Geocoding addresses from a large population-based study: lessons learned

Agricultural paraquat dichloride use and Parkinson's disease in California's Central Valley

BACKGROUND

Paraquat dichloride is currently among the most widely used commercial herbicides in the USA. In the present study, we provide epidemiological assessment of ambient paraquat exposure and Parkinson's disease (PD) risk in a population-based study of PD in agricultural regions of Central California.

METHODS

Based on 829 PD patients and 824 community controls, we assessed associations between ambient paraquat dichloride exposure and PD. We estimated residential and workplace proximity to commercial agricultural applications in three California counties since 1974 using the CA pesticide use reporting (PUR) data and land use maps. We evaluated any, duration and average intensity [pounds (0.45 kilograms) per acre per year] of exposure for paraquat in four time windows.

RESULTS

Ambient paraquat exposure assessed at both residence and workplace was associated with PD, based on several different exposure measures. The PD patients both lived and worked near agricultural facilities applying greater amounts of the herbicide than community controls. For workplace proximity to commercial applications since 1974, working near paraquat applications every year in the window [odds ratio (OR) = 2.15, 95% confidence interval (CI) = 1.46, 3.19] and a higher average intensity of exposure [per 10 pounds (4.54 kilograms), OR = 2.08, 95% CI = 1.31, 3.38] were both associated with an increased odds of PD. Similar associations were observed for residential proximity (duration: OR = 1.91, 95% CI = 1.30, 2.83; average intensity: OR = 1.72, 95% CI = 0.99, 3.04). Risk estimates were comparable for men and women, and the strongest odds were observed for those diagnosed at ≤60 years of age.

CONCLUSION

This study provides further indication that paraquat dichloride exposure increases the risk of Parkinson's disease.

A pesticide and iPSC dopaminergic neuron screen identifies and classifies Parkinson-relevant pesticides

Parkinson's disease (PD) is a complex neurodegenerative disease with etiology rooted in genetic vulnerability and environmental factors. Here we combine quantitative epidemiologic study of pesticide exposures and PD with toxicity screening in dopaminergic neurons derived from PD patient induced pluripotent stem cells (iPSCs) to identify Parkinson's-relevant pesticides. Agricultural records enable investigation of 288 specific pesticides and PD risk in a comprehensive, pesticide-wide association study. We associate long-term exposure to 53 pesticides with PD and identify co-exposure profiles. We then employ a live-cell imaging screening paradigm exposing dopaminergic neurons to 39 PD-associated pesticides. We find that 10 pesticides are directly toxic to these neurons. Further, we analyze pesticides typically used in combinations in cotton farming, demonstrating that co-exposures result in greater toxicity than any single pesticide. We find trifluralin is a driver of toxicity to dopaminergic neurons and leads to mitochondrial dysfunction. Our paradigm may prove useful to mechanistically dissect pesticide exposures implicated in PD risk and guide agricultural policy.

Thyroid Cancer and Pesticide Use in a Central California Agricultural Area: A Case Control Study

OBJECTIVE

To examine environmental factors that influence risk of thyroid cancer.

METHODS

We performed a case-control study utilizing thyroid cancer cases from the California Cancer Registry (1999-2012) and controls sampled in a population-based manner. Study participants were included if they were diagnosed with thyroid cancer, lived in the study area at their time of diagnosis, and were ≥35 years of age. Controls were recruited from the same area and eligible to participate if they were ≥35 years of age and had been living in California for at least 5 years prior to the interview. We examined residential exposure to 29 agricultural use pesticides, known to cause DNA damage in vitro or are known endocrine disruptors. We employed a validated geographic information system-based system to generate exposure estimates for each participant.

RESULTS

Our sample included 2067 cases and 1003 controls. In single pollutant models and within a 20-year exposure period, 10 out of 29 selected pesticides were associated with thyroid cancer, including several of the most applied pesticides in the United States such as paraquat dichloride [odds ratio (OR): 1.46 (95% CI: 1.23, 1.73)], glyphosate [OR: 1.33 (95% CI: 1.12, 1.58)], and oxyfluorfen [OR: 1.21 (95% CI: 1.02, 1.43)]. Risk of thyroid cancer increased proportionately to the total number of pesticides subjects were exposed to 20 years before diagnosis or interview. In all models, paraquat dichloride was associated with thyroid cancer.

CONCLUSIONS

Our study provides first evidence in support of the hypothesis that residential pesticide exposure from agricultural applications is associated with an increased risk of thyroid cancer.

A Flexible Method for Identifying Spatial Clusters of Breast Cancer Using Individual-Level Data

Prior research has shown that cancer risk varies by geography, but scan statistics methods for identifying cancer clusters in case-control studies have been limited in their ability to identify multiple clusters and adjust for participant-level risk factors. We develop a method to identify geographic patterns of breast cancer odds using the Wisconsin Women's Health Study, a series of 5 population-based case-control studies of female Wisconsin residents aged 20-79 enrolled in 1988-2004 (cases=16,076, controls=16,795). We create sets of potential clusters by overlaying a 1 km grid over each county-neighborhood and enumerating a series of overlapping circles. Using a two-step approach, we fit a penalized binomial regression model to the number of cases and trials in each grid cell, penalizing all potential clusters by the least absolute shrinkage and selection operator (Lasso). We use BIC to select the number of clusters, which are included in a participant-level logistic regression model. We identify 15 geographic clusters, resulting in 23 areas of unique geographic odds ratios. After adjustment for known risk factors, confidence intervals narrowed but breast cancer odds ratios did not meaningfully change; one additional hotspot was identified. By considering multiple overlapping spatial clusters simultaneously, we discern gradients of spatial odds across Wisconsin.

Measuring Spatial Accessibility to Hospitals of Acute Myocardial Infarction in Multi Period Scale: A Case Study in Shijingshan District, Beijing, China

The hospital accessibility of Acute Myocardial Infarction (AMI) emergency treatment is of great importance, not only for improving the survival rate of patients but also for protecting the basic human right to health care. Traditional AMI emergency treatment research often does not consider ways to shorten the travel time to hospitals for AMI patients and does not reflect the actual time it takes to travel to hospitals, which is critical to AMI emergency treatment. To avoid these shortcomings, this study proposes a method of accessibility measurement based on Web Mapping API (Application Programming Interface) to obtain travel time to hospitals during different periods, then calculated the AMI hospital accessibility based on these detailed data. This study considered the Shijingshan District, Beijing, China, as an empirical case. The study discovered significant differences in the temporal and spatial characteristics of the AMI hospital accessibility on weekdays and weekends. The analysis revealed that travel time to hospitals and traffic congestion are the two main factors affecting AMI hospital accessibility. The research results shed new light on the accessibility of urban medical facilities and provide a scientific basis with which local governments can optimize the spatial structure of medical facilities.

Improving a Street-Based Geocoding Algorithm Using Machine Learning Techniques

Address matching is a crucial step in geocoding; however, this step forms a bottleneck for geocoding accuracy, as precise input is the biggest challenge for establishing perfect matches. Matches still have to be established despite the inevitability of incorrect address inputs such as misspellings, abbreviations, informal and non-standard names, slangs, or coded terms. Thus, this study suggests an address geocoding system using machine learning to enhance the address matching implemented on street-based addresses. Three different kinds of machine learning methods are tested to find the best method showing the highest accuracy. The performance of address matching using machine learning models is compared to multiple text similarity metrics, which are generally used for the word matching. It was proved that extreme gradient boosting with the optimal hyper-parameters was the best machine learning method with the highest accuracy in the address matching process, and the accuracy of extreme gradient boosting outperformed similarity metrics when using training data or input data. The address matching process using machine learning achieved high accuracy and can be applied to any geocoding systems to precisely convert addresses into geographic coordinates for various research and applications, including car navigation.

