Citation
Cancer incidence patterns and environmental contamination in north Casper, Wyoming

Material Information

Title:
Cancer incidence patterns and environmental contamination in north Casper, Wyoming integrating qualitative and quantitative methods and geographic information systems (GIS)
Creator:
Tajik, Mansoureh
Publication Date:
Language:
English
Physical Description:
xiii, 196 leaves : illustrations, maps ; 28 cm

Subjects

Subjects / Keywords:
Cancer -- Wyoming -- Casper ( lcsh )
Tetrachloroethylene -- Environmental aspects -- Wyoming -- Casper ( lcsh )
Groundwater -- Pollution -- Health aspects -- Wyoming -- Casper ( lcsh )
Groundwater -- Pollution -- Health aspects -- Public opinion -- Wyoming -- Casper ( lcsh )
Cancer -- Public opinion -- Wyoming -- Casper ( lcsh )
Public opinion -- Wyoming -- Casper ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 185-196).
General Note:
Department of Health and Behavioral Sciences
Statement of Responsibility:
by Mansoureh Tajik.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
54850936 ( OCLC )
ocm54850936
Classification:
RC277.W8 T355 2003 ( lcc )

Full Text
CANCER INCIDENCE PATTERNS AND ENVIRONMENTAL
CONTAMINATION IN NORTH CASPER, WYOMING: INTEGRATING
QUALITATIVE AND QUANTITATIVE METHODS AND GEOGRAPHIC
INFORMATION SYSTEMS (GIS)
by
Mansoureh Tajik
B.S., University of Texas at Arlington
M.S., University of Colorado at Denver
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Health and Behavioral Sciences
2003
l
un


This thesis for the Doctor of Philosophy
degree by
Mansoureh Tajik
has been approved
by
Deborah Thomas
April Z8,2oq3
Date


Tajik, Mansoureh (Ph.D., Health and Behavioral Sciences)
Cancer Incidence Patterns and Environmental Contamination in North
Casper, Wyoming: Integrating Qualitative and Quantitative Methods and
Geographic Information Systems (GIS)
Thesis directed by Adjunct Associate Professor Karen Gottlieb
ABSTRACT
The purpose of this study is to a) determine if the geographic patterns
of specific cancer types, perceived by community members as elevated,
suggest a spatial relationship to two known tetrachloroethylene (PCE)
contamination plumes in North Casper, Wyoming using Geographic
Information Systems (GIS) and b) identify factors that contribute to
community discontent with government agencies in their response to
citizens' health concerns via a qualitative, bounded case study. Quantitative
data included cancer incidence data (1980 -1999) from the Wyoming
Cancer Surveillance Program, PCE groundwater plume data from the
Environmental Protection Agency, and 1990 and 2000 US census data.
Qualitative data included community member testimonials, observations,
group discussions, interviews, and government correspondence.
The observed patterns of cancer distribution in North Casper suggest
a statistically significant departure from the expected patterns based on the
city of Casper and national rates. People in North Casper experience higher
cancer incidence, two to three times higher, than those in other areas in
Casper. A direct correlation of these patterns and the PCE plumes could not
be established. However, it does appear two factors may have contributed
ni


to the higher than expected rates of cancers. These are the geographic
location in general and the poverty status of the population. Qualitative
analysis reveals the most pronounced characteristic of the interaction
between the community members and the government agencies is lack of
trust on the part of the community. Several themes emerged that explain the
community's discontent: delay in response time, perceived cover up and
lack of transparency, perception of pro-industry stance, vague and
fragmented communication, and perceived unfair treatment.
The present study shows a research design integrating qualitative,
quantitative, and GIS approaches can increase the power of a study to
advance our understanding of the relationship between the environment and
human health. We how know more about the dynamics of health,
environment, and the community in North Casper than we did before.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Signed
Karen Gottlieb
IV


ACKNOWLEDGMENT
First and foremost, I would like to thank my dissertation committee members
for their guidance and continuous assistance throughout this process.
Karen, I feel honored and my educational experience ever so enriched for
having you as my mentor. Sometimes words cannot do justice; this is one of
those times. Thank you.
Deb, you have always known when and how to say the right thing and kept
reminding me of the light at the end of the tunnel despite numerous times I
argued with you that it was a train! Thank you for sticking with me
throughout this process.
Steve (Koester), it has always been a pleasure to get the "inside" scoop
from you on important issues seasoned with a special wit that is very unique
to you. Thank you for agreeing to be a part of this process so late in the
game.
Steve (Sain), your door was always open and you were very keen in
providing me with your expertise and guidance with a sweet smile at all time.
Thank you!
David (Tracer), you were there throughout my journey and provided me with
some of the most insightful comments. I wish that logistics would have
allowed you to have been part of the final stage as well. I appreciate all your
help and do thank you.
I would also like to acknowledge and thank the following people and
organizations for their support and contributions: John Wyckoff, Ph.D. from
the Department of Geography at the University of Colorado at Denver for his
contributions in the early stages of research; the Department of Geography
at the University of Colorado at Denver for the Global Change and
Environmental Quality Grant; Jim Ruttenbur, M.D., Ph.D. with the University
of Colorado Health Sciences Center for introducing me to this project; Leslie
Kaas with the Law Fund of the Rockies for all her help and support; Judy
Brockhouse, MPH, at the Wyoming Department of Health Cancer
Surveillance Program for the cancer data; Mike H. Jun and Liz Hepp at the
City of Casper, Wyoming for providing the base map layer; and the North


Casper Citizens Outreach members, Gloria Stewart, Nita Lowndes, and
Richard Innes for their contribution and moral support. My most heartfelt
appreciation goes to citizens of North Casper whose lives and stories
touched my heart.
Last, but not least, I would like to thank an exceptional lady, Chris Pon.
Chris, your words (and deeds) of encouragement meant the world to me.
Thank you for being the kind and thoughtful person that you are.


TABLE OF CONTENTS
Figures ........................................................xi
Tables..........................................................xii
CHAPTER
1. INTRODUCTION
A. Background............................................ 1
B. Dissertation Layout....................................3
C. The Problem.......................................... 4
D. Goals and Specific Aims ...............................5
E. Overview of Research Methods ..........................6
2. REVIEW OF THE LITERATURE
A. Introduction....................................... 10
B. Community-Level Environmental Health Concerns.........11
Times Beach, Missouri (1971)........................12
Love Canal, New York (1979).........................13
Three-Mile Island, Pennsylvania (1979)..............14
Woburn, Massachusetts (1986)........................15
Globeville Neighborhood, Denver, Colorado (1993)....16
C. About Tetrachloroethylene.............................17
PCE Toxicity........................................19
PCE Carcinogenicity............................... 20
D. Epidemiology and Disease Cluster Analysis.............23
vii


Successful Disease Cluster Studies ...................24
Disease Cluster Investigations................... ....26
E. Application of GIS in Environmental and Public Health....28
GIS and Public Health Examples........................31
F. Spatial and Temporal Statistical Methods in Health ......37
Summary...............................................42
G. Environmental Equity.....................................42
H. Social Action Model in Community Organization............44
I. Qualitative Methods......................................49
Triangulation ........................................50
Case Study ...........................................50
J. Summary .................................................52
3. THE PEOPLE AND THE ENVIRONMENT
A. Introduction ........................................... 53
B. Socio-demographic Characteristics........................55
Total Population Distribution and Ratios..............55
Sex Ratio ............................................61
Age Distribution......................................63
Race ..... ...... ..:...,..........................69
Poverty Status........................................74
C. Contamination History................................... 76
D. Summary..................................................78
4. QUALITATIVE ANALYSIS
A. Introduction.............................................80
B. Methods ............................................... 81
C. Thematic Analysis Results............................... 84
viii


Themes Early Period: Information Gathering Phase..85
Themes Action Period............................. 88
D. Discussion ............................................93
Themes...............................................93
Social Action Model Themes...........................93
Other Themes.........................................96
E. Summary................................................98
Aim 1 ...............................................98
Aim 2 ............................................ 99
Aim 3 ....:...:.....................................100
5. QUANTITATIVE ANALYSIS
A. Introduction......................................... 102
B. Methods............................................ 103
Data Sources........................................103
GIS Database Construction ..........................105
Spatial Analysis....................................106
C. Results............................................. 111
Point Patterns .....................................111
Standardized Morbidity Ratios and Relative Risks....118
Spatio-temporal Cluster Analysis....................127
D. Discussion ...........................................128
E. Summary...............................................131
6. CONCLUSION
A. Introduction..........................................133
B. Triangulation: a Validity Model.......................134
C. Triangulation: a Complementarity Model................135
IX


D. Triangulation: a Trigonometry Model.........136
E. Summary.....................................138
APPENDIXES
A. GLOSSARY OF ACRONYMS .........................141
B. SUMMARY OF ENVIRONMENTAL CONTAMINATION
STUblES NORTH CASPER, WYOMING..................144
C. SUMMARY DOCUMENTS USED IN
QUALITATIVE ANALYSIS ......................... 150
D. EXTRAPOLATED POPULATION-AT-RISK BY AGE
CATEGORIES FOR THE CITY OF CASPER AND
TRACTS 2, 3, 4, 6, AND 7.......................166
E. INTERNAL REVIEW BOARD AND
CONFIDENTIALITY FORMS :........................173
BIBLIOGRAPHY..... ................................ 185


FIGURES
Figure
1.1 Study multi-methodology diagram.................................9
2.1 Cluster analysis techniques based on types of cluster..........40
2.2 Community organization and community-building typology.........47
3.1 Map of North Casper, Wyoming, Identifying the plumes and
locations of potential polluters ...........................54
3.2 Population change (in percent) in the city of Casper and five
census tracts from 1990 to 2000.............................58
3.3 Sex ratio in the city of Casper and across five census
tracts (sex ratios are shown as males per 100 females)........62
4.1 Qualitative analysis diagram.................................. 83
5.1 Linear extrapolation of the population of Casper
from Censuses 1990 and 2000.................................108
5.2 Dot map of all cancers in the city of Casper .................112
5.3 Dot map of cancers of brain and central nervous system,
liver, and kidney in the. city of Casper....................113
5.4 Dot map of leukemia (all types) in the city of Casper.........114
5.5 Dot map of all cancers in the Census Tracts 2
North Casper), 3, 4, 6, and 7 ..............................115
5.6 Dot map of cancers of brain and central nervous system,
liver, and kidney in the Census Tracts 2 (North Casper),
3, 4, 6, and 7..............................................116
5.7 Dot map of cancers of leukemia (all types) in the Census
Tracts 2 (North Casper), 3, 4, 6, and 7 ....................117
5.8 Standard morbidity ratios for all cancers comparing the city
of Casper and North Casper, Wyoming.........................120
XI


TABLES
2.1 Tetrachloroethylene (PCE) physical and chemical information....18
2.2 Questions to consider in cluster investigations................27
2.3 Different stages of investigation in environmental health
using Geographic Information Systems (GIS) ....................30
3.1 A comparison of population in the 1990 and 2000 censuses
among five census tracts.......................................57
3.2 Population dynamic across different tracts and the city of
Casper based on 1990 and 2000 census data......................60
3.3 Population counts by age categories based on
the 1990 census data .............. ...........................65
3.4 Population counts by age categories based on the 2000
census data ....................................................66
3.5 Age distribution comparison between North Casper and
the city of Casper (a = 0.05; degrees of freedom = 1)
in 2000.........................................................67
3.6 Age distribution comparison between North Casper and
the city of Casper (a = 0.05; degrees of freedom = 1)
in 1990 ..................................................... 68
3.7 Race distribution in the population in the city of Casper
and the selected census tracts (Census 1990)....................70
3.8 Race distribution in the population in the city of Casper
and the selected census tracts (Census 2000)...................71
3.9 Comparison of racial distribution in population between
the city of Casper and North Casper (Census 1990)
using chi-square statistics (a = 0.05;
degrees of freedom = 1) ........................................72
3.10 Comparison of racial distribution in population between
the city of Casper and North Casper (Census 2000)
using chi-square statistics (a = 0.05;
degrees of freedom = 1) ........................................73
xii


3.11 Comparison of poverty status across the city of Casper
and Census Tracts 2, 3, 4, 6, and 7 in 1989 and 1999
(Censuses 1990 and 2000) .........................................75
5.1 Age-adjusted standardized morbidity ratios for all cancers
in North Casper from 1989 to 1999.........................119
5.2 Standardized morbidity ratios for each age category
in the city of Casper.....................................121
5.3 Standardized morbidity ratios for each age
category in North Casper..................................122
5.4 Age-adjusted standardized morbidity ratios (SMR)
of specific cancers in the city of Casper.................124
5.5 Relative risks of all cancers in North Casper compared
with the city of Casper and Census Tracts 3, 4, 6, and 7..126
5.6 Mantel's product and test statistics for select cancers
in the city of Casper, Wyoming (1980-1999)................128
xiii


