, we assess individual-level effects, direct effects of contextual factors, and interaction effects between individual and contextual factors.

Results

Pooling all women across states leads to quite different conclusions than state-specific models. Commuter intensity, community acculturation, and community elderly impoverishment have significant direct impacts on mammography use which vary across states. Minorities living in isolated enclaves with others of the same race/ethnicity may be either advantaged or disadvantaged, depending upon the place studied.

Conclusion

Careful analysis of place-specific context is essential for understanding differences across communities stemming from different causal factors. Optimal policy interventions to change behavior (improve screening rates) will be as heterogeneous as local community characteristics, so no "one size fits all" policy can improve population health. Probability modeling with correction for clustering of individuals within multilevel contexts can reveal important differences from place to place and identify key factors to inform targeting of specific communities for further study.     

Evaluation of geoimputation strategies in a large case study

Background

Health data usually has missing or incomplete location information, which impacts the quality of research. Geoimputation methods are used by health professionals to increase the spatial resolution of address information for more accurate analyses. The objective of this study was to evaluate geo-imputation methods with respect to the demographic and spatial characteristics of the data.

Methods

We evaluated four geoimputation methods for increasing spatial resolution of records with known locational information at a coarse level. In order to test and rigorously evaluate two stochastic and two deterministic strategies, we used the Texas Sex Offender registry database with over 50,000 records with known demographic and coordinate information. We reduced the spatial resolution of each record to a census block group and attempted to recover coordinate information using the four strategies. We rigorously evaluated the results in terms of the error distance between the original coordinates and recovered coordinates by studying the results by demographic sub groups and the characteristics of the underlying geography.

Results

We observed that in estimating the actual location of a case, the weighted mean method is the most superior for each demographic group followed by the maximum imputation centroid, the random point in matching sub-geographies and the random point in all sub-geographies methods. Higher accuracies were observed for minority populations because minorities tend to cluster in certain neighborhoods, which makes it easier to impute their location. Results are greatly affected by the population density of the underlying geographies. We observed high accuracies in high population density areas, which often exist within smaller census blocks, which makes the search space smaller. Similarly, mapping geoimputation accuracies in a spatially explicit manner reveals that metropolitan areas yield higher accuracy results.

Conclusions

Based on gains in standard error, reduction in mean error and validation results, we conclude that characteristics of the estimated records such as the demographic profile and population density information provide a measure of certainty of geographic imputation.

     

Quantifying geocode location error using GIS methods

Background

The Metropolitan Atlanta Congenital Defects Program (MACDP) collects maternal address information at the time of delivery for infants and fetuses with birth defects. These addresses have been geocoded by two independent agencies: (1) the Georgia Division of Public Health Office of Health Information and Policy (OHIP) and (2) a commercial vendor. Geographic information system (GIS) methods were used to quantify uncertainty in the two sets of geocodes using orthoimagery and tax parcel datasets.

Methods

We sampled 599 infants and fetuses with birth defects delivered during 1994–2002 with maternal residence in either Fulton or Gwinnett County. Tax parcel datasets were obtained from the tax assessor's offices of Fulton and Gwinnett County. High-resolution orthoimagery for these counties was acquired from the U.S. Geological Survey. For each of the 599 addresses we attempted to locate the tax parcel corresponding to the maternal address. If the tax parcel was identified the distance and the angle between the geocode and the residence were calculated. We used simulated data to characterize the impact of geocode location error. In each county 5,000 geocodes were generated and assigned their corresponding Census 2000 tract. Each geocode was then displaced at a random angle by a random distance drawn from the distribution of observed geocode location errors. The census tract of the displaced geocode was determined. We repeated this process 5,000 times and report the percentage of geocodes that resolved into incorrect census tracts.

