Chris DuClos

Linking health and environmental data in geographical analysis: it's so much more than centroids.

Programs and studies increasingly use existing data from multiple sources (e.g., surveillance systems, health registries, or governmental agencies) for analysis and inference. These data usually have been collected on different geographical or spatial units, with each varying from the ones of interest. Combining such disparate data creates statistical challenges. Floridas efforts to move toward implementing the Centers for Disease Control and Prevention (CDC)s Environmental Public Health Tracking (EPHT) program aptly illustrate these concerns, which are typical of studies designed to measure the association between environmental and health outcomes. In this paper, we develop models of spatial associations between myocardial infarctions (MIs) and ambient ozone levels in Florida during August 2005 and use these models to illustrate the problems that can occur when making inferences from aggregated data, the concept of spatial support, and the importance of correct uncertainty assessment. Existing data on hospital discharges and emergency department visits were obtained from Floridas Agency for Health Care Administration. Environmental data were obtained from Floridas Department of Environmental Protection; sociodemographic data were obtained from the US Census Bureau; and data from CDCs Behavioral Risk Factor Surveillance System were used to provide additional information on other risk factors. We highlight the opportunities and challenges associated with combining disparate spatial data for EPHT analyses. We compare the results from two different approaches to data linkage, focusing on the need to account for spatial scale and the support of spatial data in the analysis. We use geographically weighted regression, not as a visual mapping tool, but as an inferential tool designed to indicate the need for spatial coefficients, a test that cannot be made by using the majority of Bayesian models. Finally, we use geostatistical simulation methods for uncertainty analysis to demonstrate its importance in models with predicted covariates. Our focus is on relatively simple methods and concepts that can be implemented with ESRIs(®) ArcGIS(®) software.

Associations between exposure to ambient benzene and PM2.5 during pregnancy and the risk of selected birth defects in offspring

OBJECTIVE A growing number of studies have investigated the association between air pollution and the risk of birth defects, but results are inconsistent. The objective of this study was to examine whether maternal exposure to ambient PM2.5 or benzene increases the risk of selected birth defects in Florida. METHODS We conducted a retrospective cohort study of singleton infants born in Florida from 2000 to 2009. Isolated and non-isolated birth defect cases of critical congenital heart defects, orofacial clefts, and spina bifida were identified from the Florida Birth Defects Registry. Estimates of maternal exposures to PM2.5 and benzene for all case and non-case pregnancies were derived by aggregation of ambient measurement data, obtained from the US Environmental Protection Agency Air Quality System, during etiologically relevant time windows. Multivariable Poisson regression was used to estimate adjusted prevalence ratios (aPRs) and 95% confidence intervals (CIs) for each quartile of air pollutant exposure. RESULTS Compared to the first quartile of PM2.5 exposure, higher levels of exposure were associated with an increased risk of non-isolated truncus arteriosus (aPR4th Quartile, 8.80; 95% CI, 1.11-69.50), total anomalous pulmonary venous return (aPR2nd Quartile, 5.00; 95% CI, 1.10-22.84), coarctation of the aorta (aPR4th Quartile, 1.72; 95% CI, 1.15-2.57; aPR3rd Quartile, 1.60; 95% CI, 1.07-2.41), interrupted aortic arch (aPR4th Quartile, 5.50; 95% CI, 1.22-24.82), and isolated and non-isolated any critical congenital heart defect (aPR3rd Quartile, 1.13; 95% CI, 1.02-1.25; aPR4th Quartile, 1.33; 95% CI, 1.07-1.65). Mothers with the highest level of exposure to benzene were more likely to deliver an infant with an isolated cleft palate (aPR4th Quartile, 1.52; 95% CI, 1.13-2.04) or any orofacial cleft (aPR4th Quartile, 1.29; 95% CI, 1.08-1.56). An inverse association was observed between exposure to benzene and non-isolated pulmonary atresia (aPR4th Quartile, 0.19; 95% CI, 0.04-0.84). CONCLUSION Our results suggest a few associations between exposure to ambient PM2.5 or benzene and specific birth defects in Florida. However, many related comparisons showed no association. Hence, it remains unclear whether associations are clinically significant or can be causally related to air pollution exposures.

Assessing the association between environmental impacts and health outcomes: a case study from Florida.