CHAPTER 1
INTRODUCTION
A. Background
North Casper is a small community of about 3,000 people located on
the north side of Casper, Wyoming. Two large contaminated groundwater
plumes of tetrachioroethylene (PCE)1 lie beneath ten residential blocks of
North Casper. The contaminated groundwater is a legacy from now closed
industries in the neighborhood. Community members believe there is an
elevated cancer incidence in children and young adults for lymphoma, brain
tumors, and leukemia. They attribute these perceived elevations in cancer
cases to the known contamination in their groundwater and their community.
Yet, health officials do not think the contamination has an adverse effect on
the communitys health. This dissertation will investigate the pattern of
cancer cases and its relation to the PCE contamination and explore the
health concerns of the citizens;
Since 1988, several environmental investigations sponsored by
government agencies and local industries have made it clear that North
Casper has serious chemical contamination problems. What has not been
so clear, however, is whether these contaminations have negatively affected
the health of the community members. In September 1998, in a response to
the citizens petition to conduct a symptom-prevalence health study in North
Casper, the Agency for Toxic Substances and Disease Registry (ATSDR) 1
1Note: refer to Appendix A for a glossary of acronyms.
1


conducted a review of the existing data from the Wyoming Department of
Environmental Quality (WDEQ) and the United States Environmental
Protection Agency (EPA). After the review, ATSDR concluded that the
pollution in the North Casper PCE plume site is not an apparent public
health hazard and a study to determine if residents health has been
affected by exposure to PCE is not recommended (ATSDR, Public
Comment Release, 1998). ATSDR based its determination on the fact North
Caspers primary source of water was the city water. Regular testing had
shown the city water was unaffected by the contamination.
The ATSDRs succinct conclusion, however, did not reassure the
concerned citizens. They became quite skeptical about the process and
developed a distrust of government agencies and the local industries.
Subsequently, the North Casper community began to form various citizen
groups, such as Citizens Outreach, and embarked on looking elsewhere for
an independent study of its health concerns.
I met the members of the Citizens Outreach group for the first time in
December of 1999. After reviewing several boxes of documents, attending
many meetings, conducting informal interviews, exchanging emails and
phone calls, and combining four different databases into one, I learned an
invaluable lesson. There are multiple realities in any given situation and it is
extremely hard, if not impossible, to weave together the whole picture in
any given situation no matter how bounded the picture is within a time
period and a geographic space. It is even harder to establish what some of
those realities are, given the complexities of establishing a possible link
between a particular environmental contamination and adverse health
outcome in addition to assessing the affected communitys perception of its
risk.
2


In this research, I developed a strategy to identify and weave together
some of those realities integrating qualitative, qualitative, and quantitative
methods and Geographic Information Systems (GIS). GIS allows for an
integration of environmental and disease data. Inclusion of the qualitative
component supplemented the quantitative and spatial analyses by providing
the citizen's perspective. This way, I was able to bring focus to my study
goal as well as examine the overall picture that emerged. What follows is
the result of that strategy.
B. Dissertation Layout
This dissertation is organized into six different chapters. The current
chapter, Chapter 1, continues with a statement of the problem; presents the
specific aims of this study, and provides a brief overview of the research
methods. Chapter 2 reviews the relevant literature in areas of:
community-level environmental health concerns,
tetrachloroethylene (PCE) properties, toxicity, and
carcinogenicity,
epidemiology and disease cluster analysis,
disease cluster investigations,
application of GIS in environmental and public health,
spatial and temporal statistical methods in health,
environmental equity,
the social action model in community organization, and
qualitative methods.
Chapter 3 reviews the population characteristics as well as the history of
environmental contamination in North Casper, Wyoming. Chapter 4 contains
the qualitative analysis, methods, results, and discussion. Chapter 5 focuses
3


on the quantitative and spatial analyses of the data and the respective
methods, results, and discussions. The last chapter, Chapter 6, brings
together all components of the study results and provides a conclusion for
the dissertation research and recommendations for further research.
C. The Problem
From 1961 to 1982, the Centers for Disease Control and Prevention
(CDC) investigated 108 reported cancer clusters in 29 states and 5 foreign
countries with the final determination that no clear cause was established for
the reported cases. (NCEH, Pub. No. 02-0594). Since the mid-1980s, no
CDC staff members have been dedicated to working full-time to identify and
investigate cancer clusters. This, however, does not imply a diminishing or
an absence of such cases. State health departments in many states receive
numerous phone calls every year from people worried about too much
cancer in their community, neighborhood, or county. A survey of state
health departments in 1997 reported that approximately 1,100 cluster
investigation requests were made in that year alone. Most requests had
been initiated citizens and no pattern emerged for cancer types or
suspected hazards (Trumbo, 2000).
Of these 1000 pjus cluster reports, very few may receive intense
media exposure and become known nationwide. Some of the high profile
cases include a childhood leukemia cluster in Woburn, Massachusetts
(Lagakos, et al., 1986), a brain tumor cluster in Toms River, New Jersey
(Cohn, 1993), and a brain tumor cluster in Amoco research laboratory
workers in Naperville, Illinois (Cohen, 1999). Conversely, great many other
cases that generate grave concerns within the communities may only
receive local publicity (Henry et al., 2001).
4


Regardless of their degree of visibility, these cases represent citizens
fear and a belief that something in their town, community, school, or
workplace is causing cancer. Most often, investigations of these cases fail to
show a clear causal path between the disease and a particular
environmental factor. So, they are rarely resolved to the satisfaction of the
community. Such outcomes are mainly due to the complex nature of
establishing a relationship between exposure to an environmental hazard
and disease. Advancing our knowledge and understanding the relationship
between the environment and human health, in addition to finding better
ways to communicate with concerned citizens, are therefore of paramount
importance. The North Casper community is a crystal-clear example of a
community with an environmental contamination problem and a perceived
disease cluster.
D. Goals and Specific Aims
The primary goal of this study is twofold. One is to advance our
understanding of relationships between the environment and human health
in a population that has not been previously studied, the community of North
Casper in Wyoming, using Geographic Information Systems and spatial
analysis techniques. The other is to gain insight into how a simple reporting
of a perceived cancer cluster evolves into community dissatisfaction and
distrust of environmental agencies and public health officials. This second
component includes a retroactive look at citizen mobilization based on the
social action model using qualitative methods. The secondary goal of this
research is to provide information to both public health officials and the
communities concerned with perceived cancer clusters about ways to
5


improve understanding of their situation and how to eliminate some of the
difficulties in communication.
The specific aims of this study are:
Aim 1: to describe the context within which concerned citizens
interact with government agencies in a perceived disease cluster
case,
Aim 2: to identify factors contributing to community dissatisfaction
with the official responses to its health concerns,
Aim 3: to determine if there is any evidence of environmental inequity,
and
Aim 4: to investigate if the geographic patterns of specific types of
cancer, which community members perceive as elevated, suggest a
spatial relation to two known tetrachloroethylene (PCE) groundwater
plumes in North Casper, Wyoming, using routinely collected
surveillance program data.
E. Overview of Research Methods
To achieve these aims, I used both qualitative and quantitative
methods. The theoretical framework for the qualitative portion is based on
the social action model in community organization (Rothman & Tropman,
6


1987) within an intrinsic bounded case study2. The objective of the
qualitative part is to identify themes that emerge in the case study regarding
the community members concerns about environmental health and its
interaction with the authorities. The themes are identified based on the five
concepts of participation and relevance, empowerment, critical
consciousness, community competence, and issue selection within the
social action model in community organization. Detailed description of the
model and the relevant concepts are presented in Chapter 2.
Data collected to meet the objectives of the qualitative part were:
official correspondence among community members and relevant health
and environmental agencies, written testimonials from the citizens, group
discussions, and informal interviews and discussions with members of the
citizen group. Important questions explored included:
Were the community members treated as equals by other key players
in the process? (participation and relevance)
Did the community members assume greater power to create desired
changes? (empowerment)
Did the community members and officials engage in dialogue to make
a change? (critical consciousness)
Were the community members and officials able to engage in
effective problem solving? (community competence)
Were the community members able to identify winnable and specific
targets of change? (issue selection)
2 A case study is a qualitative study of a bounded system based on a wide variety of data
sources; an intrinsic case study draws the researcher toward understanding of what is
important about the case itself within its own world, rather than focusing on an issue.
7


For the quantitative analyses, I used cancer case data obtained from
the Wyoming Cancer Surveillance Program (WCSP) database, PCE plume
data from Region VIII EPA, digital maps from the city of Casper, and socio-
demographic data from the US Census Bureau. The plume and cancer data
as well as the digital maps of the city of Casper were all imported into a GIS
database as layers. I analyzed the data using Mantel's Method3 and
ClusterSeer Software4 in conjunction with ArcGIS5. Specific questions
explored with the quantitative methods were:
Exploratory Spatial Analysis Do the geographic patterns of cancer
suggest any relationship with two known PCE plumes in North
Casper, Wyoming?
Cluster Analysis Do the observed patterns of disease distribution
suggest a statistically significant departure from the expected
patterns (i.e., the null distribution) in a particular geographic area?
Figure 1.1 is a diagram of the study multi-methodology.
3 A spatio-temporal method to analyze space-time clustering. It is thoroughly described in
Chapter 5.
4 A statistical software package used to analyze space, time, and space-time relationships.
5 A geographical information systems software package.
8


Figure 1.1. Study multi-methodology diagram
1 What are the ways to improve interaction and communication between government agencies j
and community members in cases of environmental health and perceived disease clusters? |
Is there a spatial relationship between geographic patterns of specific types of cancer-
perceived as elevated by community membersand the geographic distribution of known
environmental contaminations?
9


CHAPTER 2
REVIEW OF THE LITERATURE
A. Introduction
During the 20th century and into the 21st century, the human species
has been facing a new battery of challenges resulting principally from its
own successes in colonizing and dominating the earth. Rampant
industrialization and unsustainable development have made environmental
pollution one of the principal threats to human health. In the past several
decades, there has been an increased awareness and attention to a
possible link between human health and various environmental factors
including those of a chemical nature. Numerous nationwide requests for
disease cluster investigations are just one sign of the broad public concern
about the role anthropogenic6 environmental factors play in the development
of various chronic diseases such as cancers.
In this chapter, I review the history of the relationship between human
health and the environment focusing on research on community-level
environmental health concerns and reports of disease clusters. Specifically,
I review several high profile disease cluster7 studies that have presented a
special challenge to the field of environmental epidemiology in linking
chronic exposure to low levels of contaminants-with no immediate
6 Human made or human produced.
7 A disease cluster is a greater than expected number of disease cases in a geographic
area.
10


discernible effects on health-to real (or perceived) adverse health outcomes
in the affected communities.
Although the city of Casper, the location for this study, has been
plagued by several known environmental contaminants, the two
tetrachloroethylene (PCE) plumes in North Casper have made the most
significant contribution to the communitys concerns. Thus, the third section
of this chapter provides a summary of relevant research in PCE toxicity and
its adverse health effects. The fourth section is devoted to a literature review
on epidemiology and disease cluster studies. I review those studies that
have been successful in establishing a link between exposures and health
outcomes. The section that follows contains GIS application in the area of
environmental and public health. Some research examples and methods are
discussed briefly. In the two final sections, I first discuss the literature
pertaining to environmental equity (also known as environmental justice)
followed by a discussion regarding the social action model and its
application in community organization as well as the appropriate qualitative
methods used in this study.
B. Community-Level Environmental Health Concerns
As was pointed out earlier, from 1961 to 1982, the CDC investigated
108 reported cancer clusters in 29 states and 5 foreign countries with the
final determination that no clear cause was established for the reported
cases (NCEH, Pub. No. 02-0594). However, as the number of investigations
increased, so did the number of factors that might explain them. In most
cases, identifying one single, discernible external cause was not possible. A
1997 survey of state health departments reported approximately 1,100
cluster investigation requests were made in that year alone. Most requests
11


had been initiated by citizens and no pattern emerged for cancer types or
suspected hazards (Trumbo, 2000). Some of the most publicized cluster
cases that have received nationwide and local notoriety in the past are
summarized here.
Times Beach. Missouri (1971)
In 1971, thousands of gallons of dust-controlling oil, made out of the
chemical wastes at a Missouri plant that produced Agent Orange8, were
sprayed on the unpaved roads of Times Beach, a small town in eastern
Missouri. A few days after the oil was sprayed, animals began dying and
later children became ill. By 1981, many horses, and several dogs, cats,
chickens, rodents, and birds had perished. Some government scientists
were recommending that people should leave the area.
In November of 1982, soil tests confirmed the presence of high levels
of dioxin in Times Beach. Dioxin is a by-product of Agent Orange production
and one of the most toxic chemical known to science. At that time, the CDC
urged temporary evacuation while they conducted additional tests. Finally, in
February 1983, the federal government decided to use 33 million dollars of
Superfund9 money to purchase Times Beach and relocate the people. The
town was officially closed by April 1985 and the cleanup efforts started (EPA
ROD, 1988; MO Executive Order 92-09, 1992).
In 1988, hundreds of residents of Times Beach filed lawsuits against
Syntex, the chemical company believed to be the source of the
contaminated oil. Their lawsuits were based on the perceived health-related
8 Agent Orange is a herbicide the US military used in Vietnam to destroy foliage cover.
9 A federal trust fund for use in the cleanup of spills or sites containing hazardous waste
posing significant threat to the public health or the environment (also known as the
hazardous substance Superfund).
12