A network-based method with privacy-preserving for identifying influential providers in large healthcare service systems

In data science, networks provide a useful abstraction of the structure of many complex systems, ranging from social systems and computer networks to biological networks and physical systems. Healthcare service systems are one of the main social systems that can also be understood using network-based approaches, for example, to identify and evaluate influential providers. In this paper, we propose a network-based method with privacy-preserving for identifying influential providers in large healthcare service systems. First, the provider-interacting network is constructed by employing publicly available information on locations and types of healthcare services of providers. Second, the ranking of nodes in the generated provider-interacting network is conducted in parallel on the basis of four nodal influence metrics. Third, the impact of the top-ranked influential nodes in the provider-interacting network is evaluated using three indicators. Compared with other research work based on patient-sharing networks, in this paper, the provider-interacting network of healthcare service providers can be roughly created according to the locations and the publicly available types of healthcare services, without the need for personally private electronic medical claims, thus protecting the privacy of patients. The proposed method is demonstrated by employing Physician and Other Supplier Data CY 2017, and can be applied to other similar datasets to help make decisions for the optimization of healthcare resources in the response to public health emergencies.

Unconventional natural gas development and adverse birth outcomes in Pennsylvania: The potential mediating role of antenatal anxiety and depression

BACKGROUND

Studies have reported associations between unconventional natural gas development (UNGD) and adverse birth outcomes. None have evaluated potential mediating mechanisms.

OBJECTIVES

To evaluate associations between (1) UNGD and antenatal anxiety and depression and (2) antenatal anxiety and depression and preterm birth (<37 weeks gestation) and reduced term birth weight, (3) stochastic direct and indirect effects of UNGD on preterm birth and term birth weight operating through antenatal anxiety and depression, and (4) effect modification by family-level socioeconomic status.

METHODS

This retrospective cohort study included mothers without prevalent anxiety or depression at time of conception, who delivered at Geisinger in Pennsylvania between January 2009-January 2013. We assembled phase-specific UNGD activity data from public sources. Mothers were categorized as exposed (quartile 4) or unexposed (quartiles 1-3) based on average daily inverse distance-squared UNGD activity metric between conception and the week prior to anxiety or depression (cases) or the pregnancy-average daily metric (non-cases). We estimated associations with a doubly robust estimator (targeted minimum loss-based estimation) and adjusted for potential individual- and community-level confounding variables.

RESULTS

Analyses included 8,371 births to 7,715 mothers, 12.2% of whom had antenatal anxiety or depression. We found 4.3 additional cases of antenatal anxiety or depression per 100 women (95% CI: 1.5, 7.0) under the scenario where all mothers lived in the highest quartile of UNGD activity versus quartiles 1-3. The risk difference appeared larger among mothers receiving Medical Assistance (indicator of low family income) compared to those who did not, 5.6 (95% CI: 0.5, 10.6) versus 2.9 (95% CI: -0.7, 6.5) additional cases of antenatal anxiety or depression per 100 women. We found no relationship between antenatal anxiety or depression and adverse birth outcomes and no mediation effect either overall or when stratifying by Medical Assistance.

CONCLUSION

We observed a relationship between UNGD activity and antenatal anxiety and depression, which did not mediate the overall association between UNGD activity and adverse birth outcomes.

Geographic access to mammography facilities and frequency of mammography screening

PURPOSE

To assess the association between geographic access to mammography facilities and women's mammography utilization frequency.

METHODS

Using data from the population-based 1995-2007 Wisconsin Women's Health study, we used proportional odds and logistic regression to test whether driving times to mammography facilities and the number of mammography facilities within 10 km of women's homes were associated with mammography frequency among women aged 50-74 years and whether associations differed between Rural-Urban Commuting Areas and income and education groups.

RESULTS

We found evidence for nonlinear relationships between geographic access and mammography utilization (nonlinear effects of driving times and facility density, P-values .01 and .005, respectively). Having at least one nearby mammography facility was associated with greater mammography frequency among urban women (1 vs. 0 facilities, odds ratio 1.26, 95% confidence interval, 1.09-1.47), with similar effects among rural women. Adding more facilities had decreasing marginal effects. Long driving times tended to be associated with lower mammography frequency. We found no effect modification by income, education, or urbanicity. In rural settings, mammography nonuse was higher, facility density smaller, and driving times to facilities were longer.

CONCLUSIONS

Having at least one mammography facility near one's home may increase mammography utilization, with decreasing effects per each additional facility.

The availability of medical marijuana dispensary and adolescent marijuana use

PURPOSE

To examine the association between medical marijuana dispensary (MMD) availability and adolescent marijuana use.

METHODS

The study sample was comprised of 8th, 10th, and 12th graders (N=14,953) from 141 schools in the 2014 Monitoring the Future study, who resided in the 18 states that had legalized medical marijuana as of January 1, 2014. Multilevel logistic regressions with random effects were conducted to quantify the cross-sectional associations of the availability of MMD within 5- and 25-mile buffers from the centroid of school zip codes with self-reported recent use (past-year) and current use (past-month) of marijuana, controlling for individual characteristics and school, zip code, and state contextual factors.

RESULTS

In the combined sample, the availability of MMD was not associated with recent or current use of marijuana. Subsample analyses suggested that the availability within a 5-mile buffer was associated with a higher likelihood of recent use in 8th graders (OR=1.93, 95% CI=1.11-3.33) and the availability within a 5- to 25-mile buffer was associated with a higher likelihood of recent use in 10th graders (OR=1.33, 95% CI=1.00-1.77). The availability of MMD was not associated with recent use in 12th graders or current use in any grades.

CONCLUSIONS

The availability of MMD was not associated with current use of marijuana among adolescents. There was some evidence suggesting that the availability of MMD within short to medium traveling distance may be associated with a higher level of recent use in middle schoolers who are also at a high risk of experimenting with marijuana.

Spatial measurement errors in the field of spatial epidemiology

Background

Spatial epidemiology has been aided by advances in geographic information systems, remote sensing, global positioning systems and the development of new statistical methodologies specifically designed for such data. Given the growing popularity of these studies, we sought to review and analyze the types of spatial measurement errors commonly encountered during spatial epidemiological analysis of spatial data.

Methods

Google Scholar, Medline, and Scopus databases were searched using a broad set of terms for papers indexed by a term indicating location (space or geography or location or position) and measurement error (measurement error or measurement inaccuracy or misclassification or uncertainty): we reviewed all papers appearing before December 20, 2014. These papers and their citations were reviewed to identify the relevance to our review.

Results

We were able to define and classify spatial measurement errors into four groups: (1) pure spatial location measurement errors, including both non-instrumental errors (multiple addresses, geocoding errors, outcome aggregations, and covariate aggregation) and instrumental errors; (2) location-based outcome measurement error (purely outcome measurement errors and missing outcome measurements); (3) location-based covariate measurement errors (address proxies); and (4) Covariate-Outcome spatial misaligned measurement errors. We propose how these four classes of errors can be unified within an integrated theoretical model and possible solutions were discussed.

Conclusion

Spatial measurement errors are ubiquitous threat to the validity of spatial epidemiological studies. We propose a systematic framework for understanding the various mechanisms which generate spatial measurement errors and present practical examples of such errors.

Tools for address georeferencing - limitations and opportunities every public health professional should be aware of

Various address georeferencing (AG) tools are currently available. But little is known about the quality of each tool. Using data from the EPIPorto cohort we compared the most commonly used AG tools in terms of positional error (PE) and subjects' misclassification according to census tract socioeconomic status (SES), a widely used variable in epidemiologic studies. Participants of the EPIPorto cohort (n = 2427) were georeferenced using Geographical Information Systems (GIS) and Google Earth (GE). One hundred were randomly selected and georeferenced using three additional tools: 1) cadastral maps (gold-standard); 2) Global Positioning Systems (GPS) and 3) Google Earth, single and in a batch. Mean PE and the proportion of misclassified individuals were compared. Google Earth showed lower PE than GIS, but 10% of the addresses were imprecisely positioned. Thirty-eight, 27, 16 and 14% of the participants were located in the wrong census tract by GIS, GPS, GE (batch) and GE (single), respectively (p<0.001). Misclassification according to SES was less frequent but still non-negligible −14.4, 8.1, 4.2 and 2% (p<0.001). The quality of georeferencing differed substantially between AG tools. GE seems to be the best tool, but only if prudently used. Epidemiologic studies using spatial data should start including information on the quality and accuracy of their georeferencing tools and spatial datasets.

