Ozone Health Assessment Plan: Scope and
Methods for Exposure Analysis and Risk
Assessment
Draft
April 2005
Office of Air Quality Planning and Standards
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-------
DISCLAIMER
This draft scope and methods plan has been prepared by staff from the Health and
Ecosystems Effects Group, Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, in conjunction with Abt Associates Inc. (through Contract No. 68-D-03-002,
WA 2-21). Any opinions, findings, conclusions, or recommendations are those of the authors
and do not necessarily reflect the views of the EPA or Abt Associates. This document is being
circulated to obtain review and comment from the Clean Air Scientific Advisory Committee
(CASAC) and the general public. Comments on this document should be addressed to Harvey
Richmond, U.S. Environmental Protection Agency, Office of Air Quality Planning and
Standards, C539-01, Research Triangle Park, North Carolina 27711 (email:
richmond.harvey@epa.gov).
-------
Table of Contents
1 INTRODUCTION 1
1.1 PURPOSE OF SCOPE AND METHODS PLAN 2
1.2 BACKGROUND 2
2 AIR QUALITY CONSIDERATIONS 3
3 SCOPE AND APPROACH FOR POPULATION EXPOSURE ANALYSIS 4
3.1 OVERVIEW 4
3.2 THE POPULATION EXPOSURE MODEL 4
3.3 USE OF HUMAN ACTIVITY DATA 6
3.3.1 Longitudinal human activity data 6
3.3.2 Representativeness of activity data 6
3.4 OUTCOMES TO BE GENERATED 7
3.5 SELECTION OF URBAN AREAS 7
3.6 EXPOSURE PERIODS 8
3.7 POPULATIONS TO BE ANALYZED 8
3.8 UNCERTAINTY AND VARIABILITY 8
4 SCOPE AND APPROACH FOR HEALTH RISK ASSESSMENT 9
4.1 OVERVIEW 9
4.2 STRUCTURE OF THE RISK ASSESSMENT 10
4.3 ASSESSMENT OF RISK BASED ON CONTROLLED HUMAN EXPOSURE STUDIES 11
4.3.1 Selection of health endpoints 12
4.3.2 Selection of exposure-response functions 12
4.3.3 Approach to calculating risk estimates 13
4.3.4 Selection of urban areas 14
4.4 ASSESSMENT OF RISK BASED ON EPIDEMIOLOGICAL AND FIELD STUDIES 14
4.4.1 Selection of health endpoints 15
4.4.2 Selection of urban areas 15
4.4.3 Selection of epidemiological and field studies 16
4.4.4 A summary of selected health endpoints, urban areas and studies 17
4.4.5 Selection of concentration-response functions 22
4.4.6 Baseline health effects incidence considerations 23
4.4.7 Assessing risk in excess of policy-relevant background 25
4.5 UNCERTAINTY AND VARIABILITY 26
5 SCHEDULE AND MILESTONES 28
6 REFERENCES 29
Figure 1. Overview of the APEX Model 34
Figure 2. Major Components of Ozone Health Risk Assessment Based on Controlled
Human Exposure Studies 37
Figure 3. Major Components of Ozone Health Risk Assessment Based on Epidemiology
and Field Studies 38
-------
Ozone Health Assessment Plan: Scope and Methods for Exposure
Analysis and Risk Assessment
1 INTRODUCTION
The U.S. Environmental Protection Agency (EPA) is presently conducting a review of
the national ambient air quality standards (NAAQS) for ozone (Os). Sections 108 and 109 of the
Clean Air Act (Act) govern the establishment and periodic review of the NAAQS. These
standards are established for pollutants that may reasonably be anticipated to endanger public
health and welfare, and whose presence in the ambient air results from numerous or diverse
mobile or stationary sources. The NAAQS are to be based on air quality criteria, which are to
accurately reflect the latest scientific knowledge useful in indicating the kind and extent of
identifiable effects on public health or welfare which may be expected from the presence of the
pollutant in ambient air. The EPA Administrator is to promulgate and periodically review, at
five-year intervals, "primary" (health-based) and "secondary" (welfare-based) NAAQS for such
pollutants.' Based on periodic reviews of the air quality criteria and standards, the Administrator
is to make revisions in the criteria and standards, and promulgate any new standards, as may be
appropriate. The Act also requires that an independent scientific review committee advise the
Administrator as part of this NAAQS review process, a function now performed by the Clean Air
Scientific Advisory Committee (CASAC).
EPA's overall plan and schedule for this Os NAAQS review is presented in a Plan for
Review of the National Ambient Air Quality Standards for Ozone (EPA, 2005a), which is
available at: http://www.epa.aov/ttn/naaqs/standards/ozone/s o3 cr pd. html. That plan
discusses the preparation of two key documents in the NAAQS review process: an Air Quality
Criteria Document (AQCD) and a Staff Paper. The AQCD provides a critical assessment of the
latest available scientific information upon which the NAAQS are to be based, and the Staff
Paper evaluates the policy implications of the information contained in the AQCD and presents
staff conclusions and recommendations for standard-setting options for the Administrator to
consider. In conjunction with preparation of the Staff Paper, staff in EPA's Office of Air Quality
Planning and Standards (OAQPS) conducts various policy-relevant assessments, including in
this review a quantitative exposure analysis and a human health risk assessment. This draft
document describes the scope and methods that staff is planning to use for these assessments.
The final section of this scope and methods plan identifies the major milestones and interim steps
involved in the planning, conduct, and documentation of these assessments.
'Section 109(b)(l) [42 U.S.C. 7409] of the Act defines a primary standard as one "the attainment and
maintenance of which in the judgment of the Administrator, based on such criteria and allowing an adequate margin
of safety, are requisite to protect the public health."
-------
1.1 Purpose of Scope and Methods Plan
This plan is designed to outline the scope and approaches and highlight key issues in the
estimation of population exposures and health risks posed by 03 under existing air quality levels
("as is" exposures and health risks), upon attainment of the current O3 primary NAAQS, and
upon meeting various alternative standards in selected sample urban areas. This plan is intended
to facilitate consultation with the CASAC, as well as public review, and to obtain advice on the
overall scope, approaches, and key issues in advance of the completion of such analyses and
presentation of results in the first draft of the 03 Staff Paper.
The planned 03 exposure analysis and health risk assessment address short-term
exposures to 03 and associated health effects. These assessments cover a variety of health
effects for which there is adequate information to develop quantitative risk estimates. However,
there are some health endpoints for which there currently are insufficient information to develop
quantitative risk estimates. Staff plans to discuss these additional health endpoints qualitatively
in the O3 Staff Paper. The risk assessment is intended as a tool that, together with other
information on these health endpoints and other health effects evaluated in the 03 AQCD and O3
Staff Paper, can aid the Administrator in judging whether the current primary standard is
requisite to protect public health with an adequate margin of safer}', or whether revisions to the
standard are appropriate.
1.2 Background
As part of the last O3 NAAQS review, EPA conducted exposure analyses for the general
population, children who spent more time outdoors, and outdoor workers. Exposure estimates
were generated for 9 urban areas for "as is" air quality and for just meeting the existing 1-hour
standard and several alternative 8-hour standards. Several reports (Johnson et al., 1996a,b,c;
Johnson, 1997) that describe these analyses can be found at:
htlp./Avwvv.epa.gov/lln/naaq.s/standards/o/.one/s_o3_pr_ld.hlml. EPA also conducted a health
risk assessment that produced risk estimates for the number of children and percent of children
experiencing lung function and respiratory symptoms associated with the exposures estimated
for these same 9 urban areas. This portion of the risk assessment was based on exposure-
response relationships developed from analysis of data from several controlled human exposure
studies. The risk assessment for the last review also included risk estimates for excess
respiratory-related hospital admissions related to O3 concentrations for New York City based on
a concentration-response relationship reported in an epidemiology study. Risk estimates for lung
function decrements, respiratory symptoms, and hospital admissions were developed for "as is"
air quality and for just meeting the existing 1-hour standard and several alternative 8-hour
standards. Reports describing the health risk assessment (Whitfield et al., 1996; Whitfield, 1997)
can be found at: http://www.epa.gov/rtivnaaqs/standards/ozone/s o3 pr td.html.