effects from the dioxin. Although a Missouri jury awarded 1.5 million dollars
to the family of a trucker who died of cancer as a result of dioxin at his St.
Louis workplace, the Times Beach residents were not successful. A St.
Louis circuit court jury rejected their claims because there was not enough
medical evidence to support the residents' claims of health-related injuries.
Health and dioxin pollution issues are still a concern for Times Beach
residents who moved back (Callahan, 1989; Chemical & Engineering News,
1997).
Love Canal. New York (19791
Perhaps the best known, low level, chronic exposure to hazardous
chemicals is Love Canal. In the 1940s, Love Canal was an open pit used as
a dumping facility for over 40,000 tons of toxic waste. The site was
ultimately filled in and sold to the local Board of Education who built a school
there. In 1977, the site began leaking, forcing the evacuation of more than
seven hundred families (Kolata, 1980). Love Canal became a national
symbol of poor toxic waste management, focusing media attention on
abandoned waste sites throughout the United States. The media attention
combined with the growing environmental movement forced a legislative
response from the US Congress. By 1980, Congress had passed landmark
legislation authorizing the establishment of Superfund to finance the cleanup
of hazardous abandoned sites.
Several health studies were conducted to address the residents
concerns. Some studies found increased incidences of low birth weight
babies (Paigen et al., 1985; Vianna & Polan, 1984), and significantly more
seizures, learning problems, hyperactivity, eye irritation, skin rashes,
abdominal pain, and incontinence among children living in Love Canal
13


homes (Paigen et al., 1985). Other studies found no significant increase in
the incidence rates of premature births (Paigen & Goldman, 1987), or any
specific cancers, including liver, lymphoma, or leukemia (Janerich et al.,
1981).
Three-Mile Island. Pennsylvania (19791
The accident at the Three Mile Island Unit 2 (TMI-2) nuclear power
plant near Middletown, Pennsylvania, on March 28,1979, is considered to
be one of the most serious in U.S. commercial nuclear power plant
operating history. The accident caused concerns about the possibility of
radiation-induced health effects, principally cancer, especially in the area
surrounding the plant. For 18 years, the Pennsylvania Department of Health
maintained a registry of more than 30,000 people who lived within five miles
of Three Mile Island at the time of the accident because of these concerns
(Hatch et al., 1991). The state's registry was discontinued in June, 1997
without any evidence of unusual health trends in the area (NEI, 2000; TMI,
2001).
After the accident, thousands of environmental air, water, milk,
vegetation, soil, and foodstuff samples were collected by various groups
monitoring the area. Very low levels of radionuclides10 were attributed to the
releases from the accident. However, comprehensive investigations and
assessments by several well-respected organizations concluded
in spite of serious damage to the reactor, most of the radiation was
contained and the actual release had negligible effects on the physical
health of individuals or the environment (NRC, 1979a; NRC, 1979b). Yet, in
10 A radionuclide is an unstable form of an element that undergoes radioactive decay and
can be carcinogenic.
14


1997, Wing et al. analyzed data from the area nearest the Three Mile Island
nuclear installation, and showed elevated cancer incidence rates five years
after the 1979 accident. The researchers refuted earlier assumptions that
low-level radioactive emissions from the accident were too minute to
produce observable effects. Their analysis showed excess cases for all
cancers combined, lung cancers, and leukemia.
Woburn. Massachusetts (1986)
Woburn, Massachusetts, is a. working-class, lower-middle income
town 12 miles north of Boston. In the early 1970s, the residents feared their
children were contracting leukemia at an exceedingly high rate (Brown,
1993). In 1979, two of eight water wells (Wells G and H) serving the
residents of Woburn were found to be polluted with trichloroethylene (TCE).
TCE is a nonflammable, colorless liquid that is used mainly as a solvent to
remove grease from metal parts, but it is also an ingredient in adhesives,
paint removers, typewriter correction fluids, and spot removers. Its effect on
health ranges from headaches, lung irritation, and dizziness to impaired
heart function and unconsciousness, to nerve, kidney, and liver damage to
death (ATSDR ToxFAQs 79-01-6, 1997).
Subsequent to the discovery of the pollution, Lagakos et al. (1986)
carried out a study to determine if the elevated rates of leukemia in children
could be correlated to well water use from the contaminated wells. They
estimated exposures (to water from Wells G and H) and calculated observed
versus expected rates of disease. They found statistically significant
increases of leukemia and several other disorders as a function of
increasing exposure to the contaminated water. The researchers found no
difference in leukemia rates in children from West Woburn, who had never
15


been exposed to water from Wells G and H, compared to the leukemia rates
in children who lived in East Woburn before the opening of the wells. In
another study on the Woburn disease cluster, the childhood leukemia rate
was found to be fourfold higher than the national average (Durant et al.,
1995).
Globeville Neighborhood. Denver. Colorado (1993)
Globeville is a mixed of industrial and residential area that has been
i
home to metal smelting and refining operations since 1886. An
environmental survey showed that the smelting and refining operations
produced high levels of chemical contamination at the plant site and the
surrounding neighborhood. An EPA-based risk assessment conducted
jointly by a Globevjlle refinery and the Colorado Department of Public Health
and Environment (CDPHE) (1989) showed that children were at an
increased risk for lead, arsenic, and cadmium exposure through several
media. A federally funded exposure study conducted by the CDPHE found
children living in Globeville had slightly higher blood lead levels compared to
neighboring communities, but there was no clear association with any of the
environmental variables (Gottlieb et al., 1993; Gottlieb & Koehler, 1994). As
a grassroots response, the community members filed a lawsuit against a
Globeville refinery and were awarded property, but not health, damages
(CDPHE, 1993).
The above cases are just some examples illustrating the complexity
of the relationship between low-level exposure to various contaminants in
the environment and the health of the affected communities. They also
demonstrate the intricacies of the communities' concerns about, and their
16


involvement with, their environment from economic, social, legal, and
political perspectives.
C. About Tetrachloroethvlene
Tetrachloroethylene (PCE), also known as perchloroethylene and
ethylene tetrachloride, is a nonflammable, colorless, heavy liquid at room
temperature with a sweet, chloroform-like odor (HSDB, 1993). It is relatively
insoluble in water and volatile. It is the most widely used chemical in the
textile industry for dry-cleaning, processing, and finishing fabrics (HSDB,
1993). It is also used industrially for degreasing of metals and as a chemical
intermediate in the synthesis of fluorocarbons. Electric transformers contain
PCE as an insulating fluid and cooling gas. Chemical and physical
information about PCE is summarized in Table 2.1.
17


Table 2.1. Tetrachloroethvlene (PCE) physical and chemical information
Characteristic/Property Information Reference
Chemical Name T etrachloroethylene
Chemical Formula C2CI4 HSDB 1993
Chemical Structure Cl Cl \ ./ C=C HSDB 1993
/ \ Cl Cl
Molecular Weight 165.83 HSDB 1993
Color Clear, Colorless Liquid HSDB 1993
Odor Description Ethereal, Chloroform-like, HSDB 1993
Mildly Sweet
Odor Threshold (in air) 47 ppm AIHA 1989
Physical State Liquid at Room Temperature HSDB 1993
EPA limit (in drinking 0.001 ppm EPA 1985
water)
18


PCE Toxicity
PCE is known to be an eye, skin, and respiratory irritant. However,
the most sensitive endpoint of PCE toxicity is the central nervous system
(Calabrese & Kenyon, 1991). Cardiac sensitization and arrhythmias have
been reported following acute exposure to high concentrations of PCE
(Reprotext, 1999). Pulmonary edema and coma followed a single exposure
to an unknown concentration of PCE (Patel et al., 1977); hepatic necrosis
and renal failure were observed following inhalation exposure (Gosselin et
al., 1984); and symptoms such as tiredness, weakness, and nausea and
vomiting were associated with acute exposure to lower levels of PCE
(Reichert, 1983).
In a study by Rowe et al. (1952), four human volunteers exposed to
206-235 ppm PCE for two hours acclimatized to the odor within minutes and
they all reported eye irritation and congestion of the frontal sinuses after
twenty to thirty minutes of exposure. Two of the four test subjects
experienced dizziness. A separate group of four subjects exposed to 280
ppm of PCE for two hours reported lightheaidedness and one subject
reported nausea.
Human subjects exposed to 100 ppm of PCE for seven hours
exhibited central nervous system effects as indicated by an abnormal
modified Romberg test11 and symptoms including headache and light-
headedness after the first three hours of exposure (Stewart et al., 1970).
Subjects exposed for seven hours per day for five days reported decreased
odor perception of PCE over the course of each exposure.
Mild and transient hepatitis was diagnosed in a worker found
unconscious following a thirty-minute exposure to an unknown concentration 11
11 The Romberg test is a test of position sense.
19


of PCE (Stewart, 1969). Elevated serum enzymes (an indication of impaired
liver function) were observed in a worker rendered semi-comatose by
exposure to unknown levels of PCE for three hours (Stewart et al., 1961).
Persons with preexisting skin, eye, respiratory, heart, liver, kidney,
skin, or neurological conditions may be more sensitive to the effects of PCE
exposure. Individuals with hypertension may be at increased risk of
exhibiting elevated blood pressure following exposure to PCE (Reprotext,
1999).
The PCE studies reviewed above all involve acute exposure to PCE
(i.e., exposure marked by short duration and to very high concentrations) as
might be found in an occupational accident. Much less is known about
chronic exposure to PCE (i.e., exposure marked by long duration and to
very low levels) as might be found in a community setting with a hazardous
waste site nearby, or situated on the top of a groundwater plume. However,
one study examining chronic exposure to PCE is a case-control study
among 247 women employed in dry cleaning operations. The results
indicate an increased risk of spontaneous abortion resulting from chronic
PCE exposure (Kyronen et al., 1989).
PCE Carcinogenicity12
Blair et al. (1979) analyzed the causes of death for 330 deceased
laundry and dry cleaning workers by the proportionate mortality method to
make a preliminary determination as to whether a potential health hazard
exists for workers chronically exposed to dry cleaning solventscarbon
tetrachloride, trichloroethylene (TCE), and PCE. They found an increased
12 Cancer causing properties.
20


risk for malignant neoplasms, primarily an excess of lung and cervical
cancer and slight excesses of leukemia and liver cancer.
According to a study conducted by Byers et al. (1988), an unusually
high incidence of leukemia and recurrent infections were noted in children
exposed in utero to a domestic water supply contaminated with industrial
solvents including TCE, PCE and 1,2-transdichloroethylene. Medical and
laboratory investigations with an emphasis on the immunological system
were carried out on 28 family members of leukemia patients to determine if
the relatives displayed symptoms associated with acute or chronic exposure
to these chlorinated hydrocarbons. The principal organ systems affected
were neurological, immunological, and cardiac systems. Damage to these
systems was found ih all subjects by history, physical, and laboratory
parameters. Damage to the immunological system was manifest by altered
ratios of T lymphocyte subpopulations, increased incidence of auto-
antibodies, increased infections, and recurrent rashes.
To investigate whether exposure to TCE, PCE, or 1,1,1-
trichloroethane increased carcinogenic risk, a cohort of 2,500 male and
1,924 female workers monitored for occupational exposure to these agents,
was followed for cancer incidences from 1967 to 1992 in a Finnish study
(Anttila et al., 1995)1 The overall cancer incidences within the cohort were
similar to those of the general Finnish population. However, there was an
excess of cancers of the cervix uteri13 arid lymphohematopoietic tissues. An
excess of pancreatic cancer and non-Hodgkins lymphoma (NHL) was seen
after ten years from the first personal measurement. The overall cancer
incidence increased in a 20-year follow-up period for those exposed to TCE.
There was an excess of cancers of the stomach, liver, prostate, and
13 Cervix uteri is neck of the uterus (the lower and narrowed end of the uterus).
21


lymphohematopoietic tissues combined. Workers exposed to 1,1,1-
trichloroethane had increased risk of multiple myeloma and cancer of the
nervous system. The study provided support to the hypothesis that TCE and
other halogenated hydrocarbons (e.g., PCE) are carcinogenic for the liver
and lymphohematopoietic tissues, especially for NHL occurrence. The study
also documented an excess of cancers of the stomach, pancreas, cervix
uteri, prostate, and the nervous system among workers exposed to solvents.
A study of drinking water contamination and leukemia and NHL
incidences from 1979 to 1987 found statistically significant results (Cohn et
al., 1994). A comparison of incidences in towns ranked in the highest TCE
stratum (>5 p.g/1) to towns without detectable TCE yielded an age-adjusted
rate ratio for total leukemia among females Of 1.43 (95% Cl141.07-1.90). For
females under twenty years of age, the rate ratio for acute lymphocytic
leukemia was 3.26 (95% Cl 1.27-8.15). Elevated rate ratios were observed
for chronic myelogenous leukemia among females and for chronic
lymphocytic leukemia among males and females. NHL incidence among
women was also associated with the highest TCE stratum (rate ratio = 1.36;
95% Cl 1.08-1.70). For diffuse large cell NHL and non-Burkitt's high-grade
NHL among females, the rate ratios were 1.66 (95% Cl 1.07-2.59) and 3.17
(95% Cl 1.23-8.18), respectively. Among males, the rate ratios were 1.59
(95% Cl 1.04-2.43) and 1.92 (95% Cl 0.54-6.81) for diffuse large cell NHL
and non-Burkitt's high-grade NHL respectively. PCE exposure was
associated with a higher incidence of non-Burkitt's high-grade NHL among
females, but co-linearity with TCE made it difficult to assess relative
influences. The results suggest a link between TCE/PCE and leukemia/ NHL
incidence (Cohn et al., 1994).
14 Cl, or Confidence Interval, that does not include a value of one is considered statistically
significant.
22