Geocoding large population-level administrative datasets at highly resolved spatial scales

Using geographic information systems to link administrative databases with demographic, social, and environmental data allows researchers to use spatial approaches to explore relationships between exposures and health. Traditionally, spatial analysis in public health has focused on the county, zip code, or tract level because of limitations to geocoding at highly resolved scales. Using 2005 birth and death data from North Carolina, we examine our ability to geocode population-level datasets at three spatial resolutions - zip code, street, and parcel. We achieve high geocoding rates at all three resolutions, with statewide street geocoding rates of 88.0% for births and 93.2% for deaths. We observe differences in geocoding rates across demographics and health outcomes, with lower geocoding rates in disadvantaged populations and the most dramatic differences occurring across the urban-rural spectrum. Our results suggest highly resolved spatial data architectures for population-level datasets are viable through geocoding individual street addresses. We recommend routinely geocoding administrative datasets to the highest spatial resolution feasible, allowing public health researchers to choose the spatial resolution used in analysis based on an understanding of the spatial dimensions of the health outcomes and exposures being investigated. Such research, however, must acknowledge how disparate geocoding success across subpopulations may affect findings.

What factors influence minority use of National Cancer Institute-designated cancer centers?

BACKGROUND

National Cancer Institute (NCI) cancer centers provide high-quality care and are associated with better outcomes. However, racial and ethnic minority populations tend not to use these settings. The current study sought to understand what factors influence minority use of NCI cancer centers.

METHODS

A data set containing California Cancer Registry (CCR) data linked to patient discharge abstracts identified all patients with colorectal cancer (CRC) who were treated from 1996 through 2006. Multivariable models were generated to predict the use of NCI settings by race. Geographic proximity to an NCI center and patient sociodemographic and clinical characteristics were assessed.

RESULTS

Approximately 5% of all identified patients with CRC (n = 79,231) were treated in NCI settings. The median travel distance for treatment for all patients in all hospitals was ≤ 5 miles. A higher percentage of minorities lived near an NCI cancer center compared with whites. A baseline multivariable model predicting use showed a negative association between Hispanic ethnicity and NCI center use (odds ratio, 0.71; 95% confidence interval, 0.64-0.79). Asian/Pacific Islander patients were more likely to use NCI centers (odds ratio, 1.41; 95% confidence interval, 1.28-1.54). There was no difference in use noted among black patients. Increasing living distance from an NCI cancer center was found to be predictive of lower odds of use for all populations. Medicare and Medicaid insurance statuses were positively associated with NCI center use. Neighborhood-level education was found to be a more powerful predictor of NCI use than poverty or unemployment.

CONCLUSIONS

Select minority groups underuse NCI cancer centers for CRC treatment. Sociodemographic factors and proximity to NCI centers are important predictors of use. Interventions to address these factors may improve minority attendance to NCI cancer centers for care.

The association between ambient exposure to organophosphates and Parkinson's disease risk

OBJECTIVES

There is a general consensus that pesticides are involved in the aetiology of Parkinson's disease (PD), although associations between specific pesticides and the risk of developing PD have not been well studied. This study examines the risk of developing PD associated with specific organophosphate (OP) pesticides and their mechanisms of toxicity.

METHODS

This case-control study uses a geographic information system-based exposure assessment tool to estimate ambient exposure to 36 commonly used OPs from 1974 to 1999. All selected OPs were analysed individually and also in groups formed according to their presumed mechanisms of toxicity.

RESULTS

The study included 357 incident PD cases and 752 population controls living in the Central Valley of California. Ambient exposure to each OP evaluated separately increased the risk of developing PD. However, most participants were exposed to combinations of OPs rather than a single pesticide. Risk estimates for OPs grouped according to different presumed functionalities and toxicities were similar and did not allow us to distinguish between them. However, we observed exposure-response patterns with exposure to an increasing number of OPs.

CONCLUSIONS

This study adds strong evidence that OPs are implicated in the aetiology of idiopathic PD. However, studies of OPs at low doses reflective of real-world ambient exposure are needed to determine the mechanisms of neurotoxicity.

An evaluation framework for comparing geocoding systems

BACKGROUND

Geocoding, the process of converting textual information describing a location into one or more digital geographic representations, is a routine task performed at large organizations and government agencies across the globe. In a health context, this task is often a fundamental first step performed prior to all operations that take place in a spatially-based health study. As such, the quality of the geocoding system used within these agencies is of paramount concern to the agency (the producer) and researchers or policy-makers who wish to use these data (consumers). However, geocoding systems are continually evolving with new products coming on the market continuously. Agencies must develop and use criteria across a number axes when faced with decisions about building, buying, or maintaining any particular geocoding systems. To date, published criteria have focused on one or more aspects of geocode quality without taking a holistic view of a geocoding system's role within a large organization. The primary purpose of this study is to develop and test an evaluation framework to assist a large organization in determining which geocoding systems will meet its operational needs.

METHODS

A geocoding platform evaluation framework is derived through an examination of prior literature on geocoding accuracy. The framework developed extends commonly used geocoding metrics to take into account the specific concerns of large organizations for which geocoding is a fundamental operational capability tightly-knit into its core mission of processing health data records. A case study is performed to evaluate the strengths and weaknesses of five geocoding platforms currently available in the Australian geospatial marketplace.

RESULTS

The evaluation framework developed in this research is proven successful in differentiating between key capabilities of geocoding systems that are important in the context of a large organization with significant investments in geocoding resources. Results from the proposed methodology highlight important differences across all axes of geocoding system comparisons including spatial data output accuracy, reference data coverage, system flexibility, the potential for tight integration, and the need for specialized staff and/or development time and funding. Such results can empower decisions-makers within large organizations as they make decisions and investments in geocoding systems.

Components of air pollution and cognitive function in middle-aged and older adults in Los Angeles

While experiments in animals demonstrate neurotoxic effects of particulate matter (PM) and ozone (O3), epidemiologic evidence is sparse regarding the relationship between different constituencies of air pollution mixtures and cognitive function in adults. We examined cross-sectional associations between various ambient air pollutants [O3, PM2.5 and nitrogen dioxide (NO2)] and six measures of cognitive function and global cognition among healthy, cognitively intact individuals (n=1496, mean age 60.5 years) residing in the Los Angeles Basin. Air pollution exposures were assigned to each residential address in 2000-06 using a geographic information system that included monitoring data. A neuropsychological battery was used to assess cognitive function; a principal components analysis defined six domain-specific functions and a measure of global cognitive function was created. Regression models estimated effects of air pollutants on cognitive function, adjusting for age, gender, race, education, income, study and mood. Increasing exposure to PM2.5 was associated with lower verbal learning (β=-0.32 per 10 μg/m(3) PM2.5, 95% CI=-0.63, 0.00; p=0.05). Ambient exposure to NO2 >20 ppb tended to be associated with lower logical memory. Compared to the lowest level of exposure to ambient O3, exposure above 49 ppb was associated with lower executive function. Including carotid artery intima-media thickness, a measure of subclinical atherosclerosis, in models as a possible mediator did not attenuate effect estimates. This study provides support for cross-sectional associations between increasing levels of ambient O3, PM2.5 and NO2 and measures of domain-specific cognitive abilities.

Modeling travel impedance to medical care for children with birth defects using Geographic Information Systems

BACKGROUND

Children with birth defects may face significant geographic barriers accessing medical care and specialized services. Using a Geographic Information Systems-based approach, one-way travel time and distance to access medical care for children born with spina bifida was estimated.