The planned exposure analysis and health risk assessment described in this Scope and
Methods Plan build upon the methodology and lessons learned from the exposure and risk work
-------
conducted for the last review. These plans are also based on the information currently available
in the first draft Os AQCD; as such, some aspects of these plans may change based on changes
that may be incorporated in the final O3 AQCD.
2 AIR QUALITY CONSIDERATIONS
Staff plans to perform exposure and health risk analyses using the most recent year
(2004) of air quality data available at this time. The time period to be analyzed will be the 63
season, which in the urban areas to be included in this assessment, varies from April to October
to the entire year depending on the region of the country.
The following air quality scenarios will be considered:
• "As is" air quality in each urban area for 2004,
• meeting the current 8-h 0.08 ppm, average 4th daily maximum standard, and
• meeting alternative Os standards.
In order to conduct exposure and risk analyses for the last two scenarios, staff will adjust
the air quality data to simulate just meeting the current and alternative standards. The adjustment
of air quality data will be based on three years of data (2002 - 2004). Staff is currently
considering various approaches to making such adjustments, including the quadratic air quality
adjustment approach that was evaluated and used in the last review (Johnson, 1997).
A key issue to be addressed in the Os Staff Paper is the characterization of policy-
relevant background 63 levels in the U.S, which is defined as the distribution of Os
concentrations that would be observed in the U.S. in the absence of anthropogenic (man-made)
emissions of Oa precursors in the U.S., Canada, and Mexico. This definition appropriately
allows for analyses that focus on the effects and risks associated with pollutant levels that have
the potential to be controlled by U.S. regulations, through international agreements with border
countries, or by voluntary emissions reductions in the U.S. and elsewhere. Staff estimates of
policy-relevant background, including consideration of regional and seasonal differences, will be
informed by information and analyses in the draft Os AQCD, consideration of the results of air
quality simulation models, and analyses of measured ambient Os concentrations. In particular,
the results of the global tropospheric O3 model GEOS-CHEM will be used to estimate monthly
average background Os levels for different geographic regions across the U.S. These GEOS-
CHEM simulations include a background simulation in which North American anthropogenic
emissions of nitrogen oxides, non-methane volatile organic compounds, and carbon monoxide
are set to zero, as described in Fiore et al. (2003).
-------
3 SCOPE AND APPROACH FOR POPULATION EXPOSURE ANALYSIS
3.1 Overview
Population exposure to ambient Oa levels will be evaluated using anew version of the
Air Pollutants Exposure (APEX) model, also referred to as the Total Risk Integrated
Methodology/Exposure (TRIM. Expo) model. Exposure estimates will be developed for current
Os levels, based on 2004 air quality data, and for Os levels associated with just meeting the
current 8-h O3 NAAQS and alternative O3 standards, based on adjusting 2002-2004 air quality
data Exposure estimates will be modeled for 12 urban areas located throughout the U.S. for 1)
the general population, 2) all school-age children, 3) active school-age children, and 4) asthmatic
school-age children. This choice of population groups includes a strong emphasis on children,
which reflects the results of the last review in which children, especially those who are active
outdoors, were identified as the most important at-risk group.
The exposure estimates will be used as an input to that part of the health risk assessment
that is based on exposure-response relationships derived from controlled human exposure
studies, discussed in Section 4.3 below. The exposure analysis will also provide information on
population exposure exceeding levels of concern that are identified based on evaluation of health
effects that are not included in the quantitative risk assessment. The methodology used to
conduct the exposure analysis as well as summary results and key findings from the exposure
analysis will be presented in the QI Staff Paper. In addition, an exposure analysis technical
support document with a more detailed description of the methodology and results will
accompany the 03 Staff Paper.
3.2 The Population Exposure Model
The EPA has developed APEX as a tool for estimating human population exposure to
criteria and air toxic pollutants. APEX serves as the human inhalation exposure model within
the Total Risk Integrated Methodology2 (TRIM) framework (Richmond et al., 2002; EPA 2003).
APEX is a PC-based model that was derived from the probabilistic NAAQS Exposure Model
(pNEM) used in the last 03 NAAQS review (Johnson et al., 1996a, 1996b). Figure 1 provides a
schematic overview of the APEX model.
APEX simulates the movement of individuals through time and space and their exposure
to a given pollutant in indoor, outdoor, and in-vehicle microenvironments. The model
stochastically generates simulated individuals using census-derived probability distributions for
demographic characteristics (Figure 1, steps 1-3). The population demographics are from the
2000 Census at the tract level, and a national commuting database based on 2000 census data
2 The Total Risk Integrated Methodology is described at http://w\vw.cpa.cov/ttr./fera.
-------
provides home-to-work commuting flows between tracts. Any number of simulated individuals
can be modeled, and collectively they represent a random sample of the study area population.
Diary-derived time activity data from the Consolidated Human Activity Database
(CHAD) (McCurdy et al., 2000; EPA, 2002; Graham and McCurdy, 2004) are used to construct
a sequence of activity events (each event < 60 minutes) for each simulated individual consistent
with the individual's demographic characteristics and accounting for effects of day type and
temperature on daily activities (Figure 1, step 4). APEX calculates the concentration in the
microenvironment associated with each event in an individual's activity pattern and sums the
event-specific exposures by hour to obtain a continuous time series of hourly exposures spanning
the time period of interest (Figure 1, steps 5, 6).
APEX has a flexible approach for simulating microenvironmental concentrations, where
the user can define the microenvironments to be modeled. For the application to Os, the
following microenvironments will be modeled:
• Indoors - residence
• Indoors - bars and restaurants
• Indoors - schools
• Indoors - day care centers (commercial)
• Indoors - other (e.g., offices, shopping)
• Outdoors - near road
• Outdoors - other (e.g., playgrounds, parks)
• In vehicle - cars, trucks, etc.
• In vehicle - mass transit vehicles
The concentrations in each microenvironment are calculated using either a factors or
mass-balance approach, and the user specifies the probability distributions of the parameters that
go into the concentration calculations (e.g., indoor-outdoor air exchange rates). These
distributions can depend on the values of other variables in the model. For example, the
distribution of air exchange rates in a home, office, or car depends on the type of heating and air
conditioning present, which are also stochastic inputs to the model. The user can choose to keep
the value of a stochastic parameter constant for the entire simulation (e.g., house volume), or can
specify that anew value shall be drawn hourly, daily, or seasonally from specified distributions.
APEX also allows the user to specify diurnal, weekly, or seasonal patterns for various
microenvironmental parameters.
The calculation of microenvironmental concentrations in APEX is dependent not only on
the parameter distributions for the mass balance and factors approaches, but also on the ambient
(outdoor) Os concentrations and temperatures. Hourly Oj concentration measurements from the
fixed-site monitoring data maintained in EPA's Air Quality System and surface temperatures
from the National Weather Service will be spatially interpolated for each study area for input to
APEX.
-------
Exposure modeling will be conducted based on Os concentrations measured in 2004 and
for air quality scenarios reflective of meeting alternative Oa standards. Exposure modeling will
also be performed based on policy-relevant background concentration levels alone, in order to be
able to assess health risks due to 63 concentrations in excess of background.