Although most of these studies are based on either chronic
occupational exposure or acute exposure, they provide a benchmark to
examine the North Casper situation if the residents have been chronically
exposed to low levels of PCE either by inhalation or via well water use. The
WDEQ documents indicate North Casper residents are on the city water,
which is not contaminated. Based on the 1990 census, however,
approximately nineteen well waters were in use in the Census Tract 7, the
census tract covering the entire North Casper. It is unknown as to how many
residents were drinking well water during or before 1990 census, or in what
capacity the well waters were being used by the residents. Furthermore, the
community members believe that there might be an inhalation pathway from
the PCE plumes to their homes. To date, ho studies have investigated
chronic inhalation of PCE in its health effects in residential settings.
D. Epidemiology and Disease Cluster Analysis
Epidemiology is the study of disease in time, space, and population.
A basic premise of epidemiology is that disease distributes predictably in
human populations (Doll, 1981)! A disease cluster should therefore exhibit a
greater rate of disease occurrence than that which is predicted. The CDC
defines a disease cluster as a greater-than-expected number of disease
cases that occur within a group of people in a geographic area over a period
of time (NCEH, 2002).
Some cancer cluster investigations have led researchers to an
identifiable cause. Successful investigations involve clusters of malignancies
that are either extremely rare, involve a signature15 disease, or are not
15 A signature disease is a disease that is unique to a specific exposure.
23


ordinarily seen in a particular group of people. Most cluster investigations,
however, are notoriously difficult, even under the best of circumstances, and
do not lead to a successful conclusion. Even if a cluster is confirmed, it is
extremely hard to link a disease, especially a complex disease such as
cancer, to one single environmental cause or hazard. The perceived higher
cancer incidences in North Casper do not involve either a rare form of
cancer, or a signature disease. This makes the establishment of a link
between the PCE contamination and disease outcome much more difficult.
Successful Disease Cluster Studies
DES. A1997 investigation of a specific and rare type of vaginal
cancer cluster, clear cell adenocarcinoma, in young women was traced to
the use of the drug, diethylstilbestrol (DES), by their mothers during
pregnancy (Herbst & Scully, 1970). DES was prescribed to expectant
women who had a history of repeat miscarriages. This investigation was
successful because the cancer under investigation, clear cell
adenocarcinoma, is a very rare cancer.
Polwinvl Chloride. Four cases of angiosarcoma of the liver were
diagnosed among men employed in the polyvinyl chloride (PVC)
polymerization section of a B.F. Goodrich plant near Louisville, Kentucky
between September 1967 and December 1973. A cluster investigation in
1974 of those cases led to the identification of vinyl chloride as a carcinogen
for workers involved in its manufacture (MMWR, 1974). Again, the
successful cluster investigation was mainly due to the rarity of angiosarcoma
of the liver.
24


HIV. Another successful cluster investigation occurred in 1981 when
American physicians noticed clusters of young homosexual men with a rare
form of cancer, Kaposi's sarcoma,16 together with a rare pneumonia,
Pneumocystis pneumonia (MMWR, 1981). The investigation of the cluster
revealed the syndrome known as Acquired Immune Deficiency Syndrome
(AIDS) and led to the discovery of its cause, the human immunodeficiency
virus (HIV).
Asbestos. Cancers that have a known cause may also lead to
successful cluster investigations. Mesothelioma, an extremely rare cancer of
the lining of the chest wall and lung surface, is almost always associated
with asbestos exposure. In the mid-1980s, an investigation of five
mesothelioma cases in a Native American pueblo of about 2,000 in New
Mexico revealed that silversmiths in the pueblo used asbestos mats to
protect their worktables from the heat of brazing torches and molten metal.
The workers cleaned the mats by rubbing them together which produced a
fine cloud of asbestos fibers. In addition, members of the pueblo whitened
deerskin leggings and moccasins used for traditional dances by rubbing
them with cakes of asbestos. When clumps of asbestos began to form, the
individuals held up the leggings and slapped them, dispensing asbestos
fibers in the air. The pueblo's source of asbestos was discarded steam-pipe
insulation (Driscoll et al., 1988). The above mesothelioma cluster involved
only five cases. However, the disease is so rare and the cause so well
known that even a small number could lead to a successful investigation.
16 Kaposi's sarcoma is an unusual type of cancer that usually occurred in elderly men in the
Mediterranean.
25


Disease Cluster Investigations
Despite difficulties in confirming a disease cluster and subsequently
identifying the cause, it is important to address concerns about disease
clusters in the communities that raise such concerns. Important questions to
consider when searching for an evidence of disease clustering have been
discussed by Aldrich et al. (1993). When investigating a disease cluster, we
need to know, for example, if the disease occurrence is a random
phenomenon in time and/or in space. Further, we need to explore whether
or not the disease pattern changes over time; or, if the observed rates of the
disease show a statistically significant departure from the expected rates.
Finally, we need to know how to interpret the statistical significance.
Regardless of what method is used, considering each of these areas can
inform the investigation of disease clusters. A summary of the questions
along with the measures they examine is provided in Table 2.2. I used
these questions to guide my disease cluster research in North Casper.
26


Table 2.2. Questions to consider in cluster investigations
Measures Questions
General clustering Is this cluster a random phenomenon? Are these cases more closely grouped in space than might result from random placement? Do these cases group closer in time than would occur by random distribution? Do these cases aggregate closer in space and time than might be expected from a random pattern of occurrence?
Specific clustering Do some geographic units (e.g., communities, counties, zip codes) contain more cases than might be expected from random incidence alone? Do some population groups (e.g., communities, ethnic groups) experience more cases than their proportion of the population might indicate would be expected from random incidence?
Incidence patterns Does the pattern of case occurrence indicate that over time a change in disease frequency has occurred?
Morbidity rates Do the observed rates of disease suggest that a statistically significant departure from the expected rates has occurred?
Interpretations How do we interpret statistical significance?
27


E. Application of G1S in Environmental and Public Health
All social and environmental processes occur within a particular
geographic space. Accordingly, location is a fundamental dimension of the
phenomena examined by environmental and social scientists. The first
environmental health study based on establishing a spatial link between
environment and disease was John Snows classic study of a cholera
outbreak in London, England, in 1854. In his study, Snow was able to link
cholera with drinking from specific wells using a simple mapping technique
(Gordis, 2000). Almost 150 years after Snows work, modern computer tools
such as Geographic Information Systems (GIS) have replaced manual
spatial analysis and mapping. Traditional environmental health research that
seeks to establish a relationship between environmental hazards and
disease has been limited to statistical analysis of case-specific data. Little
attention has been given to determinirig the spatial components of disease
etiology through spatial ahalysis.
More and more, researchers assert that GIS can advance
epidemiological science by increasing our understanding of disease
etiology. With GISs ability to explicitly link health outcomes to demographic,
social, and environmental variables and provide a consistent framework, the
limitations of traditional environmental epidemiology could be overcome
(Briggs, 1995; Briggs, 1997; Cliff & Haggett, 1988; Douven, 1995; Loslier,
1995).
Geoffrey Jacquez (1998) refers to GIS as an enabling technology in
disease cluster investigations. He asserts:
My vision of GIS is of an enabling technology that
may lead to fundamental advances in our
understanding of relationships between the
environment and human health. The approach
28


incorporates disease cluster statistics and other
tests for spatial patterns, with the objective of
generating and testing epidemiological hypotheses.
This paradigm is evolving, and its potential is best
understood using the water drop lens as an
historical analog. In the 1600s Anton Van
Leeuwenhoek glimpsed the first images of
microscopic organisms using a water drop lens.
These 'animalcules' were a curiosity, and no one
suspected their role in infection and disease.
Improvement in technology led to the compound
microscope, which in the 1800s enabled Pasteur
and his colleagues to reveal the link between
bacteria and infection. It was the application of the
technology in the context of a systematic approach
that made scientific advances possible, (p. 3)
Douven and Scholten (1995) have identified six stages of
investigation involved in an environmental health study using GIS. These
stages comprise of data collection and preparation of spatial and non-spatial
data, mapping, spatial and temporal exploration of the data, establishment
of association (if any) among various factors, data interpretation, and a
search for possible causal relationship in disease etiology. These stages are
summarized in Table 2.3. In the section that follows, examples of studies
involving GIS in the area of public and environmental health detail some of
the stages.
29


Table 2.3. Different stages of investigation in environmental health using
Geographic Information Systems (GIS)
Stage Process
Stage 1 Collecting and preparing spatial and non-spatial data (called attributes in GIS)
Stage 2 Mapping of data to identify spatial patterns of disease at various geographic scales
Stage 3 Examining spatial and temporal patterns in data using statistical analysis
Stage 4 Establishing if any association between disease and other spatial factors exists
Stage 5 Interpreting the results of data analysis to gain new insights and to identify gaps in knowledge
Stage 6 Searching for possible causal relationship in disease etiology
30


GIS in Public and Environmental Heath
Although GIS application had a slow start in public and environmental
health, many researchers in the field have been using GIS techniques and
its capabilities in various capacities during recent years. Data integration,
manipulation, processing, aggregation, visualization, and spatial analysis
are greatly facilitated with GIS. Such capabilities have allowed for more
precise investigations of environmental and public health problems
especially in focusing the exploration of incidences of various diseases in
smaller geographic units and around specific contamination problems.
Moreover, many investigators have used GIS to integrate large data sets
from various databases into one database in order to determine the patterns
of specific cancers or various diseases and their relations to specific
contaminations.
Becker et al. (1998) conducted an epidemiology study in which they
developed a GIS system linked to the disease surveillance database at the
Baltimore Health Department and used this system to evaluate the
geographic epidemiology of gonorrhea in Baltimore, Maryland, during 1994.
In the GIS system, gonorrhea cases were geocoded by reported addresses
using digitized maps, and were assigned to specific census tracts. Census
tract-specific rates for persons aged fifteen through thirty nine were
calculated using 1990 census data. The incidence rates of census tracts that
had more than thirty cases were then calculated and ranked. Analysis
showed that as radial distance from the core17 areas increased,
17 Core areas are areas with the highest concentration of the disease.
31


geographically defined hyperendemic areas of gonorrhea increased,
incidence rates decreased, and male/female ratio increased.
Gartell and Dunn (1990), for example, used GIS in a study of cancer
of the larynx in northwestern England. They incorporated large data sets
from the national and regional cancer registries and identified cancer cases
by site, sex, and age. The data regarding the location of incinerators also
were incorporated for the public health officials to be able to easily map and
identify the cancer cases based on the location of a particular incinerator as
those cases occur.
In order to investigate residential environmental risk factors for Lyme
disease in Baltimore County, Maryland, several researchers used GIS to
overlay six different land use databases containing fifty-three environmental
variables (Glass et al., 1995). They combined GIS methods with case-
control methods and were able to identify high risks areas for Lyme disease
in Baltimore County. Guthe et al. (1992) used existing computerized data to
predict populations of children at high risk of lead exposure in the Newark,
East Orange, and Irvington areas of New Jersey. Data from various sources
were integrated using Arclnfo GIS software. They constructed maps of the
data from existing databases containing sources of lead exposure (e.g.,
industrial emissions, hazardous waste sites, and traffic volume) and maps
showing census tracts of suspected high lead exposure. This enabled the
residents to identify the population of children at high risk of lead exposure.
In another study, Dutkiewicz et al. (1998) used GIS methods to
analyze the spatial distribution of environmental and health hazards in large
geographic areas, urban and rural regions of forty nine voivodships (i.e.,
administrative units) in Poland. They used four health-related environmental
32


indicators18 and eight environmental-related health indicators19. With the
application of GIS technique, kriging,20 and data interpolation, the
researchers were able to determine the least environmentally healthy
situations in the southwest regions of Poland.
Some researchers have also used exploratory methods in GIS to
explore the rates of various diseases. Kafadar (1996) used map smoothing
technique, an exploratory method in GIS, to explore the rates of melanoma
in white females and white males in the United States. She obtained
melanoma-related mortality data from 1973 through 1987 from the National
Cancer Institute. The data were then adjusted for age and latitude and
smoothed using two different smoothing methods. Maps of the continental
United States were subsequently created showing regions of high rates
even after adjusting for age and latitude. This suggested the possibility of
other variables influencing the mortality rates for melanoma.
A valuable technique in GIS has been map creation for the purpose
of rapid visualization. In a 2001 study (Kistemann et al.), the researchers
conducted a GIS study concerning water-supply structure in the Rhein-Berg
District in North Rhine-Westphalia, Germany. A GIS database consisting of
water supply structure was constructed. This database allows for a rapid
visualization and analysis of drinking-water supply structure and offers
advantages in microbial monitoring of raw and drinking water. It also allows
for the recognition and investigation of incidents and outbreaks that in turn
18 Environmental hazards such as industrial emissions of dusts and gases [S02, NOx, CO,
dust] that have been known to affect health.
19 Health indicators are measures that indicate a health outcome due to exposure to an
environmental hazard such as mortality rate from trauma and chemical poisoning.
20 Kriging is a mapping method that represents the variable under study as a continuous
process, unconstrained by the borders of geographic units and where sudden transitions
between levels of two neighboring areas are avoided. It provides the variance of the
estimated values from the spatial variability of the actual data, for example, a standard error
map, and these error maps can be useful to introduce new sample values for analysis.
33