The Centers for Disease Control and Prevention (CDC) created the Environmental Public Health Tracking (EPHT) program to integrate hazard monitoring, exposure, and health effects surveillance into a cohesive tracking network. Part of Floridas effort to move toward implementation of EPHT is to develop models of the spatial and temporal association between myocardial infarctions (MIs) and ambient ozone levels in Florida. Existing data were obtained from Floridas Agency for Health Care Administration, Floridas Department of Environmental Protection, the U.S. Census Bureau, and CDCs Behavioral Risk Factor Surveillance System. These data were linked by both ignoring spatial support and using block kriging, a support-adjusted approach. The MI data were indirectly standardized by age, race/ethnicity, and sex. The state of Florida was used as the comparison standard to compute the MI standardized event ratio (SER) for each county and each month. After the data were linked, global models were used initially to relate MIs to ambient ozone levels, adjusting for covariates. The global models provide an estimated relative MI SER for the state. Realizing that the association in MIs and ozone might change across locations, local models were used to estimate the relative MI SER for each county, again adjusting for covariates. Results differed, depending on whether the spatial support was ignored or accounted for in the models. The opportunities and challenges associated with EPHT analyses are discussed and future directions highlighted.

Tracking childhood exposure to lead and developmental disabilities: examining the relationship in a population-based sample.

Elevated levels of lead detected in the blood are associated with harmful effects on childrens learning and behavior. The goal of the current Environmental Public Health Tracking Project was to examine the relationship between selected developmental disabilities and childhood blood lead levels in a population-based sample. Using extant datasets from the Florida Department of Health, Childhood Lead Poisoning Prevention Program, and the Florida Department of Education, we were able to isolate a linked dataset of children who were tested for lead poisoning and attended public schools. Special education categories served as a proxy for developmental disabilities; the prevalence of these disabilities in the sample of children with blood lead levels was compared with that in children who attended the same schools but were not tested for lead poisoning. Results indicated that children screened for lead poisoning were more likely to be receiving services for behavior problems, mental retardation, learning disabilities, or a speech-language impairment than other children attending the same schools. Implications for using administrative datasets to examine this relationship are discussed.

Identifying Heat Waves in Florida: Considerations of Missing Weather Data.

Background Using current climate models, regional-scale changes for Florida over the next 100 years are predicted to include warming over terrestrial areas and very likely increases in the number of high temperature extremes. No uniform definition of a heat wave exists. Most past research on heat waves has focused on evaluating the aftermath of known heat waves, with minimal consideration of missing exposure information. Objectives To identify and discuss methods of handling and imputing missing weather data and how those methods can affect identified periods of extreme heat in Florida. Methods In addition to ignoring missing data, temporal, spatial, and spatio-temporal models are described and utilized to impute missing historical weather data from 1973 to 2012 from 43 Florida weather monitors. Calculated thresholds are used to define periods of extreme heat across Florida. Results Modeling of missing data and imputing missing values can affect the identified periods of extreme heat, through the missing data itself or through the computed thresholds. The differences observed are related to the amount of missingness during June, July, and August, the warmest months of the warm season (April through September). Conclusions Missing data considerations are important when defining periods of extreme heat. Spatio-temporal methods are recommended for data imputation. A heat wave definition that incorporates information from all monitors is advised.

Using an environmental public health tracking biomonitoring study to validate safe water restoration efforts in Florida.

CONTEXT Inorganic arsenic is a carcinogen when consumed over many years and is the type of arsenic that is associated with well water. Private wells in many central Florida counties have been found to contain levels of arsenic above 10 μg/L, which is the maximum contaminant level (MCL) established by the US Environmental Protection Agency. OBJECTIVE The purpose of this study was to test whether individuals living in homes with arsenic levels above the MCL who use bottled water or have a point of use (POU) filter on their main source of water in the house (eg, kitchen sink) are ingesting unsafe levels of arsenic through other unfiltered water sources in the home. DESIGN Case-control study of residents with private wells above (case) and below (control) the MCL for arsenic in drinking water (10 μg/L) conducted between April and July 2013. SETTING AND PARTICIPANTS Residents with private wells living in Hernando County, Florida, where nearly 400 unique wells of 1200 tested have been found to have elevated arsenic levels (targeted sampling). INTERVENTION Participants with elevated water arsenic levels were referred to the Florida Safe Water Restoration Program for assistance in obtaining bottled water vouchers and/or POU filters. MAIN OUTCOME MEASURES Creatinine-corrected total urinary arsenic levels and information on water use and consumption behaviors, dietary intake, and other possible exposure sources. RESULTS Total urinary arsenic levels were similar for cases using POU filters or bottled water when compared with controls (geometric means of 7.17 and 7.19 μg/L, respectively). CONCLUSIONS The current practice used by the Florida Safe Water Restoration Program of supplying POU filters or bottled water to households with arsenic levels in private wells between 10 and 50 μg/L appears to be sufficient to protect residents from arsenic exposure through tap water.

Developing a smartphone interface for the Florida Environmental Public Health Tracking Web portal.