Results

Median location error was less than 100 meters for both OHIP and commercial vendor geocodes; the distribution of angles appeared uniform. Median location error was approximately 35% larger in Gwinnett (a suburban county) relative to Fulton (a county with urban and suburban areas). Location error occasionally caused the simulated geocodes to be displaced into incorrect census tracts; the median percentage of geocodes resolving into incorrect census tracts ranged between 4.5% and 5.3%, depending upon the county and geocoding agency.

Conclusion

Geocode location uncertainty can be estimated using tax parcel databases in a GIS. This approach is a viable alternative to global positioning system field validation of geocodes.     

Health conditions in regions of Eastern and Western Europe

Objectives

We report health conditions in the regions of the European Union (EU) from both objective and subjective perspectives and verify whether drawing conclusions regarding the health conditions in different countries without considering the countries’ sub-national variability may lead to inaccurate results.

Methods

To depict health conditions in the EU regions, two classification methods are used: (1) hierarchical clustering with Ward’s method and squared Euclidean distance and (2) k-mean clustering. To illustrate the intra-country variability of health conditions, the coefficients of variation are computed.

Results

Health conditions are considerably better in the western regions of the EU. Objective and subjective health measures do not always coincide. Extensive within-country differences in health conditions exist.

Conclusions

The EU is clearly split into the EU-15 and Central and Eastern European countries; however, this division is observed with respect to objective health conditions only. Inclusion of self-perceived health status in the analysis measures considerably changes this picture.     

Spread pattern of the first dengue epidemic in the city of Salvador, Brazil

Background

The explosive epidemics of dengue that have been occurring in various countries have stimulated investigation into new approaches to improve understanding of the problem and to develop new strategies for controlling the disease. The objective of this study was to evaluate the characteristics of diffusion of the first dengue epidemic that occurred in the city of Salvador in 1995.

Methods

The epidemiological charts and records of notified cases of dengue in Salvador in 1995 constituted the source of data. The cases of the disease were georeferenced according to census areas (spatial units) and epidemiological weeks (temporal unit). Kernel density estimation was used to identify the pattern of spatial diffusion using the R-Project computer software program.

Results

Of the 2,006 census areas in the city, 1,400 (70%) registered cases of dengue in 1995 and the spatial distribution of these records revealed that by the end of 1995 practically the entire city had been affected by the virus, with the largest concentration of cases occurring in the western region, composed of census areas with a high population density and predominantly horizontal residences compared to the eastern region of the city, where there is a predominance of vertical residential buildings.

Conclusion

The pattern found in this study shows the characteristics of the classic process of spreading by contagion that is common to most infectious diseases. It was possible to identify the epicenter of the epidemic from which centrifugal waves of the disease emanated. Our results suggest that, if a more agile control instrument existed that would be capable of rapidly reducing the vector population within a few days or of raising the group immunity of the population by means of a vaccine, it would theoretically be possible to adopt control actions around the epicenter of the epidemic and consequently reduce the incidence of the disease in the city. This finding emphasizes the need for further research to improve the technology available for the prevention of this disease.     

Care-seeking behavior of Japanese gynecological cancer survivors suffering from adverse effects

Background

Post-treatment follow-up visits for gynecological cancer survivors should provide opportunities for management of adverse physical/psychological effects of therapy and early recurrence detection. However, the adequacy of such visits in Japan is poorly documented. We qualitatively explored care-seeking experiences of Japanese gynecological cancer survivors and deduced factors influencing care-seeking behaviors and treatment access.

Methods

We conducted 4 semi-structured focus groups comprising altogether 28 Japanese gynecological cancer survivors to collect a variety of participants’ post-treatment care-seeking behaviors through active interaction with participants. Factors influencing access to treatment for adverse effects were analyzed qualitatively.

Results

Survivors sought care through specialty clinic visits when regular post-treatment gynecological follow-ups were inadequate or when symptoms seemed to be non-treatment related. Information provided by hospital staff during initial treatment influenced patients’ understanding and response to adverse effects. Lack of knowledge and inaccurate symptom interpretation delayed help-seeking, exacerbating symptoms. Gynecologists’ attitudes during follow-ups frequently led survivors to cope with symptoms on their own. Information from mass media, Internet, and support groups helped patients understand symptoms and facilitated care seeking.