allows for better outbreak management and timely response to such
outbreaks.
Kohli et al. (2000) used GIS to perform a spatial and temporal
analysis to evaluate the relationship between exposure to background radon
levels and leukemia among children in the county of Ostergotland in
Sweden. Every child born in the county between 1979 and 1992 was
mapped to the property centroid21 coordinates by linking addresses in the
population and property registers. Population maps were then overlaid with
the maps of background radon and exposure at birth and each subsequent
year was quantified as high, normal, low, or unknown. Standardized
morbidity ratios (SMRs) were calculated using the age- and sex-specific
rates for Sweden for the year 1995. SMRs for acute lymphatic leukemia
among children born in high, normal, and low risk areas were 1.43, 1.17,
and 0.25 respectively. The relative risk for the normal risk group and high-
risk group as compared with the low risk group was 4.64 (95% Cl 1.29,
28.26) and 5.67 (95% Cl 1.06, 42.27). The children who were bom in and
stayed in areas where the risk from background radon had been classified
as: low were less likely to develop acute lymphatic leukemia than those bom
in areas classified as normal and high risk..
Earlier, Kohli et al. (1997) had used GIS to identify the population risk
in areas with various radon risks. They utilized the 1991 Swedish annual
population registration records for the county of Ostergotland and the
property register available at the Center Statistical Bureau of Sweden. In
addition, they obtained a paper map of background radon radiation for each
communethe lowest independent self-governing units under the Swedish
systemwithin the province from the commune offices. All communes in the
21A centroid is a point location at the center of a feature used to represent that feature.
34


county assessed the risk for background radiation from radon with the help
of the Swedish Geological Survey Office. The location of each individual
was then mapped by linking the address in the population register to the
property register. The map of the population coverage was then overlaid
with digitized radon maps. This allowed the researchers to quickly and
efficiently identify the populations who were living in high, normal, or low risk
areas with respect to radon exposure based on their commune, sex, and
age. The researchers believe the results of their study could be used by the
Swedish public health officials to take appropriate actions.
In 1993, Wartenberg and colleagues developed a method to identify
and characterize populations with potential for high exposure to magnetic
fields. Their intent was to generate a sample size large enough for
epidemiologic cohort studies, They digitized into GIS each transmission
station along a twenty-kilometer segment of a 230-kV power line in New
Jersey. Using GIS, they chose a hundred-meter buffer on each side of the
transmission line and used demographic data from the US Census and
TIGER22 files to map characteristics of the census blocks that were
contained wholly or partially within the buffer. Using these methods, they
were able to identify populations suitable for a cohort study to assess the
health hazard associated with electromagnetic field exposures.
GIS has also been used to explore environmental equity, or inequity,
in different communities. Sheppard et ai. (1999) used GIS to study
environmental equity in Minneapolis, Minnesota. They combined 1990
census data on population characteristics and the 1995 EPAs Toxic
Release Inventory (TRI) data in conjunction with two commonly used
22 TIGER is a topologically integrated geographic encoding and referencing system; it is a
map data format based on zero, one, and two cells, used by the US Census Bureau in
street level mapping of the United States.
35


proximity measures in GIS. The results enabled the researchers to make a
preliminary assessment of environmental equity, or inequity, in potential
exposure to airborne toxic chemicals.
Spatial and temporal analyses have been greatly facilitated using
GIS. Tempalski and McLafferty (1997), for example, used GIS to analyze
spatial and temporal low birth weight (LBW) trends in New York City during
the 1980s and 1990s. The study was in response to a growing concern
about an alarming increase in LBW babies in New York City, particularly
among African-Americans. Two types of geographic data were used. First
were data about LBW, prenatal care, and drug use obtained from the New
York City Department of Healths vital statistics records by health area; and
second were data from the 1980 and 1990 US Censuses that describes
economic and social characteristics for census tract populations. After
spatial analysis, the researchers were able to identify areas most in need of
services and programs to target and address LBW in New York City
neighborhoods.
Although the above studies illustrate many merits of using GIS in
environmental health, GIS-aided environmental research is not without
drawbacks. Data integration capability of GIS, for example, allows for
incorporation of several large databases into one database. However, the
difficulty arises when data layers from various databases are of different
scales. This scale incompatibility among data layers makes the integration
of data either quite difficult or, at times, impossible. Furthermore,
constructing the GIS database could be a daunting task and a demanding
procedure. At times, as much as eighty percent of the time in a GIS-aided
research is spent in constructing the database.
Other disadvantages in using GIS technology are its high software
costs and extensive time and labor required for the researcher to become
36


skilled in using the various techniques. Despite the disadvantages, GIS
techniques are an improvement over the traditional epidemiological methods
both in modeling the exposure and providing a context for both the
environmental pollutions and their health outcomes.
F. Spatial and Temporal Statistical Methods in Health
Rapid growth during 1990s in spatial analysis methods in general and
of disease cluster statistics in particular has advanced spatial epidemiology.
Concerning spatial epidemiology, Lawson (2001) states:
Since mapped data are spatial in nature, the
application of spatial statistical methods forms a core
part of the subject area. The reason for this lies in
the fact that the study of any data which are
georeferenced may have properties which relate to
the location of the individual data items and also the
surrounding data. (p. 3)
Lawson and Waller (1996) provide a review of appropriate statistical
methodology for the assessment of putative health effects of air pollution or
ionizing radiation sources. In analyzing events around a pollution source,
they recommend defining a fixed window or geographical region with a given
area and all events that occur within that region within a particular time
period to be modeled (i.e., mapped).
In testing for disease clusters, there is an abundance of cluster tests
in the peer-reviewed literature. To determine which test is appropriate, the
main question to ask is: what kind of clustering is hypothesized? A general
classification of those tests is space tests, time tests, and space-time tests.
Space tests identify clusters over particular locations; time tests examine
temporal clustering within a single time series or in several time series
37


simultaneously; and space-time tests are used to detect clustering in space-
time (Sankoh & Becher, 2002).
Kulldorff et al. (1998) used a space-time scan statistic to evaluate a
brain cancer cluster alarm in Los Alamos, New Mexico. Incidence and
population data came from the Surveillance, Epidemiology, and End Results
(SEER) program of the National Cancer Institute, collected by the New
Mexico Tumor Registry. Calculations were performed with saTScan23 using
9,999 Monte Carlo24 replications. The most likely cluster was in the
Albuquerque-Santa Fe area during 1985 through 1989. The p-value of this
cluster, p = 0.074, was not statistically significant at the p < 0.05 level.
Figure 2.1 is a schematic summary of some of the cluster analysis
techniques based on different types of analysis. These analyses are
computationally quite intense and use Monte Carlo randomization
techniques to evaluate observed values. With the help of GIS and newly-
developed spatial analysis software, a computer can quickly randomize
datasets and perform the calculations. These methods, however, are limited
to exploratory data analysis rather than rigorous hypothesis testing. In
addition, spatial locations of cases under analysis often simply serve as a
proxy or indirect estimation for exposure to a risk factor. Also, the precise
date of disease onset is often unavailable and the analyses most often rely
on the date of the diagnosis. Because of these limitations, the methods can
better help identify pattern and generate hypotheses rather than formally
test pre-existing hypotheses.
For the purpose of this research, I used Mantel's method, which is
one of the methods used in spatio-temporal analysis of clusters. This
23 saTScan is a spatial analysis software.
24 Monte Carlo randomization is a special case of randomization, or bootstrap statistics
(also called re-sampling with replacement using an assumed model).
38


selection was based on the fact that individual-level data (with case location
and time of diagnosis) were available and there were uncertainties regarding
the latency period for the development of a specific type of cancer. This
method will be discussed in details in Chapter 5, the quantitative analysis
chapter.
39


Figure 2.1. Cluster analysis techniques based on types of cluster
Temporal
How was disease occurrence measured?
What type of spatial clustering?
1. Global What type
of data?
V1
Individual
Group
Case Count &
population at risk
Disease
Frequency (rate,
risk, prevalence)
2. Local
What type /.
of data?
Individual
\Case Count & rr
------1/ population at risk {_
Group r
r*
Disease
Frequency (rate, ^
risk, prevalence)
Cuzick and Edward's
Ripley's K-function
Besag & Newell's
Moran's I (pop)
Moran's I
Grimson's
Getis-Ord Local G Test
KulldorfTs Scan
SatScan
Anselin's Local Moran
Besag & Newell's
Tmnbull's
Kulldorff Spatial Scan
Diggle's
Bithell's
Lawson & Waller's
Score
40


Figure 2.1. Cluster analysis techniques based on types of cluster (Cont.)
Spatio-temporal
41


Summary
As evident by the above examples, GIS technology and spatial
analysis techniques have made substantial and continuing contributions in
various areas of spatial epidemiology. Some of those areas include
exposure assessment, study population identification, disease map and
atlas construction, and disease surveillance to identify the locations of
possible outbreaks. In addition, since most epidemiological data have a
location and time reference, GIS technology along with spatial and temporal
analyses could prompt timely public health action. Certainly, the limitations
such as scale incompatibility among data layers, high software cost, time-
and labor-intense processes involved in database constructions that were
mentioned earlier still remain as limiting factors in pervasive use of this
innovative technology.
G. Environmental Equity
Although environmental problems affect everyone, a review of the
literature reveals that some communities, especially communities of color
and those at a lower socio-economic status, are likely to be
disproportionately exposed to environmental threats and suffer
disproportionate impacts from environmental degradation. Environmental
equity, also known as environmental justice, refers to the distribution and
effects of environmental problems and the policies and processes that
reduce differences in who bears environmental risks. In a general sense, it
includes concern for disproportionate risk burden placed upon any
population group, as defined by gender, age, income and race (Special
Report, 1995).
42


In recent years, there has been an increased interest in the impact of
environmental pollution on particular segments of the society. A growing
body of research on environmental equity addresses the social distribution
of adverse environmental conditions (Government Accounting Office, 1995).
Various community-based groups, academic researchers, and government
agencies have drawn attention to the fact that minority and/or low-income
populations potentially bear a disproportionately high burden of adverse
health and environmental effects from pollution. Although there have been
studies that show no relationship between the toxic waste sites (based on
the hazard ranking of the EPAs Toxic Release Inventory) and socio-
economic characteristics of the community in which they are located
(Neumann et al., 1998), several studies support present and past
environmental disparities. Among them are: higher incidences of childhood
lead poisoning among African-American children and among lower-income
children (CDC, 1997); higher exposures by people of color to air pollution in
areas that do not meet the Environmental Protection Agency's health-based
standards for air quality (EPA Report, 1992; Tiefenbacher & Hagelman,
1999); higher penalties for violations of federal environmental laws levied in
white communities compared to the communities with predominantly
minority population (National Law Journal, 1992); and a larger proportion of
non-whites and individuals below the poverty line residing in the vicinities of
contaminants,plumes (Chakraborty & Armstrong, 1997).
Bullard and Wright (1993) wrote: The environmental justice
perspective unmasks the ethical and political questions of who gets what,
why, and in what amounts. It has become evident that some of the worst
environmental and health problems in the nation are concentrated in low-
income and ethnically diverse communities. Many of these communities are
faced with a crumbling infrastructure, deteriorating housing, redlining and
43


economic disinvestment, inadequate schools, chronic unemployment, and
high poverty rate (Bullard & Wright, 1993; Been, 1995; Bowman, 1997).
It is worthy to note new industrial development may be located in
poor communities not just because they are poor and lack the resources to
oppose the large corporations, but because poor people may move to areas
with high levels of environmental contamination due to affordability of the
housing in that area. In other words, it is possible that what may have been
a reasonable neighborhood at the time of its construction becomes poor and
underserved because the wealthier people move away from the
contamination. Indeed, a study of the evolution of the environmental equity
in South Carolina (Mitchell et al.; 1999) indicates, while inequitable
situations may exist currently, the facilities came first.
Although large gaps in data still exist, enough is known with sufficient
certainty to move to ensure environmental equity for all segments of society.
In a Presidential Executive Order, issued by President Clinton in 1994, there
was a move to identify and address, as appropriate, disproportionately high
and adverse human health or environmental effects on minority populations
and low-income populations in the United States (EO 12898, 1994).
Therefore, it is a worthwhile effort to examine the possible disparities in
distribution of health outcomes and pollutions based on race and income in
any health study even if to say that no association was observed, or any
possible association remains unknown.
H. Social Action Model in Community Organization
Glanz, Lewis, and Rimer (1997) define a theory as a set of
interrelated concepts, definitions, and propositions that present a systematic
view of events or situations by specifying relations among variables in order
44


to explain and predict the events or situations. Theories, they believe, can
help investigators to "shape the pursuit of answers to why, what, and how"
(p. 22). Thus, theories and models could "explain" a phenomenon and
suggest ways to achieve "change."
There has not been a particular theory that guides research and
examines a community's experience in investigations of community-level
environmental health and disease clusters. Nor has there been a particular
theory that explores "why" and "how" most cluster investigations end in
discontent. However, some existing models of community organization, such
as the social action model, could provide adequate guidelines in examining
a community's experience. The social action model is based on the
assumption that action arises out of meanings that define social reality.
Society defines human and human defines society. Meanings are sustained
only to be continued reaffirmation in everyday action and through interaction
people also transform social meanings. Humans are aware of the meanings
others assign to their own acts. Therefore, the way people construct and
perceive the world as real and routine is important (Silverman, 1971).
Three perspectives come into play when attempting to understand
some of the ideologies held by participants in community-oriented processes
(Rothman, 1972). First is the locality development model which is most
closely identified with a democratic, participatory, grass-roots community
development model. Second is the social planning model that emphasizes
rational, planned, and technically-linked processes, with or without
community involvement. The commitment to democratic processes or
change cannot be assumed. The third and last model, the social action
model, is most closely identified with an activist model of community
development, including changes in power, resources, and decision-making
processes (Rothman & Tropman, 1987). The social action model is both
45


task and process oriented. The model concerns itself with increasing the
community's problem-solving ability and with achieving concrete changes to
redress imbalances of power and privilege between an oppressed or
disadvantaged group and the larger society.
Minkler and Wallerstein (1997) integrate various perspectives and
models, including the social action model in community organization and
community building (Figure 2.2). They define the social action (also known
as Alinsky's) model primarily as a need-based conflict-based model.
46