3.3 Use of Human Activity Data
3.3.1 Longitudinal human activity data
The human activity data will be drawn from the CHAD developed and maintained by the
Office of Research and Development's (ORD) National Exposure Research Laboratory (NERL).
The average subject in the time/activity studies provided less than two days of diary data. For
this reason, the construction of a season-long activity sequence for each individual requires some
combination of repeating data from one subject and using data from multiple subjects. A key
issue in this assessment is the development of an approach for creating Oa-season or year-long
activity sequences for individuals based on a cross-sectional activity data base that includes 24-
hour records. An appropriate approach should adequately account for the day-to-day and week-
to-week repetition of activities common to individuals while maintaining realistic variability
between individuals. Staff, in conjunction with staff from NERL, is developing a methodology
for constructing longitudinal diaries from the CHAD data which will be used in the 63 exposure
analysis. This method will be described in the exposure analysis technical support document.
3.3.2 Representativeness of activity data
The CHAD includes data from several surveys covering specific time periods at city,
state, and national levels, with varying degrees of representativeness. NERL staff plans to
supplement these data with more recent data where available.3 The extent to which the human
activity database provides a balanced representation of the population being modeled is likely to
vary across areas. Although the algorithm that constructs activity sequences attempts to account
for the effects of population demographics and local climate on activity, this adjustment
procedure is unlikely to fully account for all intercity differences in people's activities. Activity
patterns are likely to be affected by many local factors, including topography, land use, traffic
patterns, mass transit systems, and recreational opportunities. Issues related to the selection and
representativeness of the CHAD activity diaries for the 12 urban areas modeled will be addressed
in the exposure analysis technical support document and in the O3 Staff Paper.
3 For example, the time diary activity data from the 2002 Child Development Supplement (CDS-II) are available at
hUp://usidonltnc.isr.iunich.edu/Data. This survey collected activity data for one randomly sampled weekday and
one weekend day for 2,569 children and has more than 99,000 activity records.
-------
3.4 Outcomes to be Generated
There are several useful indicators of exposure of people to various levels of air
pollution. Factors that are important in defining such indicators include the magnitude and
duration of exposures, frequency of repeated high exposures, and ventilation rate (i.e., breathing
rate) of the individual at the time of exposure. In this analysis, exposure indicators will include
daily maximum 1- and 8-h average Oj exposures, stratified by equivalent ventilation rates (i.e.,
ventilation normalized by body surface area).
APEX calculates two general types of exposure estimates: counts of people and person-
occurrences. The former counts the number of individuals exposed one or more times per 63
season to the exposure indicator (e.g., exposure level and ventilation rate) of interest. In the case
where the exposure indicator is a benchmark concentration level, the model estimates the number
of people who experience that level of air pollution, or higher, at least once during the modeled
period. The person-occurrences measure counts the number of times per 63 season that an
individual is exposed to the exposure indicator of interest and then accumulates counts over all
individuals. Therefore, the person-occurrences measure confounds people and occurrences:
using this measure, 1 occurrence for 10 people is counted the same as 10 occurrences for 1
person.
3.5 Selection of Urban Areas
The selection of urban areas to include in the exposure analysis takes into consideration
the location of Oj, field and epidemiology studies, the availability of ambient Os data, and the
desire to represent a range of geographic areas, population demographics, and O3 climatology.
These selection criteria are discussed further below in Section 4. Based on these criteria, staff
plans to include the following 12 urban areas in the exposure analysis:
• Boston
• Philadelphia
• New York City
• Washington, D.C.
• Atlanta
• St. Louis
• Chicago
• Houston
• Los Angeles
• Detroit
• Cleveland.
• Sacramento
-------
3.6 Exposure Periods
The exposure periods to be modeled will be the (Vmonitoring seasons for each urban
area. These encompass the periods when high ambient 63 levels are likely to occur, and are the
periods for which routine hourly Os monitoring data are available. The Os seasons for the
selected study areas generally range from April through either September or October for most of
the locations in the eastern U.S. to all year in locations in southern California and Texas.
3.7 Populations to be Analyzed
Exposure modeling will be conducted for the general population residing in each area
modeled, as well as for school-age children (ages 5 to 18), active school-age children, and
asthmatic school-age children. Due to the increased amount of time spent outdoors engaged in
relatively high levels of physical activity, school-age children as a group are particularly at risk
for experiencing CDs-related health effects due to their increased dose rates. Levels of physical
activity will be categorized by a daily Physical Activity Index (PAI), a measure of activity
proportional to the metabolic equivalents of tasks (METS). METS is a unitless ratio of the
energy expended performing an individual task to the person's basal metabolic rate. Children
will be characterized as active if their median daily PAI over the period modeled is 1.75 or
higher, a level characterized by exercise physiologists as being "moderately active" or "active"
(McCurdy, 2000). Data from various national and state surveys undertaken by the Centers for
Disease Control and Prevention (Kami, 2000) will be used to help assure the reasonableness of
the proportions of children that are characterized as active. The proportion of the population of
school-age children characterized as being asthmatic will be estimated by statistics on asthma
prevalence rates.
3.8 Uncertainty and Variability
APEX is a Monte Carlo simulation model which explicitly incorporates the inherent
variability of the model input data. Developing appropriate distributions representing variability
and uncertainty in various model inputs (e.g., air exchange rates, Oj decay rates, physiological
parameters) is a key part of this modeling effort.
The primary difficulty in performing an uncertainty analysis is the quantitative
characterization of the uncertainties of the model inputs and model formulation. We often have
information about the variability of model inputs, and sometimes the variability and uncertainty
combined, but it is usually difficult to estimate the uncertainty separately from the variability.
However, for the APEX Os application, we have enough information to provide reasonable
bounds or ranges for the uncertainties of many of the model inputs. We plan to assess the
impacts of the uncertainties of the model inputs across these ranges, and use these results to
inform a discussion of model uncertainties.
-------
Staff plans to follow a 2-dimensional Monte Carlo/Latin hypercube sampling approach to
a combined variability and uncertainty analysis for APEX. Essentially, a Monte Carlo approach
entails performing many model runs with model inputs randomly sampled from specified
distributions reflecting variability and uncertainty of the model inputs. The 2-dimensional
Monte Carlo method allows for the separate characterization of the variability and uncertainty in
the model results (Morgan and Henri on, 1990).
Due to the large number of APEX input parameters, it is unrealistic to perform a 2-
dimensional Monte Carlo analysis of all of the inputs, due to the large number of model runs that
would be required. We plan to first perform a 1-dimensional Monte Carlo uncertainty analysis
of the model inputs to identify a limited number of input parameters that account for a major part
of the uncertainty. A 2-dimensional analysis of variability and uncertainty would then be
conducted, accounting for the uncertainty of these key inputs and the variability of all of the
inputs.
Uncertainties are inherent in modeled representations of physical reality due to
simplifying assumptions and other aspects of model formulation. The methods for assessing
input parameter uncertainty and model formulation or structure uncertainty are different. It is
difficult to incorporate the uncertainties due to the model formulation into a quantitative
assessment of uncertainty in a straightforward manner. The preferred way to assess model
formulation uncertainty is by comparing model predictions with measured values, while having
fairly complete knowledge of the uncertainty due to input parameters. In the absence of
measurements that can be used to estimate model uncertainty, one must rely on informed
judgment. Our approach to assessing model formulation uncertainty will be to partition this
uncertainty into that of the components, or sub-models, of APEX. For each of the sub-models
within APEX, we will discuss the simplifying assumptions and those uncertainties associated
with the sub-models which are distinct from the input data uncertainties. Where possible, we
will evaluate these sub-models by comparing their predictions with measured data. Otherwise,
we will formulate an informed judgment as to a range of plausible uncertainties for the sub-
models. We will quantitatively assemble the different types of uncertainties and variability to
present an integrated analysis of uncertainty and variability.