Conclusions

Post-treatment adverse effects are often untreated during follow-up visits. Awareness of possible post-treatment adverse effects is important for gynecological cancer survivors in order to obtain appropriate care if the need arises. Consultation during the follow-up visit is essential for continuity in care.     

Relative risk estimation of dengue disease at small spatial scale

Background

Dengue is a high incidence arboviral disease in tropical countries around the world. Colombia is an endemic country due to the favourable environmental conditions for vector survival and spread. Dengue surveillance in Colombia is based in passive notification of cases, supporting monitoring, prediction, risk factor identification and intervention measures. Even though the surveillance network works adequately, disease mapping techniques currently developed and employed for many health problems are not widely applied. We select the Colombian city of Bucaramanga to apply Bayesian areal disease mapping models, testing the challenges and difficulties of the approach.

Methods

We estimated the relative risk of dengue disease by census section (a geographical unit composed approximately by 1–20 city blocks) for the period January 2008 to December 2015. We included the covariates normalized difference vegetation index (NDVI) and land surface temperature (LST), obtained by satellite images. We fitted Bayesian areal models at the complete period and annual aggregation time scales for 2008–2015, with fixed and space-varying coefficients for the covariates, using Markov Chain Monte Carlo simulations. In addition, we used Cohen’s Kappa agreement measures to compare the risk from year to year, and from every year to the complete period aggregation.

Results

We found the NDVI providing more information than LST for estimating relative risk of dengue, although their effects were small. NDVI was directly associated to high relative risk of dengue. Risk maps of dengue were produced from the estimates obtained by the modeling process. The year to year risk agreement by census section was sligth to fair.

Conclusion

The study provides an example of implementation of relative risk estimation using Bayesian models for disease mapping at small spatial scale with covariates. We relate satellite data to dengue disease, using an areal data approach, which is not commonly found in the literature. The main difficulty of the study was to find quality data for generating expected values as input for the models. We remark the importance of creating population registry at small spatial scale, which is not only relevant for the risk estimation of dengue but also important to the surveillance of all notifiable diseases.

     

Validation of the Arabic version of the EORTC quality of life questionnaire among cancer patients in Lebanon

Purpose

The objective of this article was to evaluate the psychometric properties of the translated Lebanese Arabic version of the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-C30 in a sample of adult cancer patients in Lebanon.

Methods

The EORTC QLQ-C30 was translated into Lebanese Arabic and administered to a convenient sample of 200 adult patients diagnosed with cancer at a tertiary care center in Lebanon between 2009 and 2010. The psychometric indices assessed were reliability, multitrait scaling analysis, construct validity, and confirmatory factor analysis.

Results

Six out of nine subscales had Cronbach’s alpha coefficients above 0.70. Multitrait scaling analysis showed that all item-scale correlation coefficients met the set standards of convergent validity with the exception of item 5 only (95.8 %). In addition, 79.7 % of the item-scale correlation coefficients met the criterion for discriminant validity. In inter-scale correlations, all conceptually related scales had correlation coefficients of ≥0.40 with the exception of role functioning and fatigue scales having an undesirable correlation coefficient of ?0.76. In known-groups comparison, the instrument differentiated significantly between some of the subscales with respect to education, employment, and age. CFA showed an almost good fit (GFI = 0.87) with respect to our current data set.

Conclusion

The translated Lebanese Arabic version of the EORTC QLQ-C30 is a reliable and valid instrument that can be used to assess the quality of life of Lebanese cancer patients.     

Belief Elicitation to Populate Health Economic Models of Medical Diagnostic Devices in Development

Background and Objective

Bayesian methods can be used to elicit experts’ beliefs about the clinical value of healthcare technologies. This study investigates a belief–elicitation method for estimating diagnostic performance in an early stage of development of photoacoustic mammography (PAM) imaging versus magnetic resonance imaging (MRI) for detecting breast cancer.