Figure 2.2. Community organization and community-building typology
(Adapted from Health Behavior and Health Education, Minkler and
Wallerstein, 1997)
Needs-Based
Strengths-Based
Consensus
Conflict
Community Social Action
Development (Alinsky's Model)
Collaboration
Strategies
Grassroots Organizing
Professionally Driven Organizing
Coalitions
Lay Health Workers
Building Community Identity
Political and Legislative Actions
Culturally Relevant Practice
Advocacy
Community Building
and Capacity Building
(Power With)
Empowerment-oriented
Social Action
Challenging Power Over)
I*__________________iL
Community Competence
Leadership Development
Critical Awareness
47


Minkler and Wallerstein (1997) define five key concepts in models
(including social action) of community organizations: participation and
relevance, empowerment, critical consciousness, community competence,
and issue selection. Participation and relevance start with the people and
engage the community members. Based on this concept, community
members create their own agenda based on felt needs, shared power, and
awareness of resources.
Among various definitions of empowerment are: communities
achieving equity (Katz, 1984) and communities having capacity to identify
problems and solutions (Cottrell, 1983). Here, community members assume
greater power or expand their power from within to create desired changes.
In critical consciousness, consciousness is based on reflection and action in
making change and community members engage in dialogue that links root
causes and community actions (Freire, 1973). Community ability to engage
in effective problem solving, or community competence, could translate into
community members working to identify problems, create consensus, and
agree on change strategies to reach goals (Cottrell, 1994). Finally, the
concept of issue selection is defined as identification of winnable and
specific targets of change that unify and build community strength. Within
this concept, community members are thought to be able to identify issues
through community participation and decide on targets as part of a larger
strategy (Miller, 1995).
While reviewing the qualitative data in this study, each of the
constructs in the social action model will guide the identification and
extraction of themes relevant to each construct.
48


1. Qualitative Methods
In order to gain more thorough understanding of the environmental
health concerns at the community level in North Casper, Wyoming,
qualitative methods in the form of a bounded case study were used to
augment the quantitative methods in this research. Qualitative methods
have often been contrasted with quantitative ones in their respective
epistemological foundations. In the past, the two approaches were regarded
as incompatible. However, more and more researchers in health-related
fields have begun to advocate the integration of qualitative and quantitative
methods and the merits of combining the two. While researchers from the
two paradigms tend to stress either the objective (quantitative) or the
subjective (qualitative) end point, they are in the same position: they both
deal with real phenomena,.with social processes, and they both have to
ascribe meaning to their data. Rather than sequencing qualitative and
quantitative research in some way, both approaches are essentially inter-
related, with quantitative research contributing towards the precise
identification of relevant processes, and qualitative research providing the
basis for their thick description (Fielding et al., 2001).
Strauss and Corbin. (1990) broadly define qualitative research as "any
kind of research that produces findings not arrived at by means of statistical
procedures or other means of quantification." Qualitative researchers are
interested in answering those questions and are not prepared to simply
accept the quantitative answers. When placed alongside qualitative
evidence, quantitative evidence is both clear and powerful.
49


Triangulation
Combining qualitative and quantitative methods is almost by definition
an issue of across-method triangulation. It was introduced into the social
sciences by Denzin (1978). Acknowledging that no observation or
interpretations are perfectly repeatable, triangulation serves to clarify
meaning by identifying different ways the phenomenon is being seen (Flick,
1992).
Kelie (1995) identifies various meanings in using triangulation. He
aims to determine which of these meanings is most appropriate for
conceptualizing the combination of qualitative and quantitative methods. He
distinguishes three meanings or models of triangulation: (1) triangulation as
the mutual validation of results obtained on the basis of different methods
(the validity model), (2) triangulation as a means toward obtaining a larger,
more complete picture of the phenomenon under study (the
complementarity model), and (3) triangulation in its original trigonometrical
sense, indicating that a combination of methods is necessary in order to
gain any (not necessarily a fuller) picture of the relevant phenomenon at all
(the trigonometry model). These three models are in turn brought to bear
upon the potential relationships between the results yielded by qualitative
and quantitative methods employed in the same study.
Case Study
One of the major traditions in qualitative research is case study
research. The case selected for study has boundaries, often bounded by
time, place, or an activity (a bounded system). It might be an event, a
process, a program, or several people (Stake, 1995). A qualitative case
study provides an in-depth study of a bounded system based on a diverse
50


array of data collection material (Creswell, 1998). Its purpose is to shed light
on a phenomenon and to produce detailed description of it. Further, it aims
to develop possible explanations of a phenomenon and to evaluate it.
Stake (1995) classifies case study research into the three types of
intrinsic, instrumental, and collective case studies. Intrinsic case study, the
method of choice in this research, is undertaken because one wants better
understanding of a particular case. It is not undertaken primarily because
the case represents other cases or because it illustrates a particular issue or
problem, but because, in all its particularity and ordinariness, the case itself
is of interest (Denzin & Lincoln, 1998).
The more the researcher has intrinsic interest in the case, the more
the focus of study will usually be on the case's uniqueness, particular
context, issues, and story. Denzin and Lincoln (1998) summarize the major
conceptual responsibilities of the qualitative case researcher as follows:
bounding the case, conceptualizing the object of the study,
selecting phenomena, themes, or issuesthat is, the research
questionsto emphasize,
seeking patterns of data to develop the issues,
triangulating key observations and bases for interpretation,
selecting alternative interpretation to pursue, and
developing assertions or generalizations about the case.
Data analysis in a case study could take the form of interpretational
analysis or structural analysis. Interpretational analysis examines data for
constructs, themes, and patterns that can be used to describe and explain
the case. Structural analysis searches the data for patterns with little or no
inference made to the meaning of the pattern.
51


J. Summary
This literature review shows there is an abundance of literature on the
topics of perceived and actual environmental disease clusters, adverse
health effects of PCE, the application of GIS in environmental and public
health, spatial and temporal statistical methods, environmental equity (or
environmental justice), models of community organization, and qualitative
methods. My goal in this dissertation research is to integrate and meld this
information in answering the health concerns of the North Casper
community.
52


CHAPTER 3
THE PEOPLE AND THE ENVIRONMENT
A. Introduction
Casper is a city of about 49,000 residents, located in the middle of
the state of Wyoming in Natrona County. The North Casper community is
situated in the northwestern comer of the city between an active rail line that
slices east-west through the city and the North Platte River. Since the late
1800s, Casper has been a host to a variety of industries. In the early part of
the twentieth century, oil-related enterprises, including refineries, began to
dominate the industrial base in Casper. Figure 3.1 shows a street map of the
northern part of Casper with the boundaries of North Casper marked with a
thick dashed line. The map illustrates the location of two PCE plumes as
well as the street locations of the polluting industries that include an Amoco
refinery, a former dry cleaner, and a city maintenance garage.
53


Figure 3.1. Map of North Casper, Wyoming, identifying the plumes and
locations of potential polluters
54


B. Socio-demographic Characteristics
North Casper has boundaries that fall completely within Census Tract
2 in Natrona County, Wyoming. A drive through the neighborhood shows
that North Casper is clearly a lower-income area, with tightly spaced homes,
some of which are in disrepair or abandoned. Several boarded-up
businesses quite visibly line some of the streets. The residents are more
racially and ethnically diverse than other parts of the city of Casper.
In subsequent sections of this chapter, I will first construct North
Casper population indicators such as size, sex, age, race, and poverty
status distributions and ratios using 1990 and 2000 US Censuses as the
sources of data. Tract level census data for Tracts 2 (North Casper), 3, 4, 6,
and 7 were obtained. Geographically, Tracts 3, 4 and 7 are immediately
adjacent to Tract 2; and Tract 6 is distant (one tract removed) from Tract 2.
In the sections that follow, population indicators for each geographic unit are
discussed.
In the last section of this chapter, a summary of environmental
contamination studies in North Casper based on the data obtained from the
EPA and WDEQ is presented.
Total Population Distribution and Ratios
Based on the 1990 census, the population of the entire city of Casper
stood at 46,742 people. Approximately 2,115 (4.5%) of that population
resided in North Casper (Census Tract 2). By the 2000 census, the city of
Casper's population reached 49,644 people. There was also an increase in
North Casper's population to 3,692 people (7.4% of the total population of
Casper). Table 3.1 shows the total population in each tract as well as the
proportion of the Casper population that are within each tract. As indicated
55


in the table, North Casper had the lowest proportion of the total population of
Casper at 4.5% followed by Tract 7 at 5.2%.
It is also evident from Table 3.1 that, with the exception of North
Casper (Tract 2), the population of the other census tracts remained quite
stable through the ten-year period between the two censuses. Figure 3.2 is
a graph of the population change (in percent) of the entire city of Casper and
the five census tracts from 1990 to 2000. With an increase of 74.6% in
population, North Casper experienced the highest growth in population
compared to Casper overall and three other tracts. The city of Casper and
Tracts 3, 6, and 7 had a 6.2%, 6.0%, 3.2%, and 12.8% increase in their
populations respectively; Tract 4 had a 0.2% decrease in population.
56


Table 3.1. A comparison of population in the 1990 and 2000 censuses
among five census tracts
1990 Census
Percent3
Number
2000 Census
Number
Percent'
0,
a
City of Casper 46,742 100 49,644 100
Tract 2 (N. Casper) 2,115 4.5 3,692 ; 7.4
Tract 3 3,824 8.2 4,055 8.2
Tract 4 4,154 8.9 4,146 8.4
Tract 6 6,409 13.7 6,612 13.3
Tract 7 2,428 5.2 2,739 5.5
3 Proportion of total population of the city of Casper.
57


Figure 3.2. Population change (in percent) in the city of Casper and five
census tracts from 1990 to 2000
Population
Change
(%)
58


The US Census Bureau does not collect population movement in and
out of different census tracts. However, a comparison of the population (in
percent) living in the same residence five years prior to the census year
confirms the earlier assertion that North Casper has had a less stable
population during the 1990s compared to the city of Casper and other tracts.
The population dynamic across different tracts and the city of Casper based
on 1990 and 2000 census data is presented in Table 3.2. Approximately
17% fewer people, aged five or older, lived in the same house for five years
prior to the census in 1990 than they did in 2000. This difference in mobility
was significantly less pronounced in the city of Casper and other census
tracts compared to that of North Casper. Due to an absence of intra-tract
mobility data, it was not possible to assess if the change in residence
occurred within the same tract or to a different one.
59


Table 3.2. Population dynamic across different tracts arid the city of Casper
based on 1990 and 2000 census data
Living in the Same House in 1985 Living in the Same House in 1995
Number Percent of Total Population (>=5) Number Percent of Total Population (>=5) Difference in Percent Mobility
Casper 19,624 45.5 23,560 50.8 5.3
Tract 2 (N. Casper) 1,067 54.8 1,285 37.7 -17.1
Tract 3 1,626 45.2 1,375 36.2 -0.9
Tract 4 , 1,761 45.8 2,150 54.9 9.1
Tract 6 2,536 43.6 3,180 52.6 9.0
Tract 7 1,148 50.8 1,052 40.7 -10.0
60


Sex Ratio
The sex ratio is the number of males per 100 females in a popultion.
This ratio is calculated by dividing the number of males in the population
over the number of females and multiplying the product by 100. It is often
calculated to assess any imbalance that may exist between the number of
males and females in a given population. The sex ratios for the city of
Casper and the tracts were calculated to determine: (1) if the sex ratio
shows an imbalance between the number of males and females in the
population of North Casper (the area with known contamination) compared
with other tracts, (2) if there is a shift in sex ratio of the population in the
contaminated areas from one census to the next, and (3) if the shift in sex
ratio favors one gender over another:
A graph of the sex ratio calculations for Casper and the five census
tracts from both 1990 and 2000 census data is shown in Figure 3.3. As
Figure 3.3 demonstrates, the sex ratio in Tract 2 (North Casper) and Tract 7
is much lower than those in the population of Casper and other tracts. At the
time when the 1990 census was taken, there are 88 males per 100 females
in North Casper. This ratio, however, increases to 97 (males per 100
females) by Census 2000. It also appears that the shift in sex ratio narrows
the gap between the proportion of male in the population to that of female in
North Casper.
61


Figure 3.3. Sex ratio in the city of Casper and across five census tracts (sex
ratios are shown as males per 100 females)
Number
of
males
per 100
females
Casper Tract2 Tract3 Tract 4 Tract6 Trad7
(North
Casper)
62