The exposure analysis technical support document will provide a more detailed plan for
uncertainty assessment. An analysis of variability will be described in the first draft of the Os
Staff Paper; the uncertainty analysis will be presented in the second draft Os Staff Paper.
4 SCOPE AND APPROACH FOR HEALTH RISK ASSESSMENT
4.1 Overview
The health risk assessment will estimate various health effects associated with Os
exposures for current Os levels, based on 2004 air quality data, as well as reductions in risk
associated with attaining the current 8-h O3 NAAQS and alternative Os standards, based on
adjusting 2002-2004 air quality data. Risk estimates will be developed for 12 urban areas
-------
located throughout the U.S. Health endpoints to be examined in the risk assessment include:
lung function decrements, respirator}' symptoms in asthmatic children, school absences,
emergency department visits for respiratory causes, respiratory- and cardiac-related hospital
admissions, and mortality.
The methods used to conduct the risk assessment and summary results and key findings
from the assessment will be presented in the first draft Oa Staff Paper for current 63 levels and
for just meeting the current 8-h standard. The second draft Os Staff Paper will include risk
estimates associated with just meeting alternative 63 standards. In addition, a health risk
assessment technical support document with a more detailed description of the methodology and
results will accompany the Os Staff Paper.
4.2 Structure of the Risk Assessment
At this time, two general types of human studies are particularly relevant for deriving
quantitative relationships between Os levels and human health effects: controlled human
exposure studies and epidemiological and field studies. Controlled human exposure studies
involve volunteer subjects who are exposed while engaged in different exercise regimens to
specified levels of Os under controlled conditions for specified amounts of time. The responses
measured in such studies have included measures of lung function, such as forced expiratory
volume in one second (FEVt), respiratory symptoms, airway hyperresponsiveness, and
inflammation. As noted above, prior EPA risk assessments for Os have included risk estimates
for lung function decrements and respiratory symptoms based on analysis of individual data from
controlled human exposure studies. For the current health risk assessment, staff plans to use the
probabilistic exposure-response relationships developed during the last review which was based
on analysis of individual data that describes the relationship between a measure of personal
exposure to Ch and the measure(s) of lung function recorded in the study. The measure of
personal exposure to ambient Oj is typically some function of hourly exposures - e.g., 1-hour
maximum or 8-hr maximum. Therefore, a risk assessment based on exposure-response
relationships derived from controlled human exposure study data requires estimates of personal
exposure to Oa, typically on a 1-hour or multi-hour basis. Because data on personal hourly Oi
exposures are not available, estimates of personal exposures to varying ambient concentrations
are derived through exposure modeling, as described above in Section 3.
In contrast to the exposure-response relationships derived from controlled human
exposure studies, epidemiological and field studies provide estimated concentration-response
relationships based on data collected in real woild settings. Ambient Os concentration is
Typically measured as the average of monitor-specific measurements, using population-oriented
monitors. Population health responses for Os have included population counts of school
absences, emergency room visits, hospital admissions for respiratory and cardiac illness,
respiratory symptoms, and premature mortality. As described more fully below, a risk
assessment based on epidemiological studies typically requires baseline incidence rates and
population data for the risk assessment locations.
10
-------
The characteristics that are relevant to the planning and structure of a risk assessment
based on controlled human exposure studies versus one based on epidemiology or field studies
can be summarized as follows:
• A risk assessment based on controlled human exposure studies uses exposure-response
functions, and thus requires estimates of personal exposures. It therefore involves an
exposure modeling step that is not needed in a risk assessment based on epidemiology
or field studies, which uses concentration-response functions.
• Epidemiological and field studies are carried out in specific real world locations (e.g.,
specific urban areas). To minimize uncertainty, a risk assessment based on
epidemiological studies should be performed for the locations in which the studies
were carried out. Controlled human exposure studies, carried out in laboratory
settings, are generally not specific to any particular real world location. A controlled
human exposure studies-based risk assessment can therefore appropriately be carried
out for any location for which there are adequate air quality data on which to base the
modeling of personal exposures. There are, therefore, some locations for which a
controlled human exposure studies-based risk assessment could appropriately be
carried out but an epidemiological studies-based risk assessment could not.
• The adequate modeling of hourly personal exposures associated with ambient
concentrations requires more complete ambient monitoring data than are necessary to
estimate average ambient concentrations used to calculate risks based on
concentration-response relationships. Therefore, there may be some locations in
which an epidemiological studies-based risk assessment could appropriately be carried
out but a controlled human exposure studies-based risk assessment could not.
• To derive estimates of risk or risk reduction from concentration-response relationships
estimated in epidemiological studies, it is usually necessary to have estimates of the
baseline incidences of the health effects involved. Such baseline incidence estimates
are not needed in a controlled human exposure studies-based risk assessment..
Overviews of the scope and methods for each type of risk assessment are discussed below.
4.3 Assessment of Risk Based on Controlled Human Exposure Studies
The major components of the portion of the health risk assessment based on data from
controlled human exposure studies are illustrated in Figure 2. The air quality and exposure
analysis components that are integral to this portion of the risk assessment are discussed above in
Sections 2 and 3, respectively. As described in the draft 63 AQCD (EPA, 2005b), there are
numerous controlled human exposure studies reporting lung function decrements (as measured
by changes in FEVi), other measures of lung function, airway responsiveness, respiratory
symptoms, and various markers of inflammation. Most of these studies have involved voluntary
11
-------
exposures with healthy adults although a few studies have been conducted with mild and
moderate asthmatics and one study reported lung function decrements for children 8-11 years old
(McDonnell et al., 1985).
4.3.1 Selection of health endpoints
In the last review, the health risk assessment estimated both lung function decrements
(2:10, 2:15, and 220% changes in FEV() and respiratory symptoms in children 6-18 years old
associated with 1-hour exposures at moderate and heavy exertion and 8-hour exposures at
moderate exertion. At that time EPA staff and the CAS AC O3 Panel judged that it was
reasonable to estimate the exposure-response relationships for children 6-18 years old based on
data from adult subjects (18-35 years old). As discussed in the 1996 O3 Staff Paper (EPA,
1996a) and 1996 O3 AQCD (EPA, 1996b), findings from other chamber studies (McDonnell et
al., 1985) for children 8-11 year old and summer camp field studies in at least six different
locations in the United States and Canada found lung function changes in healthy children
similar to those observed in healthy adults exposed to O3 under controlled chamber conditions.
Staff intends to use the same approach in this assessment.
In the prior risk assessment, staff estimated risk for lung function decrements associated
with 1-hour heavy exertion, 1-hour moderate exertion, and 8-hour moderate exertion exposures.
Since the 8-hour moderate exertion exposure scenario clearly resulted in the greatest health risks
in terms of lung function decrements, staff plans to include only the 8-hour moderate exertion
exposures in the current risk assessment for this health endpoint.
Although respiratory symptoms in healthy children were estimated in the last review,
staff does not plan to estimate respiratory symptoms in healthy children given the lack of
symptoms found in field studies examining responses in children published since the prior
review. While a number of controlled human exposure studies have been published since the last
review reporting various other acute effects, including airway responsiveness and increases in
inflammatory indicators, none of these studies were conducted at multiple concentration levels
within the range of greatest interest (i.e., below 0.12 ppm). Thus, staff plans to limit this portion
of the risk assessment to lung function decrements in children and to again base the exposure-
response relationships on data obtained for 18-35 year old subjects.
