Introduction

The Need for an Assessment of Local Environmental Health Data

Environment is a major determinant of health. Just how “major” depends on how broadly “environmental” is defined: one study estimates that about a quarter of all disease worldwide is due to environmental factors.[1]  Locally, there are clear indications that the Pittsburgh region, once famous for having “cleaned up its act” as one of the most polluted places in the country, has been backsliding over the last decade or two in terms of certain aspects of environmental health, including, for example, air quality and urban sprawl. For example, in a recent comparison of air pollution among the 50 largest U.S. metropolitan areas, Scorecard.org recently ranked Pittsburgh as having the 18^th worst air in the nation,[2] while the Brookings Institute’s “Back to Prosperity” found that, as measured by land urbanized per new household, the Pittsburgh Metropolitan Area is by far the worst sprawling area in the country.[3] Despite these indications, certain key questions remain difficult to answer. For example, what is the current local burden of disease related to various environmental factors? How does this burden of disease in the Pittsburgh region compare to that seen in other parts of the country? How is it changing over time? Do certain communities in our region bear more of this burden than others? How can policy-makers, organizations and communities prioritize local environmental health problems so as to act most effectively to solve them?

Before we can begin to try to answer these questions, or even to know whether they are in fact answerable, it is first advisable to examine the types and quality of existing environmental health data. If available, such data would certainly be useful, for example, in helping make decisions related to:

· Personal actions,

· Planning a research agenda to better understand an issue,

· Design of specific programs to address an issue,

· Policy-making, and

· Funding strategies.

This report, then, is an effort to begin to lay the groundwork for an understanding of the data and data gaps related to local environmental health, so as to allow such decision-making to be better informed by the available data, as well as to prioritize areas where efforts to gather better data are most needed.

Project Purposes and Goals

In order to facilitate effective improvement of environmental health in local communities, we wanted to first holistically examine the available evidence base.  What information do we already have? What other information do we need? These were our questions going into this endeavor.

As we proceeded in attempting to answer them, we realized that several very recent or currently-in-progress reports and endeavors either a) already describe the availability and quality of a particular type of data in great detail, b) include a partial data availability inventory as part of a broader endeavor, or c) plan to conduct similar activities on a larger geographic level. We thus reassessed the project focus, shifting away from covering the finer details of the data to an initial focus on “making sure the left hand is aware of the right,” and providing descriptions of and references to existing works and endeavors.

Our goal in producing this report in its present form is to create the “to do” for a consolidated information inventory and data needs assessment that will serve the following purposes for environmental health researchers, citizens, funders and policy makers in the Pittsburgh region:

· Provide an overview of the sources of environmental health information in one place, along with an understanding of pertinence and interconnectedness of these sources

· Illustrate the large volume of information that is already available, and where much of it can be obtained, to lessen information seeking time and duplication of effort

· Describe some of the major strengths and weaknesses of existing information

· Outline some of the major gaps in information, so that we collectively know where the greatest efforts will be needed to fill them

· Highlight related data compilation, linkage and analysis endeavors, so that organizations may share resources and avoid duplication of efforts

· Illustrate the “real life” connection to environmental health issues via case studies of successful and unsuccessful attempts to obtain and utilize environmental health data for specific purposes

Project Scope

1. What is the scope of environmental health?

As is well-known, health is defined in the World Health Organization’s Constitution as “a state of complete physical, mental and social well-being and not merely the absence of disease or infirmity.”[4]  Also according to the World Health Organization, “environmental health comprises those aspects of human health, including quality of life, that are determined by physical, chemical, biological, social, and psychosocial processes in the environment.  It also refers to the theory and practice of assessing, correcting, controlling, and preventing those factors in the environment that can potentially adversely affect the health of present and future generations.”[5]

It is clear that these WHO definitions of “health” and of “environmental health” are broad. Yet having the full range of environmental health issues on the table is important for a truly community-based approach to environmental health. It is the communities themselves who then help program personnel to define the environmental health issues that are important, to decide which of these issues should be addressed, and to plan how best to address these issues.

2. What aspects of environmental health are covered in this report?

Because of limitations of time and resources, this report has attempted to be comprehensive in outlook, yet strategic in focus. It is recognized that the information contained in this report is not by any means a complete picture of environmental health in the Pittsburgh region. This document is thus presented as a work in progress that we nevertheless hope may serve as a strategic guide for funding research and service programs dedicated to improving local environmental health.

This report covers local data related to the following areas of environmental health:

Outdoor air pollution

Water pollution

Land-based pollution

Chemical hazards

Some problems of the built environment: (e.g., sprawl, “bad neighborhoods”)

Even within these areas of focus, this report, like the available data, does not cover the full set of environmental toxins, pollution sources, exposure pathways, environmentally-related health outcomes or economic considerations.  In addition, within each area the present report has focused on certain subtopics as examples. For example, we focus on the data related to the built environment at the neighborhood level, omitting any discussion of indoor environments.