Research Design

Eighteen experienced radiologists ranked tumor characteristics regarding their importance to detect malignancies. With reference to MRI, radiologists estimated the true positives and negatives of PAM using the variable interval method. An overall probability density function was determined using linear opinion pooling, weighted for individual experts’ experience.

Result

The most important tumor characteristics are mass margins and mass shape. Respondents considered MRI the better technology to visualize these characteristics. Belief elicitation confirmed this by providing an overall sensitivity of PAM ranging from 58.9 to 85.1 % (mode 75.6 %) and specificity ranging from 52.2 to 77.6 % (mode 66.5 %).

Conclusion

Belief elicitation allowed estimates to be obtained for the expected diagnostic performance of PAM, although radiologists expressed difficulties in doing so. Heterogeneity within and between experts reflects this uncertainty and the infancy of PAM. Further clinical trials are required to validate the extent to which this belief–elicitation method is predictive for observed test performance.     

The Impact of Rural Health Care Accessibility on Cancer-Related Behaviors and Outcomes

Background

This research seeks to identify the relationship between economic factors related to the ability to receive and pay for health services and adverse cancer outcomes, as well as preventative screening and behavioral factors that influence the risk of cancer. We focus on the Northern High Plains region, where we are able to compare regions with extremely low access to health services with those with relatively high levels of access.

Objective

This study aims to identify health disparities in rural communities, particularly among Native American populations, and, thereby, begin to determine the most effective means by which to deliver health services to areas where geography, economics, and culture might prevent traditional models of health delivery from providing sufficient incentives for the prevention of adverse cancer-related outcomes.

Methods

The Health Care Accessibility Index (HCAI) is computed through the use of principal component analysis and includes economic variables as well as variables concerning institutional and geographic access to health care. Index values are then regressed onto cancer outcomes, cancer-prevention outcomes, and cancer-related risk, using weighted least squares and quantile regressions.

Results

Counties with relatively poor access to health care (low HCAI) also have statistically (1) lower breast cancer screening rates, (2) higher smoking prevalence, (3) higher obesity prevalence, and (4) higher cancer-related mortality rates. Breast cancer screening is found to be especially sensitive to areas of low health accessibility.

Conclusions

Empirical results provide support for policy efforts to increase the accessibility of health care services that are targeted to areas with low mammography screening rates, high obesity rates, high smoking prevalence, as well as areas near Native American reservation territories.     

Breast Cancer Screening,Area Deprivation,and Later-Stage Breast Cancer in Appalachia: Does Geography Matter?

Objective

To model the relationship of an area-based measure of a breast cancer screening and geographic area deprivation on the incidence of later stage breast cancer (LSBC) across a diverse region of Appalachia.

Data Source

Central cancer registry data (2006–2008) from three Appalachian states were linked to Medicare claims and census data.

Study Design

Exploratory spatial analysis preceded the statistical model based on negative binomial regression to model predictors and effect modification by geographic subregions.

Principal Findings

Exploratory spatial analysis revealed geographically varying effects of area deprivation and screening on LSBC. In the negative binomial regression model, predictors of LSBC included receipt of screening, area deprivation, supply of mammography centers, and female population aged >75 years. The most deprived counties had a 3.31 times greater rate of LSBC compared to the least deprived. Effect of screening on LSBC was significantly stronger in northern Appalachia than elsewhere in the study region, found mostly for high-population counties.

Conclusions

Breast cancer screening and area deprivation are strongly associated with disparity in LBSC in Appalachia. The presence of geographically varying predictors of later stage tumors in Appalachia suggests the importance of place-based health care access and risk.     

Accuracy of commercial geocoding: assessment and implications

Background

Published studies of geocoding accuracy often focus on a single geographic area, address source or vendor, do not adjust accuracy measures for address characteristics, and do not examine effects of inaccuracy on exposure measures. We addressed these issues in a Women's Health Initiative ancillary study, the Environmental Epidemiology of Arrhythmogenesis in WHI.