4.3.2 Selection of exposure-response functions
Staff plans to use the same methodology used in the prior risk assessment (see
Appendices A and B in Whitfield et al., 1996) to estimate probabilistic exposure-response
relationships for lung function decrements associated with 8-h moderate exertion exposures. The
combined data set from the Folinsbee et al. (1988), Horstman et al. (1990), and McDonnell et al.
(1991) studies are used to estimate exposure-response relationships for 8-h exposures. The data
from these controlled human exposure studies are corrected for the effect of exercise in clean air
to remove any systematic bias that might be present in the data attributable to an exercise effect.
Generally, this correction for exercise in clean air is small relative to the total effects measures in
12
-------
the Oa-exposed cases. Regression techniques are then used to fit a function to the data. A
Bayesian approach is used then to characterize uncertainty attributable to sampling error based
on sample size considerations. Response rates are calculated for 21 fractiles (for cumulative
probabilities from 0.05 to 0.95 in steps of 0.05, plus probabilities of 0.01 and 0.99) at a number
of Oa concentrations.
4.3.3 Approach to calculating risk estimates
Staff plans to generate several risk measures for this portion of the risk assessment. In
addition to the estimates of the number of school age children and active children experiencing 1
or more occurrences of a lung function decrement > 15 and > 20% in an 63 season, risk
estimates also will be developed for the total number of occurrences of these lung function
decrements in school age children and active school age children. The mean number of
occurrences per child also will be calculated to provide an indicator of the average number of
times that a responder would experience the specified effect during an Os season.
A headcount risk estimate for a given lung function decrement (e.g., ^20% change in
FEVi) is an estimate of the expected number of people who will experience that lung function
decrement. Since EPA is interested in risk estimates associated with ozone concentrations in
excess of policy relevant background concentrations, staff plans to (1) estimate expected risk,
given the personal exposures associated with "as is" ambient Os concentrations, (2) estimate
expected risk, given the personal exposures associated with estimated background ambient 63
concentrations, and (3) subtract the latter from the former. As shown in Equation 4-1 below, the
headcount risk is then calculated by multiplying the resulting expected risk by the number of
people in the relevant population. Because response rates are calculated for 21 fractiles,
estimated headcount risks are similarly fractile-specific.
The risk (i.e., expected fractional response rate) for the k* fractile, Rk is:
Rt=flPJx(RRls\eJ) - f>,**(*a*k*) (Equation 4-1)
j=\ 1=1
where:
6j = (the midpoint of) the jth category of personal exposure to ozone, given "as is"
ambient 03 concentrations;
ef= (the midpoint of) the ith category of personal exposure to ozone, given background
ambient Os concentrations;
Pj = the fraction of the population having personal exposures to Os concentration of e}
ppm, given "as is" ambient 63 concentrations;
13
-------
/>/ = the fraction of the population having personal exposures to 63 concentration of
efppm. given background ambient Os concentrations;
RRk | BJ. = k-fractile response rate at 63 concentration 3;
RRk | ef= k-fractile response rate at Os concentration ef; and
N = number of intervals (categories) of Os personal exposure concentration, given "as is"
ambient 63 concentrations; and
Nb = number of intervals of 0} personal exposure concentration, given background
ambient Os concentrations.
For example, if the median expected response rate given "as is" ambient concentrations is
0.065 (i.e., the median expected fraction of the population responding is 6.5%) and the median
expected response rate given background ambient concentrations is 0.001 (i.e., the median
expected fraction of the population responding is 0.1%), then the median expected response rate
associated with "as is" ambient concentrations above policy relevant background concentrations
is 0.065 - 0.001 = 0.064. If there are 300,000 people in the relevant population, then the
headcount risk is 0.064 x 300,000 = 19,200.
4.3.4 Selection of urban areas
Staff plans to develop lung function decrement risk estimates for school age children and
active school age children living in 12 urban areas in the U.S. These areas, identified previously
in Section 3.2, represent a range of geographic areas, population demographics, and 63
climatology. As discussed further in Section 4.4.2, the selection of these areas was also
influenced by whether other health endpoints could be examined in the same urban area based on
concentration-response relationships developed from epidemiological or field studies.
4.4 Assessment of Risk Based on Epidemiological and Field Studies
As discussed in the draft 03 AQCD (EPA, 2005b), a significant number of
epidemiological and field studies examining a variety of health effects associated with ambient
Os concentrations in various locations throughout the U.S., Canada, Europe, and other regions of
the world have been published since the last NAAQS review. As a result of the availability of
these epidemiological and field studies and air quality information, staff plans to expand the Oi
risk assessment to include an assessment of selected health risks attributable to ambient 03
concentrations over policy relevant background concentration and health risk reductions
associated with attainment of current and alternative Os standards in selected urban locations in
the U.S.. The major components of the portion of the health risk assessment based on data from
epidemiological and field studies are illustrated in Figure 3. The approaches used by staff to
14
-------
select health endpoint categories, urban areas, and epidemiology and field studies to consider for
inclusion in the risk assessment are discussed below.
4,4.1 Selection of health endpoints
Staff has carefully reviewed the epidemiological evidence evaluated in Chapter 7 and
summarized in Chapter 7 Annex of the draft 03 AQCD (EPA, 2005 b). Tables AX7-1 through
AX7-5 summarize the available U.S. and Canadian studies of the effects of acute (short-term)
exposures for various health effect categories. Given the substantial number of health endpoints
and studies addressing Os effects, staff plans to include in this quantitative Oj risk assessment
only the better understood (in terms of health consequences) health endpoint categories for
which the weight of the evidence supports the inference of a causal relationship between 0? and
the effect category. In addition, staff plans to include only those categories for which there are
studies that satisfy the study selection criteria discussed below.
Based on its review of the evidence evaluated in the draft Oa AQCD, staff is considering
including in the epidemiology and field studies-based portion of the Os risk assessment the
following broad categories of health endpoints associated with short-term exposures:
• respiratory symptoms in asthmatic children;
• school absences;
• emergency department visits for respiratory illness;
• hospital admissions for respiratory illness;
• unscheduled hospital admissions for respiratory illness; and
• premature total, respiratory, and cardiovascular mortality.
4.4.2. Selection of urban areas
Several objectives were considered in selecting potential urban areas for which to
conduct the epidemiological studies-based 63 risk assessment. Staff plans to include an urban
area only if it satisfies the following criteria:
• It has sufficient air quality data for a recent year (2002 or later).
• It is the same as or close to the location where at least one concentration-response
function for one of the recommended health endpoints (see above) has been
estimated by a study that satisfies the study selection criteria (see below).
• For the hospital admission effects categories, relatively recent location-specific
baseline incidence data, specific to International Classification of Disease (ICD)
codes, are available.4
4 The absence of hospital admissions baseline incidence data does not necessarily mean that we cannot use an urban
area in the risk assessment, only that we cannot use it for the hospital admissions endpoint.
15
-------
Because baseline mortality incidence data are available at the county level, this is not a
constraint in the selection of urban areas for the Oj risk assessment. Information on the
incidence of respiratory symptoms and illnesses not requiring hospitalization, in contrast, is
generally not available, except in those locations in which studies were conducted. Data on
hospital admissions for recent years, however, specific to ICD codes, are available in some cities
but not others. This category of incidence data is therefore a consideration in the selection of
urban areas to include in the analysis.
In addition, staff plans to take into account the following considerations in selecting from
among those urban locations that satisfy the above selection criteria:
• Locations with more health endpoints are preferred over those with fewer.
• The overall set of urban locations should represent a range of geographic areas,
population demographics, and 63 climatology within the U.S..
Based on the selection criteria and additional considerations listed above, staff plans to
include the following urban areas in our assessment of risk based on epidemiological and field
studies:
• Boston
• Philadelphia
• New York City
• Washington, D.C.