3. What aspects of environmental health are not covered in this report?

Because of limited space, resources, and time, this report does NOT address the following areas of environmental health:

Indoor air pollution

Poor nutrition

Food poisoning

Abused substances

Biological hazards

Radiation hazards

Mechanical hazards, noise

Problems of geography and climate (e.g., heat- and cold-related illness, natural disasters)

We note here also, although we do not further explore them in this report, the following dimensions of environmental health:

Global-Local Linkages

Certain local issues may have large-scale or global environmental health implications (e.g., transportation as it affects global warming), and certain large-scale or global environmental health issues may have local implications, either presently or in the future (e.g., ozone depletion, global warming, acid rain, loss of biodiversity).

Future Generations

Environmental health is concerned with the health and well-being of future generations as well as present populations.

Social Environments

Unhealthy relationships with other people and the world around us (e.g., isolation, abuse, dependency, alienation, lack of control) as well as negative large-scale socio-cultural, macroeconomic, and political influences (e.g., racism, disenfranchisement, marginalization, exploitation) may clearly lead to mental illness, lowered resistance or predisposition to physical illness, substance abuse, violence, and other health problems.

Geographic Focus

In the present document, our geographic focus is Southwestern Pennsylvania, with the greatest emphasis placed upon Pittsburgh and Allegheny County.  In deciding on the geographic focus of this project, we weighed several factors. On the one hand, it made sense to focus on a smaller area because 1) the Pittsburgh metropolitan area has the greatest population density, and thus the greatest exposure risk, 2) a clear focus on our part would be likely to be more useful for our audience, and 3) different geographies have their own information systems.  On the other hand, it also made sense to view these issues regionally because a) pollution does not stop at municipal or county boundaries, often traveling across several states, b) although rural areas may be more sparsely populated, certain types of exposures may be much higher in these areas, c) economies of scale may be achieved by pooling resources, which is especially important within the current funding environment, d) many organizations build their information systems around reporting requirements, reporting upwards to state and national-level agencies that seek homogeneity of systems across jurisdictions, and e) many local datasets can be obtained only from state and federal-level agencies.   

Note that, while the definitions of the Southwestern Pennsylvania “region” for certain key agencies (the Southwestern Pennsylvania Commission, the Pennsylvania Department of Environmental Protection, and the Pennsylvania Department of Health) fall within the same 12 counties, they are slightly different (see Appendix B: Counties in Definitions of “Region”).  Thus, although we speak generically of Southwestern Pennsylvania, we do not use a steadfast definition of the region in this report.  

Report Organization and Limitations

The organization of this report reflects the idea that interactions between environments and people occur in both directions:

ENVIRONMENTS <===> PEOPLE

The various sources of data relating environmental pollutants to human health outcomes may thus be placed along nodes in the following model:

As a specific example, here is this model as it might apply to data relating mercury from power plants and neuro-cognitive impairment in children:

Despite our adoption of this model as the framework for much of the classification of data in this report, we recognize several of its limitations. For example, the model represents only one greatly simplified chain of events within a complex system. In reality, (1) multiple sources typically release multiple toxins for multiple underlying reasons, (2) a given toxin may produce multiple health outcomes to a varying degree in susceptible populations, and (3) a given health outcome may be caused by multiple toxins. In addition, the model does not take into account a host of other important contexts, conditions, and confounders, such as, for example, environmental and other factors contributing to exposures, or to susceptibility.

If sufficient information and resources were available, a more ideal approach would be to consider the data for a given environmental health issue within an “eco-social” model constructed for that particular issue which would consider its ecological and social contexts. An eco-social model would look, for example, at consumer demands and polluting activities in the contexts of economic and legal issues, pollutant releases in the contexts of cost and technological issues, exposures of susceptible populations in the contexts of environmental justice issues and socio-economic issues, and health outcomes in the contexts of health education, co-morbidities and other factors affecting resistance, and access to health care.

Moreover, this model does not easily apply to many problems of the built environment, let alone most problems of the social environment. For this reason, in our discussion of data related to the built environment and health, we have adopted a different framework for presenting our preliminary findings.

What is more, since our common goal is not only to fully characterize environmental health problems but also to address and begin to solve them, many more dimensions of data are actually needed than those which are reviewed here.  An outline of what the full evidence base might entail for a given environmental health issue is given below in Figure 1. To put what we are doing in perspective, the areas covered by the present report are shown in italics.