Results

Addresses in 49 U.S. states (n = 3,615) with established coordinates were geocoded by four vendors (A-D). There were important differences among vendors in address match rate (98%; 82%; 81%; 30%), concordance between established and vendor-assigned census tracts (85%; 88%; 87%; 98%) and distance between established and vendor-assigned coordinates (mean ρ [meters]: 1809; 748; 704; 228). Mean ρ was lowest among street-matched, complete, zip-coded, unedited and urban addresses, and addresses with North American Datum of 1983 or World Geodetic System of 1984 coordinates. In mixed models restricted to vendors with minimally acceptable match rates (A-C) and adjusted for address characteristics, within-address correlation, and among-vendor heteroscedasticity of ρ, differences in mean ρ were small for street-type matches (280; 268; 275), i.e. likely to bias results relying on them about equally for most applications. In contrast, differences between centroid-type matches were substantial in some vendor contrasts, but not others (5497; 4303; 4210) p_interaction < 10^-4, i.e. more likely to bias results differently in many applications. The adjusted odds of an address match was higher for vendor A versus C (odds ratio = 66, 95% confidence interval: 47, 93), but not B versus C (OR = 1.1, 95% CI: 0.9, 1.3). That of census tract concordance was no higher for vendor A versus C (OR = 1.0, 95% CI: 0.9, 1.2) or B versus C (OR = 1.1, 95% CI: 0.9, 1.3). Misclassification of a related exposure measure – distance to the nearest highway – increased with mean ρ and in the absence of confounding, non-differential misclassification of this distance biased its hypothetical association with coronary heart disease mortality toward the null.

Conclusion

Geocoding error depends on measures used to evaluate it, address characteristics and vendor. Vendor selection presents a trade-off between potential for missing data and error in estimating spatially defined attributes. Informed selection is needed to control the trade-off and adjust analyses for its effects.     

The relevance of spatial aggregation level and of applied methods in the analysis of geographical distribution of cancer mortality in mainland Portugal (2009–2013)

Background

Knowledge regarding the geographical distribution of diseases is essential in public health in order to define strategies to improve the health of populations and quality of life.The present study aims to establish a methodology to choose a suitable geographic aggregation level of data and an appropriated method which allow us to analyze disease spatial patterns in mainland Portugal, avoiding the “small numbers problem.” Malignant cancer mortality data for 2009–2013 was used as a case study.

Methods

To achieve our aims, we used official data regarding the mortality by all malignant cancer, between 2009 and 2013, and the mainland Portuguese resident population in 2011. Three different spatial aggregation levels were applied: Nomenclature of Territorial Units for Statistics, level III (28 areas), municipalities (278 areas), and parishes (4050 areas).Standardized Mortality Ratio (SMR) and relative risk (RR) were computed with Besag, York and Mollié model (BYM) for the evaluation of geographic patterns of mortality data. We also estimated Global Moran’s I, Local Moran’s I, and posterior probability (PP) for the spatial cluster analysis.

Results

Our results show that the occurrence of lower and higher extreme values of the standardized mortality ratio tend to increase with the decrease of data spatial aggregation. In addition, the number of local clusters is higher at small spatial aggregation levels, although the area of each cluster is generally smaller. Regarding global clustering, data forms clusters at all considered levels.Relative risk (RR) computed by Besag, York and Mollié model, in turn, also shows different results at the municipalities and parishes levels. However, the difference is smaller than the difference obtained by SMR computation. This statement is supported by the coefficient variation values.

Conclusions

Our findings show that the choice of spatial data aggregation level has high importance in the research results, as different aggregation levels can lead to distinct results.In terms of the case study, we conclude that for the period of 2009–2013, cancer mortality in mainland Portugal formed clusters. The most suitable applicable spatial scale and method seemed to be at the municipalities level and Besag, York and Mollié model, respectively. However, further studies should be conducted in order to provide greater support to these results.