METHODS

Using 2007 road information from the Florida Department of Transportation, we built a topological network of Florida roads. Live-born Florida infants with spina bifida during 1998 to 2007 were identified by the Florida Birth Defects Registry and linked to hospital discharge records. Maternal residence at delivery and hospitalization locations were identified during the first year of life.

RESULTS

Of 668 infants with spina bifida, 8.1% (n = 54) could not be linked to inpatient data, resulting in 614 infants. Of those 614 infants, 99.7% (n = 612) of the maternal residential addresses at delivery were successfully geocoded. Infants with spina bifida living in rural areas in Florida experienced travel times almost twice as high compared with those living in urban areas. When aggregated at county levels, one-way network travel times exhibited statistically significant spatial autocorrelation, indicating that families living in some clusters of counties experienced substantially greater travel times compared with families living in other areas of Florida.

CONCLUSION

This analysis demonstrates the usefulness of linking birth defects registry and hospital discharge data to examine geographic differences in access to medical care. Geographic Information Systems methods are important in evaluating accessibility and geographic barriers to care and could be used among children with special health care needs, including children with birth defects.

Impact of geocoding methods on associations between long-term exposure to urban air pollution and lung function

BACKGROUND

Errors in address geocodes may affect estimates of the effects of air pollution on health.

OBJECTIVE

We investigated the impact of four geocoding techniques on the association between urban air pollution estimated with a fine-scale (10 m × 10 m) dispersion model and lung function in adults.

METHODS

We measured forced expiratory volume in 1 sec (FEV1) and forced vital capacity (FVC) in 354 adult residents of Grenoble, France, who were participants in two well-characterized studies, the Epidemiological Study on the Genetics and Environment on Asthma (EGEA) and the European Community Respiratory Health Survey (ECRHS). Home addresses were geocoded using individual building matching as the reference approach and three spatial interpolation approaches. We used a dispersion model to estimate mean PM10 and nitrogen dioxide concentrations at each participant's address during the 12 months preceding their lung function measurements. Associations between exposures and lung function parameters were adjusted for individual confounders and same-day exposure to air pollutants. The geocoding techniques were compared with regard to geographical distances between coordinates, exposure estimates, and associations between the estimated exposures and health effects.

RESULTS

Median distances between coordinates estimated using the building matching and the three interpolation techniques were 26.4, 27.9, and 35.6 m. Compared with exposure estimates based on building matching, PM10 concentrations based on the three interpolation techniques tended to be overestimated. When building matching was used to estimate exposures, a one-interquartile range increase in PM10 (3.0 μg/m3) was associated with a 3.72-point decrease in FVC% predicted (95% CI: -0.56, -6.88) and a 3.86-point decrease in FEV1% predicted (95% CI: -0.14, -3.24). The magnitude of associations decreased when other geocoding approaches were used [e.g., for FVC% predicted -2.81 (95% CI: -0.26, -5.35) using NavTEQ, or 2.08 (95% CI -4.63, 0.47, p = 0.11) using Google Maps].

CONCLUSIONS

Our findings suggest that the choice of geocoding technique may influence estimated health effects when air pollution exposures are estimated using a fine-scale exposure model.

Racial differences in the built environment--body mass index relationship? A geospatial analysis of adolescents in urban neighborhoods

BACKGROUND

Built environment features of neighborhoods may be related to obesity among adolescents and potentially related to obesity-related health disparities. The purpose of this study was to investigate spatial relationships between various built environment features and body mass index (BMI) z-score among adolescents, and to investigate if race/ethnicity modifies these relationships. A secondary objective was to evaluate the sensitivity of findings to the spatial scale of analysis (i.e. 400- and 800-meter street network buffers).

METHODS

Data come from the 2008 Boston Youth Survey, a school-based sample of public high school students in Boston, MA. Analyses include data collected from students who had georeferenced residential information and complete and valid data to compute BMI z-score (n = 1,034). We built a spatial database using GIS with various features related to access to walking destinations and to community design. Spatial autocorrelation in key study variables was calculated with the Global Moran's I statistic. We fit conventional ordinary least squares (OLS) regression and spatial simultaneous autoregressive error models that control for the spatial autocorrelation in the data as appropriate. Models were conducted using the total sample of adolescents as well as including an interaction term for race/ethnicity, adjusting for several potential individual- and neighborhood-level confounders and clustering of students within schools.

RESULTS

We found significant positive spatial autocorrelation in the built environment features examined (Global Moran's I most ≥ 0.60; all p = 0.001) but not in BMI z-score (Global Moran's I = 0.07, p = 0.28). Because we found significant spatial autocorrelation in our OLS regression residuals, we fit spatial autoregressive models. Most built environment features were not associated with BMI z-score. Density of bus stops was associated with a higher BMI z-score among Whites (Coefficient: 0.029, p < 0.05). The interaction term for Asians in the association between retail destinations and BMI z-score was statistically significant and indicated an inverse association. Sidewalk completeness was significantly associated with a higher BMI z-score for the total sample (Coefficient: 0.010, p < 0.05). These significant associations were found for the 800-meter buffer.

CONCLUSION

Some relationships between the built environment and adolescent BMI z-score were in the unexpected direction. Our findings overall suggest that the built environment does not explain a large proportion of the variation in adolescent BMI z-score or racial disparities in adolescent obesity. However, there are some differences by race/ethnicity that require further research among adolescents.

The effect of administrative boundaries and geocoding error on cancer rates in California

Geocoding is often used to produce maps of disease rates from the diagnosis addresses of incident cases to assist with disease surveillance, prevention, and control. In this process, diagnosis addresses are converted into latitude/longitude pairs which are then aggregated to produce rates at varying geographic scales such as Census tracts, neighborhoods, cities, counties, and states. The specific techniques used within geocoding systems have an impact on where the output geocode is located and can therefore have an effect on the derivation of disease rates at different geographic aggregations. This paper investigates how county-level cancer rates are affected by the choice of interpolation method when case data are geocoded to the ZIP code level. Four commonly used areal unit interpolation techniques are applied and the output of each is used to compute crude county-level five-year incidence rates of all cancers in California. We found that the rates observed for 44 out of the 58 counties in California vary based on which interpolation method is used, with rates in some counties increasing by nearly 400% between interpolation methods.

Parkinson's disease risk from ambient exposure to pesticides

Due to the heavy and expanding agricultural use of neurotoxic pesticides suspected to affect dopaminergic neurons, it is imperative to closely examine the role of pesticides in the development of Parkinson's disease (PD). We focus our investigation on pesticide use in California's heavily agricultural central valley by utilizing a unique pesticide use reporting system. From 2001 to 2007, we enrolled 362 incident PD cases and 341 controls living in the Central Valley of California. Employing our geographic information system model, we estimated ambient exposures to the pesticides ziram, maneb, and paraquat at work places and residences from 1974 to 1999. At workplaces, combined exposure to ziram, maneb, and paraquat increased risk of PD three-fold (OR: 3.09; 95% CI: 1.69, 5.64) and combined exposure to ziram and paraquat, excluding maneb exposure, was associated with a 80% increase in risk (OR:1.82; 95% CI: 1.03, 3.21). Risk estimates for ambient workplace exposure were greater than for exposures at residences and were especially high for younger onset PD patients and when exposed in both locations. Our study is the first to implicate ziram in PD etiology. Combined ambient exposure to ziram and paraquat as well as combined ambient exposure to maneb and paraquat at both workplaces and residences increased PD risk substantially. Those exposed to ziram, maneb, and paraquat together experienced the greatest increase in PD risk. Our results suggest that pesticides affecting different mechanisms that contribute to dopaminergic neuron death may act together to increase the risk of PD considerably.