• Atlanta
• St. Louis
• Chicago
• Houston
• Los Angeles
• Detroit
• Cleveland
• Sacramento
4.4.3 Selection of epidemiological and field studies
As discussed above, staff plans to include in the Os risk assessment only the better
understood health effects for which the weight of the evidence supports a causal inference.
Thus, in cases where none of the available studies reported a statistically significant relationship,
the effect endpoint would not be included. Once it has been determined that a health endpoint
will be included in the analysis, however, inclusion of a study on that health endpoint will not be
based on statistical significance. That is, consistent with the approach being taken in the
particulate matter (PM) risk assessment (see EPA, 2005c, Chapter 4, and Abt Associates, 2005),
no credible study on an included health endpoint has been excluded from the analysis on the
basis of lack of statistical significance.
-------
Staff has applied the following selection criteria for any study that has estimated one or
more O3 concentration-response functions for a selected health endpoint in an urban location to
be used for the 03 risk assessment:
• It is a published, peer-reviewed study that has been evaluated in the draft 03 AQCD (EPA,
2005b) and judged adequate by staff for purposes of inclusion in this risk assessment
based on that evaluation.
• It directly measured, rather than estimated, O3 on a reasonable proportion of the days in
the study.
• It either did not rely on Generalized Additive Models (GAMs) using the S-Plus software
to estimate concentration-response functions or has appropriately re-estimated these
functions using revised methods.5
• For short-term mortality studies, that the study reported results for the O3 season.
Staff notes that the draft 03 AQCD is currently under review by the CASAC 03 Panel and the
general public. Accordingly, the final group of studies to be included in the planned risk
assessment may change based on the advice and recommendations resulting from this review.
4.4.4 A summary of selected health endpoints, urban areas and studies
Based on applying the criteria and considerations discussed above, the health endpoints,
urban locations, and epidemiology and field studies that staff plans to include in the O3 risk
assessment are given in Table I. More detail on the studies is given in Table 2.
5 The GAM S-Plus problem was discovered prior to the recent PM risk assessment that was carried out as part of the
PM NAAQS review. It is discussed in the PM Criteria Document (EPA, 2004), second draft PM Staff Paper (EPA,
2005c), and draft PM Health Risk Assessment (Abt Associates, 2005).
17
-------
Table 1. Locations and Health Endpoints Considered for Inclusion in the Oj Risk
Assessment Based on Epidemiological and Field Studies
Urban Area
Boston
Philadelphia
New York
Washington,
D.C.
Atlanta
St. Louis
Chicago
Houston
Los Angeles
Sacramento
Detroit
Cleveland
Short-term
Exposure
Mortality
BellA'
BellA'
Huang8"
MoolgavkarN
BellA' mm
Huang8
BellA>
BellA"
Huang8"
BellA'
BellA*
Huang8"
Schwartz0
BellA'
Huang8" _
Schwartz0'
BellA'
Huang8
BellA*
BellA<
Huang8"
Schwartz*"
ItoD
BellA'
Huang8"
Respiratory
Hospital
Admissions
Thiirston0
LinnF
Schwartz6
Emergency Room
Visits for
Respiratory Illness
Tolbert"
Friedman1
Peel'
JaffeM
School
Absences
Gilliland*
Respiratory
Symptoms
MortimerL"
MortimerL"
MortitnerL
Mortimer1""
Mortimer1'
Study reports multi-city results based on a set of cities including city listed in this row. Single-city results have
been obtained from the authors.
** Study reports multi-city results based on a set of cities including city listed in this row. Single-city results are
also reported.
A Bell et al. (2004) H Tolbert et al. (2000)
B Huang et al. (2004)
c Schwartz (2004)
DIto(2003) KGillilandetal. (2001)
E Schwartz et al. (1996) L Mortimer et al. (2002)
F Linn et al. (2000) MJaffe et al. (2003)
G Thurston et al. (1992) NMoolgavkar et al. (1995)
1 Friedman etal. (2001)
JPeel et al. (2005)
18
-------
Table 2. Overview of Ozone Epidemiologies! and Field Studies Considered for Inclusion in die 03 Health Risk Assessment
Study
Location(s)
Health Endpoint
Other Pollutants in
Model
Analyses Limited to
Ozone Season?
Statistically Significant Effects?
Mortality
Bell etal., 2004
Huang et al., 2004
Moolgavkar et al,,
1995
Schwartz, 2004
Ito, 2003
multi-city function
based on 95 cities
19 U.S. cities
multi-city with single
city and multi-city
estimates
Philadelphia
14 U.S. cities
Detroit, MI
total (non-injury) mortality
cardiovascular and
respiratory mortality
cardiovascular and
respiratory mortality
total non-accidental
mortality
total non-accidental
mortality
Total, circulatory, and
respiratory mortality
No*
CO, SO2.NO2, PM10-
each in a separate 2-
pollutant model with O3
TSP, S02
PM10
PM10
Yes
Yes (summer only)
Yes
Yes
Yes
Yes
Yes - for single-pollutant model
Mixed - for 2-pollutant models
Yes - for summer
Yes- for full yr and warm season,
but not cold season
Yes
Hospital admissions
Linn, etal. 2000
Schwartz et al, 1996
Thurston et al.,
1992
Los Angeles, CA
Cleveland, OH
New York City, NY
Hospital admissions for
pulmonary illness among
people age 30+
Hospital admission for
resp. illness among people
age 65+
unscheduled hospital
admission for respiratory
illness
No
PM10,CO,NO2
No
No
Yes
Yes
Yes
Yes (summers only)
No
No
Yes
Yes
-------
to
o
Study
Location(s)
Health Endpoint
Other Pollutants in
Model
Analyses Limited to
Ozone Season?
Statistically Significant Effects?
Emergency room visits for:
Friedman, et al.
2001
Jaffe et al., 2003
Peel et al. 2005
Tolbert et al. 2000
Atlanta, GA
Cleveland, OH
Cleveland, Cincinnati,
and Columbus
combined
Atlanta, GA
Atlanta, GA
"acute care events" for
asthma among children
ages 1 - 16***
Asthma, among people
ages 5-34
all respiratory illnesses
Asthma
Pneumonia
COPD
UR1
Pediatric ER visits for
asthma
No
No
Yes******
Y(,s******
Yes******
Yes******
Yes
No
Yes (June 21 - Sept.
1)
Yes (June - August)
Yes
Yes (summer only)
Yes - for 2-day and 3-day
cumulative exposures tor some
databases***
No
Yes
No
No
No
Yes
Yes
School absenteeism
Gilliland, et al.
2001
Los Angeles, CA
due to all illness; resp.
illness; upper resp. illness;
lower resp. illness, among
4th grade children.
No
No
(January-June)
Yes for all categories
Respiratory symptoms
Gent et al.
2003****
CT and Springfield, MA
respiratory symptoms in
asthmatic children under
1 2 (at time of enrollment
in study)
No
Yes (April-
September)
Yes - for some symptoms, among
medicated users only
-------
Study
Mortimer, et al.
2002
Location(s)
multi-city (8 locations in
U.S.)
Health Endpoint
Morning asthma
symptoms in asthmatic
children 4-9 years old
Other Pollutants in
Model
PM10, SO2.NO2
Analyses Limited to
Ozone Season?
Yes (June- August)
Statistically Significant Effects?
Yes - for single-pollutant model
only
* The authors report that the results were robust to adjustment for PMm, but do not report the multi-pollutant functions.
** This study was carried out using GAM S-Plus before the GAM S-Plus problem was realized. The data were reanalyzed as part of the general HEl-sponsored
reanalysis of studies that used GAM S-Plus (see Ito 2003), but with an emphasis on PM. Reanalyzed results for O3 are presented graphically in the publication
but numerical results have been obtained from the author.