Figure 1: Environmental Health: Evaluating the Evidence Base*

(What is known—and what may need to be known—in order to guide efforts to improve a given environmental health issue)

	Demand/ Sources	Agents	Exposures	Health outcomes and causation	Externalities	Relative importance, public perceptions
Understanding the issue
Quantitative knowledge	agent sources	locations in the environment	number of people exposed and levels	1. strength of evidence linking agents to various health effects (e.g., dose-response model) 2. population health impacts and risks from agents, with special consideration of vulnerable groups	externalities (e.g., to people elsewhere, future generations, wildlife & ecosystems)	1. contribution to total burden of disease 2. public awareness and level of concern relative to other issues
Data sources, adequacy and quality of data	Source data (e.g., TRI)	Environmental data (e.g., air  and water monitoring)	Exposure data (e.g., models using  NHAPS)	Health data, tracking of exposures and diseases (e.g. , ASTDR, cancer registry, chronic disease registry)	Measures of externalities (e.g., product life-cycle analysis)	Health data and public opinion surveys
Contexts: Factors influencing sources (e.g., driving forces), agents, exposures, health outcomes, etc. (and strength of evidence).
Trends over time. Comparisons (e.g., with other U.S. cities, MSAs).
Coping with the issue
Available intervention options	decreasing releases (e.g., less consumption and waste, cleaner systems, less toxic agents)	reducing environmental burden (e.g., remediation technologies)	reducing exposure (e.g., CDC Community Guide to Preventive Services)	Improving detection and management of health outcomes (e.g., evidence-based medicine)	reducing externalities (e.g., product design, trade policies)	increasing public awareness and understanding, engaging communities in solutions
Contexts: Factors influencing implementation of interventions
Stakeholder analysis**: level of impact of the issue on various stakeholders, level of interest of various stakeholders in the issue
Industry analysis: groups and agencies currently working on the issue—specific focuses, capacity, needs
Selecting best interventions (e.g.,” proven” interventions, likelihood of success in specific context, relative costs and benefits, strength of evidence)
Selecting best activities and organizations for funding (e.g., SWOT analysis)
Program monitoring and evaluation: measuring performance and evaluation-based planning

*Red italics denote the focus of the current report.

**Stakeholders may include: producers (businesses, workers), consumers, regulators, civic groups, foundations, citizens at large, vulnerable groups, researchers, government agencies, health care providers, payers, and insurers, wildlife and ecosystems, people in other places, future generations.

Once all these limitations are recognized, the cyclical model nevertheless remains useful as a first approximation of an organizing principle for describing much of the universe of available local environmental health data. We therefore present the remainder of this report in the following order:

Consumer Demands and Polluting Activities

Source and Release Monitoring and Emissions Estimates

Environmental Monitoring (Potential Exposure)

Human Exposure

Health Outcomes

Built Environment

Note that built environment factors do not always fit the source-pathway-exposure-outcome model.  Amenities, for example, may fit an “asset availability-awareness-behavior-outcome” model.  This has implications for the types of data that can be collected, and so the Built Environment Section is self-contained.

Depth of Information

Due to its broad scope, this report is intended merely as a starting point for future information collection and discussion.  In preparing this report, as we continued to learn of new organizations, working groups, reports and websites, we realized the following:

· It would be impossible to interview more than a small portion of the pertinent stakeholders and experts.

· Within specific areas of data, there was not enough time or space to include all pertinent expertise, or to learn about it well enough to explain it completely.  For example, some of the individuals with whom we spoke for an hour had spent decades of their career dealing with specific aspects of air quality monitoring, or specific types of diseases.

· As datasets evolve, web links change, and new information becomes available, the value of a “static” report will rapidly diminish.

· A more comprehensive data inventory, including finer detail such as specific variables within datasets, is needed, but will take significantly more effort and is better suited for a more easily updateable and queryable format (e.g., an online database).

For some types of information that we initially hoped to include, we did not receive responses to our initial requests for more in-depth information, and didn’t have time to continue follow-ups.  This may be an indicator of limited staff resources in some agencies, and will likely influence others’ ability to obtain information.  

Additionally, we do not attempt to cover the political and legislative background behind the data collection in great depth, because we didn’t want to speak out of ignorance on complex issues and jeopardize important existing and potential relationships or collaborations.  However, we recognize that this type of context does often impact the availability and quality of data, sometimes leaving individuals within organizations feeling like their hands are tied regarding what they can share, or leaving them with too little funding and staff resources to maintain the quality of data they’d like to. 

We envision this report as the template for what may become a “living document.”  This might, for example, take the form of a Wiki, an online document that allows individuals to make updates to a “shared community document” from any web browser.[6]   Alternately, the information from this document might be added to one of the dynamic online information systems already being designed by local or state organizations.