Socioeconomic status and survival after an invasive breast cancer diagnosis

BACKGROUND

Women who live in geographic areas with high poverty rates and low levels of education experience poorer survival after a breast cancer diagnosis than women who live in communities with indicators of high socioeconomic status (SES). However, very few studies have examined individual-level SES in relation to breast cancer survival or have assessed the contextual role of community-level SES independent of individual-level SES.

METHODS

The authors of this report examined both individual-level and community-level SES in relation to breast cancer survival in a population-based cohort of women ages 20 to 69 years who were diagnosed with breast cancer in Wisconsin between 1995 and 2003 (N = 5820).

RESULTS

Compared with college graduates, women who had no education beyond high school were 1.39 times more likely (95% confidence interval [CI], 1.10-1.76) to die from breast cancer. Women who had household incomes <2.5 times the poverty level were 1.46 times more likely (95% CI, 1.10-1.92) to die from breast cancer than women who had household incomes ≥5 times the poverty level. Adjusting the analysis for use of screening mammography, disease stage at diagnosis, and lifestyle factors eliminated the disparity by income, but the disparity by education persisted (hazard ratio [HR], 1.27; 95% CI, 0.99-1.61). In multilevel analyses, low community-level education was associated with increased breast cancer mortality even after adjusting for individual-level SES (HR, 1.57; 95% CI, 1.09-2.27 for ≥20% vs <10% of adults without a high school degree).

CONCLUSIONS

The current results indicated that screening and early detection explain some of the disparity according to SES, but further research will be needed to understand the additional ways in which individual-level and community-level education are associated with survival.

Analyzing spatial aggregation error in statistical models of late-stage cancer risk: a Monte Carlo simulation approach

PURPOSE

This paper examines the effect of spatial aggregation error on statistical estimates of the association between spatial access to health care and late-stage cancer.

METHODS

Monte Carlo simulation was used to disaggregate cancer cases for two Illinois counties from zip code to census block in proportion to the age-race composition of the block population. After the disaggregation, a hierarchical logistic model was estimated examining the relationship between late-stage breast cancer and risk factors including travel distance to mammography, at both the zip code and census block levels. Model coefficients were compared between the two levels to assess the impact of spatial aggregation error.

RESULTS

We found that spatial aggregation error influences the coefficients of regression-type models at the zip code level, and this impact is highly dependent on the study area. In one study area (Kane County), block-level coefficients were very similar to those estimated on the basis of zip code data; whereas in the other study area (Peoria County), the two sets of coefficients differed substantially raising the possibility of drawing inaccurate inferences about the association between distance to mammography and late-stage cancer risk.

CONCLUSIONS

Spatial aggregation error can significantly affect the coefficient values and inferences drawn from statistical models of the association between cancer outcomes and spatial and non-spatial variables. Relying on data at the zip code level may lead to inaccurate findings on health risk factors.

Moving beyond traditional measurement of mortality after injury: evaluation of risks for late death

BACKGROUND

The purpose of this study was to evaluate long-term mortality after trauma, and to determine risk factors and possible disparities related to mortality after hospital discharge.

STUDY DESIGN

Level I trauma center registry data from a 6-year period (2000 through 2005) were linked to patient electronic medical records, the National Death Index with cause of death codes, and census data using geographic information science (GIS) methodologies. Census data provided supplemental demographic and socioeconomic information from patient neighborhoods.

RESULTS

The hospital mortality rate for 15,285 patients was 3.3%, and mortality after discharge was 4.8%. Overall mortality for the study period was 8.1% (average follow-up, 2.8 years, 1-year mortality, 5.4%). Mortality after discharge was related to the initial injury in 33%, possibly related in 23%, and unrelated in 44% of patients. Logistic regression analysis demonstrated that independent predictors of hospital mortality were age, Injury Severity Score, gunshot injury, significant head injury, fall, and spinal cord injury. In contrast, independent risk factors for mortality after discharge were age, hospital length of stay, discharge from the hospital to a locale other than home, and the presence of spinal cord injury. Intoxication at hospital admission and injury due to a gunshot wound or motor vehicle collision were protective for late mortality. Bivariate analysis of census data demonstrated that lower socioeconomic status was associated with improved hospital survival, and non-native status was associated with mortality after discharge.

CONCLUSIONS

There is significant mortality attributable to trauma for up to 1 year after hospital discharge. These findings suggest that mortality after trauma needs to be measured beyond hospital discharge in order to assess the complete impact of injury.

Paraoxonase 1, agricultural organophosphate exposure, and Parkinson disease

BACKGROUND

Human, animal and cell models support a role for pesticides in the etiology of Parkinson disease. Susceptibility to pesticides may be modified by genetic variants of xenobiotic enzymes, such as paraoxonase, that play a role in metabolizing some organophosphates.

METHODS

We examined associations between Parkinson disease and the organophosphates diazinon, chlorpyrifos, and parathion, and the influence of a functional polymorphism at position 55 in the coding region of the PON1 gene (PON1-55). From 1 January 2001 through 1 January 2008, we recruited 351 incident cases and 363 controls from 3 rural California counties in a population-based case-control study. Participants provided a DNA sample, and residential exposure to organophosphates was determined from pesticide usage reports and a geographic information system (GIS) approach. We assessed the main effects of both genes and pesticides in unconditional logistic regression analyses, and evaluated the effect of carrying a PON1-55 MM variant on estimates of effects for diazinon, chlorpyrifos, and parathion exposures.

RESULTS

Carriers of the variant MM PON1-55 genotype exposed to organophosphates exhibited a greater than 2-fold increase in Parkinson disease risk compared with persons who had the wildtype or heterozygous genotype and no exposure (for diazinon, odds ratio = 2.2 [95% confidence interval = 1.1-4.5]; for chlorpyrifos, 2.6 [1.3-5.4]). The effect estimate for chlorpyrifos, was more pronounced in younger-onset cases and controls (<or=60 years) (5.3 [1.7-16]). No increase in risk was noted for parathion.

CONCLUSION

The increase in risk we observed among PON1-55 variant carriers for specific organophosphates metabolized by PON1 underscores the importance of considering susceptibility factors when studying environmental exposures in Parkinson disease.

The effects of local street network characteristics on the positional accuracy of automated geocoding for geographic health studies

Background

Automated geocoding of patient addresses for the purpose of conducting spatial epidemiologic studies results in positional errors. It is well documented that errors tend to be larger in rural areas than in cities, but possible effects of local characteristics of the street network, such as street intersection density and street length, on errors have not yet been documented. Our study quantifies effects of these local street network characteristics on the means and the entire probability distributions of positional errors, using regression methods and tolerance intervals/regions, for more than 6000 geocoded patient addresses from an Iowa county.

Results

Positional errors were determined for 6376 addresses in Carroll County, Iowa, as the vector difference between each 100%-matched automated geocode and its ground-truthed location. Mean positional error magnitude was inversely related to proximate street intersection density. This effect was statistically significant for both rural and municipal addresses, but more so for the former. Also, the effect of street segment length on geocoding accuracy was statistically significant for municipal, but not rural, addresses; for municipal addresses mean error magnitude increased with length.

Conclusion

Local street network characteristics may have statistically significant effects on geocoding accuracy in some places, but not others. Even in those locales where their effects are statistically significant, street network characteristics may explain a relatively small portion of the variability among geocoding errors. It appears that additional factors besides rurality and local street network characteristics affect accuracy in general.

Spatial autocorrelation among automated geocoding errors and its effects on testing for disease clustering

Automated geocoding of patient addresses is an important data assimilation component of many spatial epidemiologic studies. Inevitably, the geocoding process results in positional errors. Positional errors incurred by automated geocoding tend to reduce the power of tests for disease clustering and otherwise affect spatial analytic methods. However, there are reasons to believe that the errors may often be positively spatially correlated and that this may mitigate their deleterious effects on spatial analyses. In this article, we demonstrate explicitly that the positional errors associated with automated geocoding of a data set of more than 6000 addresses in Carroll County, Iowa are spatially autocorrelated. Furthermore, through two simulation studies of disease processes, including one in which the disease process is overlain upon the Carroll County addresses, we show that spatial autocorrelation among geocoding errors maintains the power of two tests for disease clustering at a level higher than that which would occur if the errors were independent. Implications of these results for cluster detection, privacy protection, and measurement error modeling of geographic health data are discussed.

