Asthma and air pollution in the Bronx: Methodological and data considerations in using GIS for environmental justice and health research
Section snippets
Geographic Information Systems (GIS) for environmental health and justice research
GIS and associated spatial analytical techniques have been used extensively to study public health issues in recent years. Uses of GIS include disease mapping, epidemiological inquiries, health services analyses and planning, environmental health and justice analyses, exposure modeling, risk assessments, disease diffusion and clustering studies, health disparities research, and investigations of many other public health issues. Examples of health research using GIS cover a wide range of topics (
The problem of air pollution and asthma in the Bronx
Asthma is extremely prevalent in the Bronx, affecting people of all ages and diminishing their quality of life. In some cases, asthma can cause death, and the asthma death rate in the Bronx (6 per 100,000) is double that of New York City (see Fig. 1). The precise causes of asthma are not known, and there may be a multiplicity of causes. Some of these are thought to be outdoor air pollution, indoor air pollution, pollen, allergies, family history, and behavioral causes such as smoking or
Environmental justice context of the Bronx
The Bronx is home to over 1.3 million people, according to the 2000 census, representing about 17 percent of the city's population (US Department of Commerce, Bureau of the Census, 2000a). Of the five boroughs of New York City, the Bronx is the least affluent, having the lowest mean household income, and the highest percentage of people below the federal poverty levels (30.7 percent, with some communities in the Bronx as high as 46 percent). The Bronx contains the highest percentage of minority
What is the geographic extent of the study (scale), and the spatial resolution (unit of analysis)?
Among the first questions to be answered when using GIS for environmental justice research are “What is the appropriate study area (the scale or geographic extent of the study)?” and “What is the appropriate unit of analysis (the spatial resolution)?” In many cases, the answers to these questions are determined by the availability of data for all possible geographies; the known or probable geographical extent of the problem to be studied; the physical integration, transportation systems,
Data aggregation and administrative boundaries
One of the long-standing issues in many GIS studies is the selection of the type of administrative unit used to aggregate demographic and socio-economic data, and how well that unit represents the community. Most often, researchers use the geographic unit that makes sense in terms of available data, but these boundaries may have little to do with defining the actual or potential impacted community. The paper, “How We Manage is How We Measure,” discusses this problem in detail (Zimmerman, 1994),
Scale and resolution of the study's spatial data sets
The geographic extent of this study is the Bronx, a New York State county and one of the five boroughs of New York City, comprising approximately 42 square miles of land mass. The Bronx was selected as the study area primarily because of its high rates of asthma hospitalizations and high quantities of noxious land uses, and the likelihood of obtaining relatively complete and accurate asthma hospitalization data sets for this area. The Bronx serves as a pilot study for the methods developed for
What is considered an environmental hazard?
In environmental justice research, it is necessary to decide which hazards will be considered as environmental burdens in order to assess if populations are disproportionately affected. Researchers have most often used databases that are publicly available and that track pollution information at the national or state level. The Toxic Release Inventory, maintained by the US Environmental Protection Agency (EPA), is used for this purpose because it is a fairly consistent database and covers the
What pollutants should be investigated?
Previous research has demonstrated that exposure to major air pollutants, including ozone, sulfur dioxide, nitrogen dioxide, and suspended particulate matter, may be associated with asthma prevalence or hospitalization, and many of these studies focused on exposure based on proximity to roadways (Edwards et al., 1994; English et al., 1997; Friedman et al., 2001; Green et al., 2004; Guo et al., 1999; Neutra, 1999; Schwartz et al., 1993; Studnicka et al., 1997; Sunyer and Spix, 1997).
There are
Data quality and data uncertainty issues
A number of data problems and data limitations are encountered with the integration of health data in GIS. A basic data quality issue is data accuracy, which takes two forms: positional accuracy and attribute accuracy. Both have substantial ramifications for the asthma and air pollution study, as discussed further below:
Positional accuracy refers to the nearness of the values describing the position of a real-world object to the object's ‘true’ position. Positional error may be introduced at
Asthma hospitalization cases—the use of record-level data
The basic data sets needed to conduct this analysis were asthma hospitalization records; the location of and emissions information about the polluting facilities or land uses; land use and zoning data; and demographic and socio-economic information. Due to issues of patient confidentiality, the patient-related data is typically the most difficult to obtain, especially at a fine level of spatial resolution.