* ** All results are given separately for the 4 separate data sources - Georgia Medicaid claims file; Health maintenance organization; Pediatric emergency
departments, and Georgia Hospital Discharge Database. The types of asthma events covered by the sources are: emergency care and hospitalizations; emergency
care, urgent care, and hospitalizations; emergency care and hospitalizations; and hospitalizations, respectively.
****Asthmatic children were divided into 2 groups: those who use maintenance medication and those who do not. This study can be used only if we are able to
obtain information about the proportion of asthmatic children who use maintenance medication.
*****A11 exposures were lagged 3 days.
******Authors conducted multi-pollutant analyses but did not present quantitative results in publication for this health endpoint.
-------
4.4.5 Selection of concentration-response functions
Studies often report more than one estimated concentration-response function for the
same location and health endpoint. Sometimes models including different sets of co-pollutants
are estimated in a study; sometimes different lags are estimated. In some cases, two or more
different studies estimated a concentration-response function for 63 and the same health endpoint
in the same location (this is the case, for example, with Os and mortality associated with short-
term exposures). For some health endpoints, there are studies that estimated multi-city 63
concentration-response functions, while other studies estimated single-city functions.
All else being equal, staff judges that a concentration-response function estimated in the
assessment location is preferable to a function estimated elsewhere, since it avoids uncertainties
related to potential differences due to geographic location. That is why the urban areas selected
for the epidemiological studies-based Os risk assessment are those locations in which
concentration-response functions have been estimated. There are several advantages, however,
to using estimates from multi-city studies versus studies carried out in single cities. Multi-city
studies are applicable to a variety of settings, since they estimate a central tendency across
multiple locations. When they are estimating a single concentration-response function based on
several cities, multi-city studies also tend to have more statistical power and provide effect
estimates with relatively greater precision than single city studies due to larger sample sizes,
reducing the uncertainty around the estimated coefficient. Because single-city and multi-city
studies have different advantages, if a single-city concentration-response function has been
estimated in a risk assessment location and a multi-city study which includes that location is also
available for the same health endpoint, staff plans to use both functions for that location in the
risk assessment.
Several Os epidemiological studies estimated concentration-response functions in which
03 was the only pollutant entered into the health effects model (i.e., single pollutant models) as
well as other concentration-response functions in which O^ and one or more co-pollutants (e.g.,
PM, nitrogen dioxide, sulfur dioxide, carbon monoxide) were entered into the health effects
model (i.e., multi-pollutant models). To the extent that any of the co-pollutants present in the
ambient air may have contributed to the health effects attributed to OT, in single pollutant models,
risks attributed to Os might be overestimated where concentration-response functions are based
on single pollutant models. However, if co-pollutants are highly correlated with 63, their
inclusion in an O3 health effects model can lead to misleading conclusions in identifying a
specific causal pollutant. When collinearity exists, inclusion of multiple pollutants in models
often produces unstable and statistically insignificant effect estimates for both Oj and the co-
pollutants. Given that single and multi-pollutant models each have both potential advantages and
disadvantages, with neither type clearly preferable over the other in all cases, staff plan to report
risk estimates based on both single and multi-pollutant models where both are available.
Epidemiological and field studies often present several concentration-response functions,
each incorporating a different lag structure. The question of lags and the problems of correctly
specifying the lag structure in a model have been discussed extensively [see, for example, the
22
-------
PM AQCD (EPA, 2004, section 8.4.4); the second draft PM Staff Paper (EPA, 2005c, sections
3.5.5.2 and 4.2.6.3); the draft O3 AQCD (EPA, 2005b, section 7.1.3.3); and Schwartz, 2000).
The draft Oa AQCD notes that "simply choosing the most significant exposure lag may bias the
air pollution risk estimates away from the null..." (EPA, 2005b, section 7.1.3.3). On the other
hand, there is recent evidence (Schwartz, 2000) that the relationship between PM and health
effects may best be described by a distributed lag (i.e., the incidence of the health effect on day n
is influenced by PM concentrations on day n, day n-1, day n-2 and so on). If this is true for Os as
well, then a model with only a single lag may bias air pollution risk estimates towards the null.
For mortality associated with short-term exposure to 63, Huang et al. (2004) present the results
for a distributed lag model that takes into account exposure from the previous 6 days. When a
study reports several single lag models for a health effect, staff plans to base our initial selection
of the appropriate lag structure for each health effect on the overall assessment provided in the
draft 03 AQCD, based on all studies reporting concentration-response functions for that health
effect.
In summary:
• if a single-city concentration-response function has been estimated in a risk assessment
location and a multi-city study which includes that location is also available for the same
health endpoint, staff plans to use both functions for that location in the risk assessment;
• risk estimates based on both single and multi-pollutant models will be used where both are
available;
• where available, distributed lag models will be used; when a study reports several single
lag models for a health effect, staff plans to base our initial selection of the appropriate lag
structure for the health effect on the overall assessment in the draft 63 AQCD, based on
all studies reporting concentration-response functions for that health effect.
4.4.6 Baseline health effects incidence considerations
The most common epidemiologically-based health risk model expresses the reduction in
health risk (Ay) associated with a given reduction in Os concentrations (Ax) as a percentage of
the baseline incidence (y). To accurately assess the impact of 63 air quality on health risk in the
selected urban areas, information on the baseline incidence of health effects (i.e., the incidence
under "as is" air quality conditions) in each location is therefore needed. Where at all possible,
staff plans to use county-specific incidences or incidence rates (in combination with county-
specific populations). Estimates of location-specific baseline mortality rates can be obtained for
each of the Os risk assessment locations for 2001 from CDC Wonder, an interface for public
health data dissemination from the Centers for Disease Control (CDC).6
Hospital admissions studies being considered for inclusion in the Os risk assessment were
conducted in Los Angeles, New York City, and Cleveland. ICD code-specific baseline hospital
6 See http://wonder.cdc.gov/.
23
-------
admission rates for Los Angeles County can be obtained from California's Office of Statewide
Health Planning and Development, which provided records of hospital admissions for the study
by Linn et al. (2000). The records provided for the Linn study included both ICD codes and All-
Patient-Refmed Diagnosis-Related Group (APR-DRG). Because Linn et al. (2000) used
diagnosis categories based on the APR-DRG, staff will ensure that information obtained from
California's Office of Statewide Health Planning and Development also contains the APR-DRG
so that the baseline incidence rates are calculated for hospital admissions categories that match
those used in the Linn study.
Schwartz et al. (1996) report several percentiles as well as the mean of the distribution of
daily hospital admissions for respiratory illness (ICD-9 codes 460-519) in Cleveland, Ohio
during the years 1988-90 (although the source of the information is not given). Staff plans to
investigate the possibility of updating these baseline incidence rates.
Thurston et al. (1992) report 1990 population, as well as average unscheduled hospital
admissions per day for respiratory illnesses and for asthma separately, for 1988 and for 1989 in
New York City (and other cities in New York State), based on data on unscheduled (emergency)
hospital admissions collected by the Statewide Planning and Research Cooperative System
(SPARCS), a division of the New York State Department of Health. Baseline incidence rates for
1989/90 can be calculated using the data presented by Thurston et al. (1992). Alternatively, staff
also plans to investigate the possibility of obtaining more recent data on unscheduled hospital
admissions for respiratory illness and asthma in New York City from SPARCS with which to
calculate more recent baseline incidence rates.