Methods for retrospective geocoding in population studies: the Jackson Heart Study

The increasing use of geographic information systems (GIS) in epidemiological population studies requires careful attention to the methods employed in accomplishing geocoding and creating a GIS. Studies have provided limited details,hampering the ability to assess validity of spatial data. The purpose of this paper is to describe the multiphase geocoding methods used to retrospectively create a GIS in the Jackson Heart Study (JHS). We used baseline data from 5,302 participants enrolled in the JHS between 2000 and 2004 in a multiphase process to accomplish geocoding2 years after participant enrollment. After initial deletion of ungeocodable addresses(n=52), 96% were geocoded using ArcGIS. An interactive method using data abstraction from participant records, use of additional maps and street reference files,and verification of existence of address, yielded successful geocoding of all but 13 addresses. Overall, nearly 99% (n=5,237) of the JHS cohort was geocoded retrospectively using the multiple strategies for improving and locating geocodable addresses. Geocoding validation procedures revealed highly accurate and reliable geographic data. Using the methods and protocol developed provided a reliable spatial database that can be used for further investigation of spatial epidemiology. Baseline results were used to describe participants by select geographic indicators, including residence in urban or rural areas, as well as to validate the effectiveness of the study's sampling plan. Further, our results indicate that retrospectively developing a reliable GIS for a large, epidemiological study is feasible. This paper describes some of the challenges in retrospectively creating a GIS and provides practical tips that enhanced the success.

Well-water consumption and Parkinson's disease in rural California

INTRODUCTION

Investigators have hypothesized that consuming pesticide-contaminated well water plays a role in Parkinson's disease (PD), and several previous epidemiologic studies support this hypothesis.

OBJECTIVES

We investigated whether consuming water from private wells located in areas with documented historical pesticide use was associated with an increased risk of PD.

METHODS

We employed a geographic information system (GIS)-based model to estimate potential well-water contamination from agricultural pesticides among 368 cases and 341 population controls enrolled in the Parkinson's Environment and Genes Study (PEG). We separately examined 6 pesticides (diazinon, chlorpyrifos, propargite, paraquat, dimethoate, and methomyl) from among 26 chemicals selected for their potential to pollute groundwater or for their interest in PD, and because at least 10% of our population was exposed to them.

RESULTS

Cases were more likely to have consumed private well water and to have consumed it on average 4.3 years longer than controls (p = 0.02). High levels of possible well-water contamination with methomyl [odds ratio (OR) = 1.67; 95% confidence interval (CI), 1.00-2.78]), chlorpyrifos (OR = 1.87; 95% CI, 1.05-3.31), and propargite (OR = 1.92; 95% CI, 1.15-3.20) resulted in approximately 70-90% increases in relative risk of PD. Adjusting for ambient pesticide exposures only slightly attenuated these increases. Exposure to a higher number of water-soluble pesticides and organophosphate pesticides also increased the relative risk of PD.

CONCLUSION

Our study, the first to use agricultural pesticide application records, adds evidence that consuming well water presumably contaminated with pesticides may play a role in the etiology of PD.

Geographic variability in geocoding success for West Nile virus cases in South Dakota

Geocoding, the process of assigning each case a set of coordinates that closely approximates its true location, is an important component of spatial epidemiological studies. The failure to accurately geocode cases adversely affects the validity and strength of conclusions drawn from the analysis. We investigated whether there were differences among geographic locations and demographic classes in the ability to successfully geocode West Nile virus (WNV) cases in South Dakota. We successfully geocoded 1354 cases (80.8%) to their street address locations and assigned all 1676 cases to ZIP code tabulation areas (ZCTAs). Using spatial scan statistics, significant clusters of non-geocoded cases were identified in central and western South Dakota. Geocoding success rates were lower in areas of low population density and on Indian reservations than in other portions of the state. Geocoding success rates were lower for Native Americans than for other races. Spatial epidemiological studies should consider the potential biases that may result from excluding non-geocoded cases, particularly in rural portions of the Great Plains that contain large Native American populations.

Dopamine transporter genetic variants and pesticides in Parkinson's disease

BACKGROUND

Research suggests that independent and joint effects of genetic variability in the dopamine transporter (DAT) locus and pesticides may influence Parkinson's disease (PD) risk.

METHODS

In 324 incident PD patients and 334 population controls from our rural California case-control study, we genotyped rs2652510, rs2550956 (for the DAT 5' clades), and the 3' variable number of tandem repeats (VNTR). Using geographic information system methods, we determined residential exposure to agricultural maneb and paraquat applications. We also collected occupational pesticide use data. Employing logistic regression, we calculated odds ratios (ORs) for clade diplotypes, VNTR genotype, and number of susceptibility (A clade and 9-repeat) alleles and assessed susceptibility allele-pesticide interactions.

RESULTS

PD risk was increased separately in DAT A clade diplotype carriers [AA vs. BB: OR = 1.66; 95% confidence interval (CI), 1.08-2.57] and 3' VNTR 9/9 carriers (9/9 vs. 10/10: OR = 1.8; 95% CI, 0.96-3.57), and our data suggest a gene dosing effect. Importantly, high exposure to paraquat and maneb in carriers of one susceptibility allele increased PD risk 3-fold (OR = 2.99; 95% CI, 0.88-10.2), and in carriers of two or more alleles more than 4-fold (OR = 4.53; 95% CI, 1.70-12.1). We obtained similar results for occupational pesticide measures.

DISCUSSION

Using two independent pesticide measures, we a) replicated previously reported gene-environment interactions between DAT genetic variants and occupational pesticide exposure in men and b) overcame previous limitations of nonspecific pesticide measures and potential recall bias by employing state records and computer models to estimate residential pesticide exposure.

CONCLUSION

Our results suggest that DAT genetic variability and pesticide exposure interact to increase PD risk.

Parkinson's disease and residential exposure to maneb and paraquat from agricultural applications in the central valley of California

Evidence from animal and cell models suggests that pesticides cause a neurodegenerative process leading to Parkinson's disease (PD). Human data are insufficient to support this claim for any specific pesticide, largely because of challenges in exposure assessment. The authors developed and validated an exposure assessment tool based on geographic information systems that integrated information from California Pesticide Use Reports and land-use maps to estimate historical exposure to agricultural pesticides in the residential environment. In 1998-2007, the authors enrolled 368 incident PD cases and 341 population controls from the Central Valley of California in a case-control study. They generated estimates for maneb and paraquat exposures incurred between 1974 and 1999. Exposure to both pesticides within 500 m of the home increased PD risk by 75% (95% confidence interval (CI): 1.13, 2.73). Persons aged < or =60 years at the time of diagnosis were at much higher risk when exposed to either maneb or paraquat alone (odds ratio = 2.27, 95% CI: 0.91, 5.70) or to both pesticides in combination (odds ratio = 4.17, 95% CI: 1.15, 15.16) in 1974-1989. This study provides evidence that exposure to a combination of maneb and paraquat increases PD risk, particularly in younger subjects and/or when exposure occurs at younger ages.

Fast food and deprivation in Nova Scotia

OBJECTIVE

To examine the relationship between density of fast food restaurants and measures of social and material deprivation at the community level in Nova Scotia, Canada.

METHODS

Census information on population and key variables required for the calculation of deprivation indices were obtained for 266 communities in Nova Scotia. The density of fast food restaurants per 1000 individuals for each community was calculated and communities were divided into quintiles of material and psychosocial deprivation. One-way analysis of variance was used to investigate associations between fast food outlet densities and deprivation scores at the community level.