As smartphone and tablet devices continue to proliferate, it is becoming increasingly important to tailor information delivery to the mobile device. The Florida Environmental Public Health Tracking Program recognized that the mobile device user needs Web content formatted to smaller screen sizes, simplified data displays, and reduced textual information. The Florida Environmental Public Health Tracking Program developed a smartphone-friendly version of the state Web portal for easier access by mobile device users. The resulting smartphone-friendly portal combines calculated data measures such as inpatient hospitalizations and emergency department visits and presents them grouped by county, along with temporal trend graphs. An abbreviated version of the public health messaging provided on the traditional Web portal is also provided, along with social media connections. As a result of these efforts, the percentage of Web site visitors using an iPhone tripled in just 1 year.

Uncertainty in maternal exposures to ambient PM2.5 and benzene during pregnancy: Sensitivity to exposure estimation decisions.

We investigate uncertainty in estimates of pregnant womens exposure to ambient PM2.5 and benzene derived from central-site monitoring data. Through a study of live births in Florida during 2000-2009, we discuss the selection of spatial and temporal scales of analysis, limiting distances, and aggregation method. We estimate exposure concentrations and classify exposure for a range of alternatives, and compare impacts. Estimated exposure concentrations were most sensitive to the temporal scale of analysis for PM2.5, with similar sensitivity to spatial scale for benzene. Using 1-12 versus 3-8 weeks of gestational age as the exposure window resulted in reclassification of exposure by at least one quartile for up to 37% of mothers for PM2.5 and 27% for benzene. The largest mean absolute differences in concentration resulting from any decision were 0.78 µg/m(3) and 0.44 ppbC, respectively. No bias toward systematically higher or lower estimates was found between choices for any decision.

Lead Exposure and Developmental Disabilities in Preschool-aged Children

Context: Lead is a preventable environmental toxin that has been previously associated with deficits in cognition, academic performance, attention, and behavior in children. Very few studies, however, have examined the relationship between exposure to lead and documented developmental disabilities. Objective: This study examined the relative risk of lead exposure on developmental disabilities in preschool-aged children. Design: A statewide lead surveillance data set containing blood lead level (BLL) was integrated with another statewide data set containing developmental disability classifications for special education placement for preschool-aged children. Participants: The participants were the 85 178 children (average age 2.6 years) whose records in both data sets were able to be linked. Forty-six percent of the participants had an identified developmental disability. Main Outcome Measure: Developmental disability classification served as the main outcome measure. Results: A high BLL, defined as 5 &mgr;g/dL or more, was associated with significantly increased risk for developmental disabilities (risk ratio [RR] = 1.04; 95% CI = 1.01-1.08), particularly intellectual disability (RR = 1.58, 95% CI = 1.10-2.25) and developmental delay (DD; RR = 1.11, 95% CI = 1.06-1.17). Conclusions: The results of this study are consistent with previous research identifying an association between lead exposure and numerous intellectual and educational outcomes and demonstrate that high BLL is associated with meeting eligibility criteria for developmental disabilities in young children. Continued research, surveillance, and prevention efforts are needed to further reduce the negative impacts of lead on individuals and society. Reducing or eliminating lead exposure would improve outcomes for individual children (eg, better academic performance) and reduce the burden to society (eg, lower enrollments in special education systems).

Health Impact Assessments and Extreme Weather—Challenges for Environmental Health

Background: The Florida Department of Health, Environmental Public Health Tracking Program, in collaboration with the Escambia County Health Department and the University of West Florida, used the Health Impact Assessment Framework to examine adverse health outcomes that may be related to an extreme flood event in Pensacola, Florida (Escambia County) during April 29 to May 3, 2014. In this 2014 flood event, portions of Pensacola received more than 15.5 in of rain in a single day. Infrastructure impacts from this extreme event included destroyed bridges and roads and the failure of many sewage lift stations. Objective: To determine whether there were associated increases in injury, illness, and death, data on reportable diseases, hospitalizations, emergency department (ED) visits, and deaths that occurred during the impact period in 2014 were compared with a control period in 2008. Design: We used an ecological design to compare impact and control periods and examined the proportion of hospitalizations, ED visits, and deaths potentially attributable to the extreme flood event. Results: The results of this comparison were mixed, with some Escambia County zip codes showing increased hospitalizations and ED visits, and some zip codes showing a decrease. However, countywide, there were increases in the proportion of both injury- and respiratory-related hospitalizations and ED visits during the impact period. Conclusions: It is challenging to characterize human health impacts from natural disasters such as extreme floods. Still, it is believed that specific policy changes could result in fewer health impacts during future flood events. For example, this study recommended raising the electric panels on lift stations above the flood elevation to keep them operational during extreme rainfall events. For more maps and tables, consult the complete project report available online at http://www.floridatracking.com/HealthTrackFL/document/Escambia_HIA_Report.pdf.