Age Distribution
The population's age distribution in North Casper was compared to
those in the city of Casper and other census tracts. Tables 3.3 and 3.4 are
summaries of the population counts divided into thirteen age categories for
each census tract as well as the entire city of Casper from the 1990 and
2000 census data respectively. The tables contain the proportion of each
age categories' count compared to the total population for each geographic
unit. The age distribution of the population in North Casper in 1990 was
compared with the city of Casper using chi-square statistics. A 2x2
contingency table was constructed that contained the number of people in
each age category (as the rows) in North Casper and a comparison
geographic unit, the city of Casper or Census Tract 3, for example (as the
columns). The analysis of the data using chi-square statistics shows the
number of people in the 15 to 19 and in age categories within 35 to 84 age
range are statistically significantly different (at a = 0.05, degrees of freedom
= 1) between the two populations.
Proportions of total population in the 15 to 19, 20 to 24, 35 to 44, and
45 to 54 age categories are statistically significantly lower in North Casper
than those in Casper (p-values of 0.0285, 0.0001, 0.004, and 0.0483
respectively). The proportion of population in the 55 to 84 age categories for
North Casper is significantly higher than Casper (p-value = 0.0001).
As it is demonstrated in Tables 3.3, 3.4, 3.5, and 3.6, overall, North
Casper has an older population when compared to the entire city of Casper
in 1990. However, by the 2000 census, there is a shift in proportions with
North Casper having significantly higher numbers of people in the under 5
(p-value = 0.002) and 65 years of age and over categories (p-values of
63


0.0038, 0.0135, and 0.0253 for 65 to 74, 75 to 84, and 84 and over
categories respectively) compared to the entire city of Casper.
64


Table 3.3. Population counts by age categories based on 1990 census data
Casper N. Casper Tract 2 Tract 3 Tract 4 Tract 6 Tract 7
Under 5 years 3,587 (8.0%)* 169 (8.0%) 224 (5.9%) 311 (7.5%) 586 (9.1%) 166 (6.8%)
5 to 9 years 4,258 (9.0%) 186 (8.8%) 241 (6.3%) '1AA (8.3%) 587 (9.2%) 168 (6.9%)
10 to 14 years 3,576 (8.0%) 152 (7.2%) 252 (6.6%) 308 (7.4%) 539 (8.4%) 141 (5.8%)
15 to 19 years 3,515 (8.0%) 131 (62%) 292 (7.6%) 249 (6.0%) 453 (7.1%) 168 (6.9%)
20 to 24 years 2,649 (6.0%) 141 (6.7%) 279 (7.3%) 235 (5.7%) 420 (6.6%) 140 (5.8%)
25 to 34 years 7,836 (17.0%) 349 (16.5%) 622 (16.3%) 642 (15.5%) 1,271 (19.8%) 391 (16.1%)
35 to 44 years 7,667 (16.0%) 242 (11.4%) 537 (14%) 609 (14.7%) 968 (15.1%) 357 (14.7%)
45 to 54 years 4,506 (10.0%) 162 (7.7%) 345 (9.0%) 363 (8.7%) 539 (8.4%) 203 (8.4%)
55 to 59 years 1,939 (4.0%) 105 (5.0%) 124 (3.2%) 214 (5.2%) 203 (3.2%) 118 (4.9%)
60 to 64 years 2,134 (5.0%) 136 (6.4%) . 185 (4.8%) 284 (6.8%) 275 (4.3%) 123 (5.1%)
65 to 74 years 3,301 (7.0%) 213 (10.1%) 365 (9.5%) 427 (10.3%) 388 (6.1%) 268 (11.0%)
75 to 84 years 1,411 (3.0%) 105 (5.0%) 254 (6.6%) 145 (3.5%) 153 (2.4%) 147 (6.1%)
85 years and over 454 (<1.0%) 24 (1.1%) 104 (2.7%) 23 (<1.0%) 27 (0.4%) 38 (1.6%)
Total 46,833 2,115 3,824 4,154 6,409 2,428
*The second number is the proportion to the total population in percent.
65


Table 3.4. Population counts by age categories based on 2000 census data
Casper N. Casper Tract 2 Tract 3 Tract 4 Tract 6 Tract 7
Under 5 years 3,264 (7.0%)* 301 (8.2%) 225 (5.5%) 272 (6.6%) 528 (8.0%) 165 (6.0%)
5 to 9 years 3,458 (7.0%) 197 (5.3%) 238 (5.9%) 311 (7.5%) 512 (7.7%) 161 (5.9%)
10 to 14 years 3,738 (8.0%) 200 (5.4%) 224 (5.5%) 293 (7.1%) 522 (7.9%) 217 (7.9%)
15 to 19 years 4,122 (8.0%) 299 (8.1%) 290 (7.2%) 265 (6.4%) 479 (7.2%) 238 (8.7%)
20 to 24 years 3,455 (7.0%) 307 (8.3%) 359 (8.9%) 269 (6.5%) 513 (7.8%) 214 (7.8%)
25 to 34 years 6,125 (12.0%) 481 (13.0%) 533 (13.9%) 550 (13.3%) 1006 (15.2%) 355 (13.0%)
35 to 44 years 7,649 (15.0%) 530 (14.4%) 559 (13.8%) 599 (14.4%) 1018 (15.4%) 422 (15.4%)
45 to 54 years 7,016 (14.0%) 488 (13.2%) 531 (13.1%) 490 (11.8%) 868 (13.1%) 403 (14.7%)
55 to 59 years 2,211 (5.0%) 155 (4.2%) 171 (4.2%) 179 (4.3%) 234 (3.5%) 118 (4.3%)
60 to 64 years 1,852 (4.0%) 135 (3.7%) 140 (3.5%) 185 (4.5%) 210 (3.2%) 90 (3.3%)
65 to 74 years 3,606 (7.0%) 317 (8.6%) 289 (7.1%) 412 (9.9%) 389 (5.9%) 190 (6.9%)
75 to 84 years 2,402 (5.0%) 210 (5.7%) 314 (7.7%) 281 (6.8%) 257 (3.9%) 126 (4.6%)
85 years and over 746 (1.5%) 72 (2.0%) 152 (3.7%) 40 (1.0%) 76 (1.1%) 40 (1.5%)
Total 49,644 3,692 4,025 4,146 6,612 2,739
*The second number is the proportion to the total population in percent.
66


Table 3.5. Age distribution comparison between North Casper and the city of
Casper (a = 0.05; two-sided; degrees of freedom = 1) in 1990
(Source: US Census 1990)
Age Category No Difference, Lower, or Higher Proportion in N. Casper Compared to the city of Casper P-value
Under 5 years No Difference 0.6547
5 to 9 years No Difference 0.6547
10 to 14 years No Difference 0.4028
15 to 19 years Lower 0.0285
20 to 24 years No Difference 0.0614
25 to 34 years No Difference 0.7518
35 to 44 years Lower 0.0001
45 to 54 years Lower 0.004
55 to 59 years Higher 0.0483
60 to 64 years Higher 0.0001
65 to 74 years Higher 0.0001
75 to 84 years Higher 0.0001
85 years and over No Difference 0.4386
67


Table 3.6. Age distribution comparison between North Casper and the city of
Casper (a = 0.05; two-sided; degrees of freedom = 1) in 2000
(Source: US Census 2000)
Age Category No Difference, Lower, or Higher Proportion in N. Casper Compared to the city of Casper P-value
Under 5 years Higher 0.0002
5 to 9 years Lower 0.0001
10 to 14 years Lower 0.0001
15 to 19 years No Difference 0.6547
20 to 24 years Higher 0.0026
25 to 34 years No Difference 0.2059
35 to 44 years No Difference 0.1069
45 to 54 years No Difference 0:1573
55 to 59 years No Difference 0.4028
60 to 64 years No Difference 1.00
65 to 74 years Higher 0.0038
75 to 84 years Higher 0.0135
85 years and over Higher 0.0253
68


Race
The racial distribution in the population of the city of Casper and
Census Tracts 2, 3, 4, 6, and 7 was calculated based on both 1990 and
2000 census data. Racial distribution based on 1990 census data is shown
in Table 3.7; racial distribution based on 2000 census data is shown in Table
3.8. The distribution and proportions of each racial category in North Casper
were compared with those of the entire city of Casper using chi-square
statistics. A 2x2 contingency table was constructed for the number of people
in each racial group (as the rows) in North Casper and another geographic
unit, Census Tract 6, for example (as the column), Except for the proportion
of American Indian/Alaskan, all other categories are statistically significantly
different between North Casper and the area with which it was compared.
The proportion of the population who were classified as white is significantly
lower in North Casper compared to those in the entire city of Casper (p-
value = 0.0001). However, the population proportion that were classified as
black/African American, American Indian/Alaskan, Asian, or other are
significantly higher in North Casper compared with those in the entire city of
Casper. Tables 3.9 and 3.10 show the results of chi-square statistics with
the p-values. As it is evident from the data, although greater proportions of
the population of North Casper and the city of Casper are white, North
Casper has significantly greater proportions of the minorities (races other
than white) than does the city of Casper as a whole.
69


Table 3.7. Race distribution in the population in the city of Casper and the
selected census tracts (Census 1990)
Casper Tract 2 (North Casper) Tract 3 Tract 4 Tract 6 Tract 7
White 45,332 (97.0%) 1,792 (84.7%) 3,638 (95.1%) 4,041 (97.3%) 6,168 (96.2%) 2,362 (97.3%)
Black/African American 355 (0.76%) 110 (5.2%) 34 (0.9%) 17 (0.4%) 77 (1.2%) 4 (0.2%)
American Indian/Aiaskan 266 (0.6%) 13 (0.6%) 38 (1.0%) 25 (0.6%) 36 (0.6%) 34 (1.4%)
Asian 190 (0.41%) 17 (0.8%) 15 (0.4%) 21 (0.4%) 53 (0.8%) 4 (0.2%)
Other 619 (1.3%) 183 (8.7%) qq (2.6%) 50 (1.2%) 75 (1.2%) 24 (1.0%)
Total 46,762 2,115 3,824 4,154 6,409 2,428
70


Table 3.8. Race distribution in the population in the city of Casper and the selected
census tracts (Census 2000)
Casper Tract 2 (North Casper) Tract 3 Tract 4 Tract 6 Tract 7
White 46,680 (94.0%) 3,049 (82.6%) 3,786 (93.4%) 3,899 (94%) 6,193 (93.7%) 2,584 (94.3%)
Black/African American 428 (0.9%) 131 (3.5%) 40 (1.0%) 28 (0.7%) 79 (1.2%) 17 (0.6%)
American Indian/Alaskan 495 (1.0%) 114 (3.1%) 60 (1.5%) ,o?%, 74 (1.1%) 37 (1.4%)
Asian 245 (0.5%) 25 (0.7%) 11 (0.3%) 23 (0.6%) 39 (0.6%) 7 (0.3%)
Other 1,021 (2.0%) 253 (6.9%) 84 (2.1%) 98 (2.4%) 106 (1.6%) 56 (2.0%)
Total 48,869 3,572 3,981 4,081 6,491 2,701
71


Table 3.9. Comparison of racial distribution in population between the city of
Casper and North Casper (Census 1990) using chi-square statistics (a =
0.05; degrees of freedom = 1)
Casper Proportion of Casper Population Tract 2 N. Casper (Observed) Tract 2 N. Casper (Expected) Chi- Square P-value
White 45,332 0.970 1,792 2,052 32.8 0.0001
Black/African American 355 0.0076 110 16 548.8 0.0001
American Indian/Alaskan 266 0.006 13 13 0.0 1.00
Asian 190 0.0041 17 9 8.0 0.0047
Other 619 0.013 183 27 879.5 0.0001
Total 46,762 2,115
72


Table 3.10. Comparison of race distribution in population between the city of
Casper and North Casper (Census 2000) using chi-square statistics (a =
0.05; degrees of freedom = 1)
Casper Proportion of Casper Population North Casper (Observed) North Casper (Expected) Chi- Square P-value
White 46,680 0.94 3,049 3357.68 28.4 0.0001
Black/African American 428 0.009 131 32.148 304.0 0.0001
American Indian/Alaskan 495 0.01 114 35.72 171.5 0.0001
Asian 245 0.005 25 17.86 2.9 0.0886
Native Hawaiian 10 0.0 0.0 0.0 0.0 1.0
Other 1,011 0.002 253 7.144 8461.0 0.0001
Total 48,869 3,572
73


Poverty Status
The North Casper Elementary School principal reported that 90
percent of the students in her school and 65 percent of students in the high
school were eligible for subsidized school lunches during the 1998-1999
school year (Citizens Outreach, 1999). This implies that the parents annual
earnings were at, or below, the poverty line. The proportion of the population
below the poverty line in 1989 and 1999obtained from Census 1990 and
Census 2000 data respectivelyfor the city of Casper and Census Tracts 2,
3, 4, 6, and 7 were calculated (Table 3.11). A comparison of the proportions
reveals that the North Casper community has a disproportionately higher
number of families and individuals below the poverty line compared to those
in the other tracts. Overall, people living in North Casper are economically
more disadvantaged than those living in other tracts or the city of Casper.
Although there are smaller proportions of families and families with a female
head of household living below the poverty line in 1999 compared to 1989.
The proportion of individuals below the poverty line, however, increased
from 1989 to 1999.
74