RESULTS

A statistically significant inverse association was found between community-level material deprivation and the mean number of fast food restaurants per 1000 people for Nova Scotia (p < 0.000). Significant positive relationships were found between density of fast food restaurants and psychosocial deprivation (p < 0.000). Both associations were principally linear with greater fast food outlet density occurring as material deprivation decreased and as psychosocial deprivation increased.

INTERPRETATION

Community-level deprivation in Nova Scotia is associated with fast food outlet density and lends support for environmental explanations for variations in the prevalence of obesity. Such findings are valuable to population health intervention initiatives targeting the modification of environmental determinants of obesity.

Exposure to Tetrachloroethylene-Contaminated Drinking Water and the Risk of Pregnancy Loss

There is little information on the impact of solvent-contaminated drinking water on pregnancy outcomes. This retrospective cohort study examined whether maternal exposure to tetrachloroethylene (PCE) - contaminated drinking water in the Cape Cod region of Massachusetts influenced the risk of clinically recognized pregnancy loss. The study identified exposed (n=959) and unexposed (1,087) women who completed a questionnaire on their residential and pregnancy histories, and confounding variables. Exposure was estimated using water distribution system modeling software. No meaningful associations were seen between PCE exposure level and the risk of clinically recognized pregnancy loss at the exposure levels experienced by the study population. Because PCE remains a common water contaminant, it is important to continue monitoring its impact on women and their pregnancies.

An effective and efficient approach for manually improving geocoded data

Background

The process of geocoding produces output coordinates of varying degrees of quality. Previous studies have revealed that simply excluding records with low-quality geocodes from analysis can introduce significant bias, but depending on the number and severity of the inaccuracies, their inclusion may also lead to bias. Little quantitative research has been presented on the cost and/or effectiveness of correcting geocodes through manual interactive processes, so the most cost effective methods for improving geocoded data are unclear. The present work investigates the time and effort required to correct geocodes contained in five health-related datasets that represent examples of data commonly used in Health GIS.

Results

Geocode correction was attempted on five health-related datasets containing a total of 22,317 records. The complete processing of these data took 11.4 weeks (427 hours), averaging 69 seconds of processing time per record. Overall, the geocodes associated with 12,280 (55%) of records were successfully improved, taking 95 seconds of processing time per corrected record on average across all five datasets. Geocode correction improved the overall match rate (the number of successful matches out of the total attempted) from 79.3 to 95%. The spatial shift between the location of original successfully matched geocodes and their corrected improved counterparts averaged 9.9 km per corrected record. After geocode correction the number of city and USPS ZIP code accuracy geocodes were reduced from 10,959 and 1,031 to 6,284 and 200, respectively, while the number of building centroid accuracy geocodes increased from 0 to 2,261.

Conclusion

The results indicate that manual geocode correction using a web-based interactive approach is a feasible and cost effective method for improving the quality of geocoded data. The level of effort required varies depending on the type of data geocoded. These results can be used to choose between data improvement options (e.g., manual intervention, pseudocoding/geo-imputation, field GPS readings).

US secondary schools and food outlets

We examined the availability of fast food restaurants and convenience stores within walking distance (0.5 miles or 805 m) of US public secondary schools. We found that one-third of schools nationwide have at least one fast food restaurant or convenience store within walking distance. In multivariate analyses, schools in the lowest-income versus the highest-income neighborhoods have more fast food restaurants and convenience stores, while schools in African-American versus White neighborhoods generally have fewer food outlets. Furthermore, urban neighborhoods with a high school versus no secondary school have more food outlets. Curbing the obesity epidemic among adolescents requires addressing the food environment surrounding schools.

A GIS-based method for household recruitment in a prospective pesticide exposure study

BACKGROUND

Recent advances in GIS technology and remote sensing have provided new opportunities to collect ecologic data on agricultural pesticide exposure. Many pesticide studies have used historical or records-based data on crops and their associated pesticide applications to estimate exposure by measuring residential proximity to agricultural fields. Very few of these studies collected environmental and biological samples from study participants. One of the reasons for this is the cost of identifying participants who reside near study fields and analyzing samples obtained from them. In this paper, we present a cost-effective, GIS-based method for crop field selection and household recruitment in a prospective pesticide exposure study in a remote location. For the most part, our multi-phased approach was carried out in a research facility, but involved two brief episodes of fieldwork for ground truthing purposes. This method was developed for a larger study designed to examine the validity of indirect pesticide exposure estimates by comparing measured exposures in household dust, water and urine with records-based estimates that use crop location, residential proximity and pesticide application data. The study focused on the pesticide atrazine, a broadleaf herbicide used in corn production and one of the most widely-used pesticides in the U.S.

RESULTS

We successfully used a combination of remotely-sensed data, GIS-based methods and fieldwork to select study fields and recruit participants in Illinois, a state with high corn production and heavy atrazine use. Our several-step process consisted of the identification of potential study fields and residential areas using aerial photography; verification of crop patterns and land use via site visits; development of a GIS-based algorithm to define recruitment areas around crop fields; acquisition of geocoded household-level data within each recruitment area from a commercial vendor; and confirmation of final participant household locations via ground truthing. The use of these procedures resulted in a sufficient sample of participants from 14 recruitment areas in seven Illinois counties.

CONCLUSION

One of the challenges in pesticide research is the identification and recruitment of study participants, which is time consuming and costly, especially when the study site is in a remote location. We have demonstrated how GIS-based processes can be used to recruit participants, increase efficiency and enhance accuracy. The method that we used ultimately made it possible to collect biological samples from a specific demographic group within strictly defined exposure areas, with little advance knowledge of the location or population.

Geocoding accuracy and the recovery of relationships between environmental exposures and health

Background

This research develops methods for determining the effect of geocoding quality on relationships between environmental exposures and health. The likelihood of detecting an existing relationship – statistical power – between measures of environmental exposures and health depends not only on the strength of the relationship but also on the level of positional accuracy and completeness of the geocodes from which the measures of environmental exposure are made. This paper summarizes the results of simulation studies conducted to examine the impact of inaccuracies of geocoded addresses generated by three types of geocoding processes: a) addresses located on orthophoto maps, b) addresses matched to TIGER files (U.S Census or their derivative street files); and, c) addresses from E-911 geocodes (developed by local authorities for emergency dispatch purposes).

Results

The simulated odds of disease using exposures modelled from the highest quality geocodes could be sufficiently recovered using other, more commonly used, geocoding processes such as TIGER and E-911; however, the strength of the odds relationship between disease exposures modelled at geocodes generally declined with decreasing geocoding accuracy.

Conclusion

Although these specific results cannot be generalized to new situations, the methods used to determine the sensitivity of results can be used in new situations. Estimated measures of positional accuracy must be used in the interpretation of results of analyses that investigate relationships between health outcomes and exposures measured at residential locations. Analyses similar to those employed in this paper can be used to validate interpretation of results from empirical analyses that use geocoded locations with estimated measures of positional accuracy.

Validation of the geographic position of EPER-Spain industries

Background

The European Pollutant Emission Register in Spain (EPER-Spain) is a public inventory of pollutant industries created by decision of the European Union. The location of these industries is geocoded and the first published data correspond to 2001. Publication of these data will allow for quantification of the effect of proximity to one or more such plant on cancer and all-cause mortality observed in nearby towns. However, as errors have been detected in the geocoding of many of the pollutant foci shown in the EPER, it was decided that a validation study should be conducted into the accuracy of these co-ordinates. EPER-Spain geographic co-ordinates were drawn from the European Environment Agency (EEA) server and the Spanish Ministry of the Environment (MOE). The Farm Plot Geographic Information System (Sistema de Información Geográfica de Parcelas Agrícolas) (SIGPAC) enables orthophotos (digitalized aerial images) of any territorial point across Spain to be obtained. Through a search of co-ordinates in the SIGPAC, all the industrial foci (except farms) were located. The quality criteria used to ascertain possible errors in industrial location were high, medium and low quality, where industries were situated at a distance of less than 500 metres, more than 500 metres but less than 1 kilometre, and more than 1 kilometre from their real locations, respectively.