Many of the previous studies relied on survey questionnaires and self- or parent-reported
Geo-referencing—mapping the locations of asthma hospitalization cases
Geo-coding, a type of geo-referencing, is a common function in most GIS applications, and is used to plot on a map the locations of phenomena or events listed in a table. Usually, street addresses listed in a table are matched by the GIS program to a spatial file of street segments, each segment having an address range. The geo-coding program generally places the point at a location mathematically computed and interpolated from the street segment file, and not necessarily at the exact location
Data limitations
A major drawback to the data used in this analysis is that asthma hospitalization records only provide instances of hospital admissions, and do not reflect the magnitude of the asthma problem. Actual cases of asthma or even emergency room visits due to severe asthma problems are not tracked consistently by doctors or hospitals, and there is no state-wide reporting of asthma and therefore no centralized asthma database. People suffering from asthma may be seen by a private doctor, a clinic, a
How is exposure potential determined?
Two commonly used methods of determining exposure potential in environmental justice research are the spatial coincidence method and proximity analysis. The spatial coincidence method entails examining and characterizing the populations within a certain geographic unit (such as a census tract, ZIP Code, or county) and noting whether or not a polluting facility exists in that unit. Populations within a unit containing a polluting facility are considered to be impacted by it, and thus potentially
GIS methods for proximity analysis
This study accounts for exposure to air pollution burdens of these noxious land uses by creating buffer zones around the TRI facilities and other listed major stationary point sources as a proxy for areas of impact. All TRI facilities and many of the listed major stationary point sources are located within industrial zones, and these zones are also usually the home of the smaller polluters. Through visual inspection of the buffers and land use and zoning data, it was determined that, in most
Results of proximity analysis
The most noticeable visual aspect of the buffers that were created around major polluting land uses is the extent of the Bronx that is covered. Approximately 66 percent of the Bronx's land mass falls within the buffers (excluding major parkland and water bodies). Since, in this study, the buffers represent those areas most impacted by air pollution, a majority of the Bronx population may be exposed. According to calculations based on the areal weighting script, 88 percent of the people within
Integration of air dispersion modeling and GIS
Exposure potential can also be estimated using a plume buffer rather than a circular or linear buffer. A plume buffer is constructed based on results from a model that estimates the extent and direction of the pollutant dispersion, as well as pollutant concentration levels. While this obviously yields more realistic results than a simple circular or linear buffer, there are several problems in using air dispersion models. The first, most difficult to solve, is the lack of readily obtainable
The need to build better databases and analytical methods
Specific data limitations were discussed above. In general, a major issue with environmental justice and health research is the difficulty in obtaining data at a resolution and accuracy level sufficient to reliably demonstrate the connections between environmental conditions and health outcomes. This is the case for both the health and the environmental data.
The lack of accessibility of health data is a significant drawback. Very few people have access to individual level health records, which
Making the connection between environmental justice and environmental health
As discussed above, this analysis found that people within the buffers were not only much more likely to be hospitalized for asthma than those living outside the buffers, but also more likely to be minority and poor than those outside the buffers. Previous research has suggested that socio-economic status itself plays a role in diseases and deaths associated with air pollution (O’Neill et al., 2003; Schulz et al., 2002). It is possible that high asthma hospitalization rates reflect minority and
Acknowledgements
This research was supported in part by the Albert Einstein College of Medicine (AECOM) and Montefiore Medical Center of the Bronx; the National Oceanic and Atmospheric Administration Cooperative Center for Remote Sensing Science and Technology (NOAA-CREST); the Professional Staff Congress-City University of New York (PSC-CUNY) Research Award; and the George N. Shuster Fellowship.
Thanks are due to Dr. Hal Strelnick, Director of the Institute for Community and Collaborative Health at Albert
References (102)
- et al.
Defining localities for health planning: a GIS approach
Social Science and Medicine
(1996) - et al.
Reassessment of lead exposures in New Jersey using GIS technology
Environmental Research
(1992) - et al.
Concentrations of submicrometer particles from vehicle emissions near a major road
Atmospheric Environment
(2000) - et al.
Individuals living in areas with high background radon: a GIS method to identify populations at risk
Computer Methods and Programs in Biomedicine
(1997) - et al.
Measuring access to primary medical care: some examples of the use of geographical information systems
Health and Place
(1998) - et al.
Motorvehicle exhaust and chronic respiratory symptoms in children living near freeways
Environmental Research
(1997) - et al.
Environmental equity: the demographics of dumping
Demography
(1994) - et al.
Geographic epidemiology of gonorrhea in Baltimore, Maryland, using a geographic information system
American Journal of Epidemiology
(1998) - et al.
Is there environmental racism? The demographics of hazardous waste in Los Angeles County
Social Science Quarterly
(1997) - et al.
Towards environmental justice: spatial equity in Ohio and Cleveland
Annals of the Association of American Geographers
(1995)