Table 3.11. Comparison of poverty status across the city of Casper and
Census Tracts 2, 3, 4, 6, and 7 in 1989 and 1999 (Censuses 1990 and
2000)
Casper (in %)a Tract 2 (North Casper)a (in %) Tract 3 (in %)a Tract 4 (in %)a Tract 6 (in %)a Tract 7 (in %)a
Families (1989) N/A 14.5 6.2 3.2 4.6 3.1
Families (1999) 8.5 5.6 4.3 2.5 1.8. 1.4
Female head of household/no husband N/A 12.4 3.3 7.7 5.1 9.9
(1989) Female head of household/no husband 27.8 3.2 2.3 1.4 1.1 0.5
(1999) Individuals (1989) N/A 6.8 9.6 2.6 2.5 4.9
Individuals (1999) . 11.4 29.6 21 12 8.8 11.1
a Numbers shown are percentages of total population in each geographic
unit.
75


C. Contamination History
Both government agencies and industries have sponsored several
environmental studies in North Casper since 1989. These studies focused
on contaminated plumes (i.e., high concentrations of contaminants in
groundwater and soil) and industrial sites in North Casper. In 1994,
Huntingdon Engineering and Environmental, Inc. performed two studies.
One study investigated contamination in an area bound by Melrose Street,
Center Street, 1-25, and the North Platte River (Huntingdon, 1994). The
second study investigated contamination in downtown Casper (Huntingdon,
1994a). Both Huntingdon studies were performed on behalf of the Wyoming
Department of Environmental Quality (WDEQ). A map of the two plumes'
locations was presented at the beginning of this chapter in Figure 3.1. Both
of the above studies provided summaries of other investigations pertaining
to contaminations in North Casper. The results of the Huntingdon studies
and the earlier studies are presented in Appendix B.
The environmental contamination studies demonstrated that there
were extensive chemical contamination and severe environmental damage
in North Casper. The contaminants ranged from significant levels of
hydrocarbon contamination to PCE plumes to benzene, toluene, xylene, and
other volatile organic compounds (VOCs). Some of the contaminations and
their locations are briefly discussed here. Appendix B provides the complete
information about the studies along with the dates they were conducted and
the sponsoring agencies.
76


Hydrocarbon contamination existed in every monitoring well at a site
located at 605 South Poplar Street. The Amoco Refinery (west of
the site) was found to be responsible for the contamination.
Major tetrachloroethylene (PCE) and trichloroethylene (TCE)
contaminations were found on North Beech Street.
Hazardous waste constituents such as benzene and toluene were
found at 613 West Yellowstone Highway on property belonging to
Amoco that contained eleven underground storage tanks of various
petroleum-based hazardous wastes'
A PCE plume moving southwest to northeast toward the North
Platte River that originated from a former dry cleaning business
located right at the center of North Casper was found.
High concentrations of petroleum contamination were found beneath
the Baroid Drilling property at 1030 West Collins Drive.
PCE in groundwater and petroleum hydrocarbons in both soil and
groundwater were found between North Center Street and North
McKinley Street.
Accumulations of petroleum hydrocarbons at the northeast corner of
South Poplar Street and Yellowstone Highway were found that had
originated, in part, from the Amoco refinery.
Most of the studies conducted failed to identify one specific source. In
some cases, multiple sources of pollution were identified. The uncertainty
about the specific sources has led to disputes among polluters as to who
has primary responsibility for the pollutiona situation in which finger
pointing supersedes efforts to actually clean up the contamination. Under
EPA's statutes and regulations, and based on the WDEQ Hazardous Waste
77


Management Division's assessment, the contaminated areas were not
"eligible for cleanup funds under the Superfund program" (EPA Doc.
404331). Therefore, the cleanup efforts in North Casper have thus far been
abysmal at best and non-existent, in most areas, at worst.
In a response to the citizens' request to reopen the North Casper
Plume site under Superfund authority; EPA reiterated its position as follows
(EPA Doc. 508437),
Upon our review and according to EPA policy
(OSWER Directive 9345.1-07, Hazard Ranking
System Guidance Manual), we cannot exercise
Superfund authority to address indoor air as a
pathway for proposal to the National Priority List
(NPL). EPA has worked diligently to address the
potential indoor air issues in North Casper and has
conferred with the Agency for Toxic Substances and
Disease Registry (ATSDR). Unless ATSDR issues a
health advisory, Superfund cannot propose this site
to the NPL. In speaking with ATSDR in the past, this
site does not warrant a health advisory.
D. Summary
A comparison of socio-demographic data from the 1990 and 2000
censuses shows that the population of North Casper has several indicators
of economic disadvantage. The community age and sex structure shows a
preponderance of women, children, and the elderly. There are greater
proportions of minorities residing in North Casper than in the other census
tracts in the city of Casper. In addition, most of the school children qualify
for the federal subsidized lunch program and almost one out of three
individuals is living at or below the poverty line in 2000. Signs of hope are
that there has been an improvement in some socio-economic indicators
between the 1990 and 2000 censuses. For example, there are fewer
78


families and households headed by women living in poverty in 2000
compared to 1990.
North Casper is also a severely contaminated area. The
contamination is not limited to the PCE plumes. Varieties of other chemicals
have contributed greatly to the environmental contamination problems in
North Casper. So far, neither the government nor the industries have taken
steps toward a comprehensive cleanup of the contamination. Economically,
the community is not in a position to be able to fund any cleanup efforts, nor
do they have the power to compel state and federal agencies to demand a
thorough cleanup. The North Casper community, as many lower income
areas in America, appears to be a victim not only of the contamination but
also of environmental inequity.
79


CHAPTER 4
QUALITATIVE ANALYSIS
A. Introduction
Many residents living in the North Casper area believe they have
been exposed to hazardous chemicals due to the contamination in the
neighborhood and report many health problems. They attribute their health
problems to [perceived] exposure to the contamination. Residents have
recounted stories of repeated health problems such as headaches, eye
problems, skin rashes, cancers (including leukemia), and birth defects to
local and state officials.
In order to understand the experience of the North Casper community
concerning its health and the neighborhood contamination problems, I will
reveal details about the members' experiences through an intrinsic case
study25 research design using qualitative methods (Creswell, 1998; Stake
1995). Consistent with case study research methodology (Stake, 1995), I
have limited the study to a specific system bounded by place and time-the
North Casper community between the years of 1989 and 2001.
25 A case study in which the interest is in understanding the particulars of the case.
80


B. Methods
To present specific and illuminating information about the community
experience, I focused on analysis of the content and exploration of available
documents. The documents consisted of the community members' records
of activities and their communication with health and environmental
agencies during the period between 1989 and 2001. It was beyond the
scope of this study to perform a more comprehensive case study, or to
conduct one-on-one interviews with key stakeholders. Instead the written
records of interaction between the community and government agencies
formed the basis of the data for the case study. Characteristic of case study
research, information from multiple sources were collected in this study.
They included:
minutes from the meetings with community representatives,
EPA, ATSDR, and WDEQ documents, and
written testimonials from North Casper citizens.
The above documents, along with their respective dates, references,
and where they were collected are summarized in Appendix C. Although the
information used for the qualitative portion of this study is part of the public
record, I eliminated the individual names and addresses from the summary
table to protect the privacy of the individual citizen or a particular agency's
staff.
I first reviewed all the collected documents, including the records of
my own interactions with the community members to analyze the data based
on qualitative strategies (Crabtree & Miller, 1992; Creswell, 1998; Olson,
81


1976) and to gain a general understanding of what had happened. Then, I
organized all the documents and records of activities in a chronological
order starting with the event and date when the community members first
became aware of the contamination problem. I marked the specific
segments in the document or record that revealed the gist of the
communication outlining the experience, view, or judgment and entered the
segments into a table. Next, I assigned codes to each text selection and
entered the codes into the adjacent column when appropriate.
The codes had been developed earlier based on the key constructs in
the social action model. The codes were "P&R" for participation and
relevance; EMP for empowerment, CRITCOM for critical consciousness,
COMCOM for community competence, and ISSEL for issue selection. In
addition, if other themes and sub-themes emerged that did not correspond
with those codes selected a priori, I coded them as OTHER to distinguish
them from social action model codes.
I present my interpretations of the themes and the case study as case
story, assertions, and reflection. Figure 4.1 is a diagram of the steps in the
qualitative analysis.
82


Figure 4.1. Qualitative analysis diagram
83


C. Thematic Analysis Results
After a review of all documents and gaining a general understanding
of the case study, it became evident that the North Casper community's
discontent with the handling of its environmental contamination and health
concerns by the officials was a gradual process that developed over time.
Therefore, I examined the themes that developed in a chronological order.
This way, I was able to capture some of the themes that were built on the
previous themes developed earlier in the process. I organized the themes
into an Early Period: Information Gathering Phase and Action Period: Active
Response to Environmental Exposure Phase.
The Early Period or the Information Gathering Phase was defined as
the period during which the community first became aware of a PCE
contamination problem in its neighborhood. During this phase, the citizens
raised questions and actively requested information regarding the impact of
the contamination on their health as well as their property values. A
proactive approach from EPA in informing and educating the community
about the pollution and its possible impact on citizens' health and property
value is lacking during this period.
The themes developed during this phase are "delay in response
time," perceived cover up" and "lack of transparency." These themes are
discussed later in this section.
The Action Period or the Active Response to Environmental Exposure
Phase was defined as the period during which the community assumed a
more active role in its response to the environmental problem in the
community and demanded a greater accountability from the agencies and
the polluting industries. During this period the community formed a citizen
84


group and began collecting information regarding the pollution. In addition,
they petitioned for health studies to assess the impacts of the contamination
The themes developed during the Action Period are "perception of
pro-industry stance, "vague and fragmented communication", "fear and
anxiety", "demand for health study", and "perceived unfair treatment." Each
theme is discussed below in its respective section.
Themes Early Period: Information Gathering Phase
Delay in response time. The shortest 'recorded' response time to a
concern or question posed by a citizen is five weeks regardless of the
urgency of the question. This does not include any oral response that may
not have been recorded or archived. A mother had inquired whether it was
safe, given the PCE contamination near her home, to bring her four-month-
old baby from the hospital to her home the following week. A response was
sent to her approximately five and a half weeks after the initial inquiry.
During fall 2001, due to a recently discovered spill and contamination
in the North Platte River, EPA and city officials ordered the closure of the
public water system in the city of Evansvillethe city adjacent to and up
river from North Casperuntil the contamination problem had been
contained and resolved. I received a telephone call from one of North
Casper citizens requesting me to telephone EPA to obtain relevant
information regarding the spill and the reason why North Casper residents
had not been put on an alert about the contamination. The citizen believed
that North Casper residents would be more likely to be exposed to any
contamination of the North Platte River than the residents of Evansville
since North Casper is located downstream from the spill.
85


The reason why the citizen wanted me to contact EPA rather than
contacting EPA directly was that.. it would take them [EPA] months to tell
us anything. Even if they tell us something, I am afraid they don't tell us the
truth and we may be drinking contaminated water."
Another citizen wrote, "[TJhere have been many attempts to inform
EPA and other coalition members of potential sources, including City of
Casper. No one ever responds to the complaints."
Perceived cover up and lack of transparency. An absence of
proactive communication with the citizens by government agencies is
perceived as cover up and lack of transparency. Some citizens declared,"..
. by accident, we discovered the existence of a preliminary plume map
which affects a major portion of the residential area and officials have
refused to release the information to the public." A citizen wrote,
... we discovered that HUD26 and all the other
entities involved knew of the ATSDR's 1993 health
advisory and had not notified the public in any
manner. Until documents were discovered verifying
this fact, we were led to believe that the redlining
was done on the spur of the moment earlier this
year. It wasn't until the fedline map was leaked to
the newspaper that the public was notified of the
supposedly short term redline. Lifting the redline was
only a tactical act. In fact, properties on the north
side of Casper remain at lower values than before.
In another statement, the same citizen asserted "EPA is setting up
new road blocks to prevent us from knowing the whole story." Another
citizen wrote",.. the citizens of North Casper have once again been notified
of a meeting first thru the news media ..." regarding a coalition meeting
26 Housing and Urban Development.
86


consisting of representatives from the government agencies and the
industries regarding the contamination.
The perception of a cover up gradually leads to citizen distrust of the
government officials. A community representative wrote, 'We cannot and will
not be in a position to rely on government officials who are blocking the truth
from the public." Finally, the distrust intensifies to a point that the citizens
view all governmental agencies1 operations and tests with great skepticism.
Following an air sample collection by EPA to determine the level of PCE
vapors in some residential units, a citizen wrote,.. ATSDR recommended
that indoor air sampling needed to be done during winter months using
Gillian pumps. Somehow during that process, the regulatory agencies
changed the type of sampling to Tedlar Bags, and in many instances
conducted the tests during the summer with the windows open." Later, a
new set of air samples using appropriate sample collection techniques was
performed by the WDEQ.
As the citizens' distrust grows, the regulatory agencies make attempts
at informing the public. However, these attempts are regarded with
suspicion. In a reaction to EPA fact sheets distributed to inform the public
about PCE exposure and toxicity, a citizen wrote, "[T]he fact sheets are only
one point of view which is designed to downplay hysteria at any cost." A
statement made by a community representative to the EPA illustrates the
enormity of community's mistrust of officials due to a perceived cover up:
We know you already have a pretty good picture of
the depth of contamination. We know you know the
site should easily qualify for the NPL [National
Priority List] if the truth were told and all the
information was used to make the determination. We
know none of you will come forward willingly to
protect the publics health and welfare. If it is the last
87