Results

Insofar as initial registry quality was concerned, 84% of industrial complexes were inaccurately positioned (low quality) according to EEA data versus 60% for Spanish MOE data. The distribution of the distances between the original and corrected co-ordinates for each of the industries on the registry revealed that the median error was 2.55 kilometres for Spain overall (according to EEA data). The Autonomous Regions that displayed most errors in industrial geocoding were Murcia, Canary Islands, Andalusia and Madrid. Correct co-ordinates were successfully allocated to 100% of EPER-Spain industries.

Conclusion

Knowing the exact location of pollutant foci is vital to obtain reliable and valid conclusions in any study where distance to the focus is a decisive factor, as in the case of the consequences of industrial pollution on the health of neighbouring populations.

Error and bias in determining exposure potential of children at school locations using proximity-based GIS techniques

BACKGROUND

The widespread availability of powerful tools in commercial geographic information system (GIS) software has made address geocoding a widely employed technique in spatial epidemiologic studies.

OBJECTIVE

The objective of this study was to determine the effect of the positional error in geocoding on the analysis of exposure to traffic-related air pollution of children at school locations.

METHODS

For a case study of Orange County, Florida, we determined the positional error of geocoding of school locations through comparisons with a parcel database and digital orthophotography. We used four different geocoding techniques for comparison to establish the repeatability of geocoding, and an analysis of proximity to major roads to determine bias and error in environmental exposure assessment.

RESULTS

RESULTS INDICATE THAT THE POSITIONAL ERROR IN GEOCODING OF SCHOOLS IS VERY SUBSTANTIAL: We found that the 95% root mean square error was 196 m using street centerlines, 306 m using TIGER roads, and 210 and 235 m for two commercial geocoding firms. We found bias and error in proximity analysis to major roads to be unacceptably large at distances of < 500 m. Bias and error are introduced by lack of positional accuracy and lack of repeatability of geocoding of school locations.

CONCLUSIONS

These results suggest that typical geocoding is insufficient for fine-scale analysis of school locations and more accurate alternatives need to be considered.

Estimating the Intensity of a Spatial Point Process from Locations Coarsened by Incomplete Geocoding

The estimation of spatial intensity is an important inference problem in spatial epidemiologic studies. A standard data assimilation component of these studies is the assignment of a geocode, that is, point-level spatial coordinates, to the address of each subject in the study population. Unfortunately, when geocoding is performed by the standard automated method of street-segment matching to a georeferenced road file and subsequent interpolation, it is rarely completely successful. Typically, 10-30% of the addresses in the study population, and even higher percentages in particular subgroups, fail to geocode, potentially leading to a selection bias, called geographic bias, and an inefficient analysis. Missing-data methods could be considered for analyzing such data; however, because there is almost always some geographic information coarser than a point (e.g., a Zip code) observed for the addresses that fail to geocode, a coarsened-data analysis is more appropriate. This article develops methodology for estimating spatial intensity from coarsened geocoded data. Both nonparametric (kernel smoothing) and likelihood-based estimation procedures are considered. Substantial improvements in the estimation quality of coarsened-data analyses relative to analyses of only the observations that geocode are demonstrated via simulation and an example from a rural health study in Iowa.

Comparing a single-stage geocoding method to a multi-stage geocoding method: how much and where do they disagree?

Background

Geocoding methods vary among spatial epidemiology studies. Errors in the geocoding process and differential match rates may reduce study validity. We compared two geocoding methods using 8,157 Washington State addresses. The multi-stage geocoding method implemented by the state health department used a sequence of local and national reference files. The single-stage method used a single national reference file. For each address geocoded by both methods, we measured the distance between the locations assigned by each method. Area-level characteristics were collected from census data, and modeled as predictors of the discordance between geocoded address coordinates.

Results

The multi-stage method had a higher match rate than the single-stage method: 99% versus 95%. Of 7,686 addresses were geocoded by both methods, 96% were geocoded to the same census tract by both methods and 98% were geocoded to locations within 1 km of each other by the two methods. The distance between geocoded coordinates for the same address was higher in sparsely populated and low poverty areas, and counties with local reference files.

Conclusion

The multi-stage geocoding method had a higher match rate than the single-stage method. An examination of differences in the location assigned to the same address suggested that study results may be most sensitive to the choice of geocoding method in sparsely populated or low-poverty areas.

Influence of geocoding quality on environmental exposure assessment of children living near high traffic roads

Background

The widespread availability of powerful geocoding tools in commercial GIS software and the interest in spatial analysis at the individual level have made address geocoding a widely employed technique in epidemiological studies. This study determined the effect of the positional error in street geocoding on the analysis of traffic-related air pollution on children.

Methods

For a case-study of a large sample of school children in Orange County, Florida (n = 104,865) the positional error of street geocoding was determined through comparison with a parcel database. The effect of this error was evaluated by analyzing the proximity of street and parcel geocoded locations to road segments with high traffic volume and determining the accuracy of the classification using the results of street geocoding. Of the original sample of 163,886 addresses 36% were not used in the final analysis because they could not be reliably geocoded using either street or parcel geocoding. The estimates of positional error can therefore be considered conservative underestimates.

Results

Street geocoding was found to have a median error of 41 meters, a 90^th percentile of 100 meters, a 95^th percentile of 137 meters and a 99^th percentile of 273 meters. These positional errors were found to be non-random in nature and introduced substantial bias and error in the estimates of potential exposure to traffic-related air pollution. Street geocoding was found to consistently over-estimate the number of potentially exposed children at small distances up to 250 meters. False positives and negatives were also found to be very common at these small distances.

Conclusion

Results of the case-study presented here strongly suggest that typical street geocoding is insufficient for fine-scale analysis and more accurate alternatives need to be considered.

The accuracy of address coding and the effects of coding errors

Studies that estimate the influence of characteristics of place on health often use geocoded addresses to identify location of study subjects. This study uses housing built before 1990 selected for the 1995-2001 National Health Interview Survey (N=252,421) to develop a standard against which geocodes obtained from an address-coding program are compared. The results show that geocoding is generally accurate and is more successful in urban areas. Blockgroups with missing codes are more rural and somewhat poorer than blockgroups with correct codes. The effect of incorrect codes on statistical analyses depends on the proportion rural in a study population.

Modeling the probability distribution of positional errors incurred by residential address geocoding

Background

The assignment of a point-level geocode to subjects' residences is an important data assimilation component of many geographic public health studies. Often, these assignments are made by a method known as automated geocoding, which attempts to match each subject's address to an address-ranged street segment georeferenced within a streetline database and then interpolate the position of the address along that segment. Unfortunately, this process results in positional errors. Our study sought to model the probability distribution of positional errors associated with automated geocoding and E911 geocoding.

Results

Positional errors were determined for 1423 rural addresses in Carroll County, Iowa as the vector difference between each 100%-matched automated geocode and its true location as determined by orthophoto and parcel information. Errors were also determined for 1449 60%-matched geocodes and 2354 E911 geocodes. Huge (> 15 km) outliers occurred among the 60%-matched geocoding errors; outliers occurred for the other two types of geocoding errors also but were much smaller. E911 geocoding was more accurate (median error length = 44 m) than 100%-matched automated geocoding (median error length = 168 m). The empirical distributions of positional errors associated with 100%-matched automated geocoding and E911 geocoding exhibited a distinctive Greek-cross shape and had many other interesting features that were not capable of being fitted adequately by a single bivariate normal or t distribution. However, mixtures of t distributions with two or three components fit the errors very well.

Conclusion

Mixtures of bivariate t distributions with few components appear to be flexible enough to fit many positional error datasets associated with geocoding, yet parsimonious enough to be feasible for nascent applications of measurement-error methodology to spatial epidemiology.