Tolbert et al. (2000) and Peel et al. (2005) report average daily emergency department
visits for the relevant health effects in the facilities participating in their studies in the Atlanta
area, which, in both studies, are reported to cover about 80 percent of the relevant emergency
department visits. Staff plans to use either the average daily rates reported in the studies, in
which case the baseline incidence will be understated; or alternatively, to adjust the baseline
incidence upward, based on the authors' assessment that the facilities included in their studies
cover about 80 percent of emergency room visits. It is less clear at this time whether baseline
incidence can be constructed or obtained for the study by Friedman et al. (2001), which
considers "acute asthma events" separately in five databases (see Table 2).
For other morbidity endpoints, such as respirator)' symptoms in children, incidence
information aggregated at higher than the city- or county-level may be all that is available. Staff
plans to use the level of aggregation closest to county-specific; however, for some morbidity
endpoints, it may be necessary to estimate county-specific incidence using national-level
incidence rates. For some health endpoints, there may be no information on incidence other than
the information provided for the city or county in which the concentration-response function was
estimated. The rationale for the choice of incidence data used for each health endpoint in each
location will be presented in the risk assessment technical support document.
Lack of location-specific incidence data will increase the uncertainty surrounding
estimates of risk for the specific cities selected for the risk assessment. To the extent possible,
24
-------
staff plans to provide a quantitative comparison to help assess the accuracy of using incidence
rates at a higher level of aggregation (e.g., national incidence rates) by comparing these rates to
county-specific incidence rates where these are available.
4.4.7 Assessing risk in excess of policy-relevant background
As noted above, staff plans to assess risks associated with 63 concentrations in excess of
policy -relevant background concentrations, and to assess risk reductions associated with just
meeting current and alternative Oj standards. Following the methods used in the prior Oa risk
assessment, risks based on a concentration-response function estimated in an epidemiological or
field study will be assessed down to the estimated policy relevant background.
To assess risks associated with Oa concentrations in excess of policy-relevant background
concentrations, staff will first calculate the difference between "as is" 63 levels and policy-
relevant background. Staff will then calculate the corresponding change in incidence of the
health effect associated with that change in ambient Os concentration. If Ax denotes the change
in Oj level from "as is"concentration to the background concentration, then the corresponding
change in incidence of the health effect, Ay, for a log-linear concentration-response function (the
most common functional form), is
= y * [e ?* - 1] (Equation 4-2)
where y denotes the baseline incidence (discussed above in Section 4.4.6) and P is the coefficient
of Os in the concentration-response function. A similar calculation would be made if the
concentration-response function is of a logistic form.
To assess the risk reduction associated with just meeting the current standard in those
locations that do not currently meet this standard, the procedure will be the same, except that in
this part of the risk assessment Ax will be the difference between "as is" 63 levels and the Os
levels that will be estimated to exist if the current standards are just met.
To assess the risk reductions associated with just meeting alternative, more stringent
standards, above and beyond the risk reductions that would be achieved by just meeting the
current standards, Ax will be the difference between 63 levels that will be estimated to exist if
the current standards are just met and 63 levels that will be estimated to exist if the alternative,
more stringent, standards are just met.
Because the 63 coefficient, P, is estimated rather than known, there is uncertainty
surrounding that estimate. This uncertainty is characterized as a normal distribution, with mean
equal to the Os coefficient reported in the study, and standard deviation equal to the standard
error of the estimate, also reported in the study. From this information, staff plans to construct a
95 percent confidence interval around the reported risk or risk reduction (number of cases of the
health effect avoided), following the method used in the draft PM risk assessment (Abt
Associates, 2005).
25
-------
4.5 Uncertainty and Variability
There are several uncertainties that affect the inputs to both the controlled studies-based
and epidemiology-studies based portion of the Os risk assessment. These include uncertainties in
the air quality adjustment procedures used to simulate attainment of standards, policy-relevant
background estimates, exposure estimates, baseline incidence rates, and appropriate model form
for the concentration- and exposure-response relationships used in the risk assessment. There
also is likely city-to-city variability in both exposure estimates, discussed previously in section
3.71, and in concentration-response and exposure-response relationships.
For the portion of the risk assessment based on exposure-response functions derived from
controlled human exposure studies, risk estimates will be prepared that incorporate the
characterization of uncertainty and variability in the exposure estimates discussed previously in
section 3.7.1. In addition, for the exposure-response relationships derived from the controlled
human exposure studies the uncertainties due to sample size considerations also will be included
in these risk estimates. Additional uncertainties for the controlled human exposure studies
portion of the risk assessment will be discussed qualitatively and the most important ones will be
addressed in sensitivity analyses described below.
With respect to the epidemiology-based portion of the risk assessment, the uncertainty
that arises due to sample size considerations, that is reflected in the confidence intervals for the
concentration-response relationships reported in the epidemiology studies, will be incorporated
in this portion of the risk assessment. In the case of short-term exposure mortality, two studies,
Bell et al. (2004) and Huang et al. (2004), provide both city-specific and overall multi-city mean
estimates. In a prior risk assessment for PM, EPA used an Empirical Bayes technique to adjust
location-specific estimates and their standard errors (Post et al., 2001). This approach effectively
moves the city-specific estimates towards the overall mean; the larger the city-specific standard
error (relative to the inter-city variability), the more the city-specific estimate is moved towards
the overall mean of the distribution. This adjustment more efficiently uses the information in the
study to yield estimates of the Os coefficient in each location and, thus better addresses concerns
about city-to-city variability. The Bell et al. (2004) city-specific estimates already reflect this
type of approach. Staff plans to use a similar approach to incorporate the Huang et al. (2004)
concentration-response coefficients, which are for a distributed lag model, in this 63 risk
assessment.
Staff also notes that several meta-analyses addressing the impact of various factors on
estimates of mortality associated with short-term exposures to Oa have recently been accepted
for publication and will be published in June 2005. Staff plans to review these analyses and
explore whether they provide additional information that can be used to assist in characterizing
the uncertainties associated with risk estimates for this health outcome.
For other sources of uncertainty in both the controlled human exposure-based and
epidemiology-based portions of the risk assessment (e.g., the use of alternative model forms for
the C-R function), there is insufficient information to incorporate these uncertainties
probabilistically into the risk assessment. Staff plans to include sensitivity analyses, briefly
26
-------
described in Table 3 below, to help characterize how important these uncertainties are to the risk
estimates. These sensitivity analyses are designed to show the impact of changing the values of
the most important uncertain inputs or assumptions underlying the analysis on the results of the
OB risk assessment. The air quality-related uncertainties impact both the controlled human
exposure and epidemiology-based portions of the risk assessment. The uncertainty about
exposure-response relationships affects the controlled human exposure based portion of the
assessment. Uncertainties about baseline incidence and concentration-response relationships
derived from epidemiology studies impacts the epidemiology-based portion of the risk
assessment.
Table 3. Planned Sensitivity Analyses
Component of the
Risk Assessment
Air Quality
Air Quality
Exposure-Response
Baseline Incidence
Concentration-
Response
Sensitivity Analysis
A sensitivity analysis of the effect of different assumptions about background O3
levels on estimated risks associated with "as is" levels of O3 above background
levels
A sensitivity analysis of the effect of different air quality adjustment procedures
on the estimated risk reductions resulting from just meeting the current 8-h
standard and alternative standards
A sensitivity analysis of the effects of alternative extrapolations of exposure-
response models (below the lowest exposure levels used in the laboratory
studies), including possible alternative hypothetical thresholds, on the estimated
lung function risks (e.g., percentages of people experiencing lung function
decrements of at least 10%, 20%, etc.) associated with "as is" levels of O3 above
background levels and risk reductions associated with just meeting alternative
standards
A comparison of using more aggregate baseline incidence data (national, state,
etc) versus county-specific information in the county with the best local baseline
incidence data
A sensitivity analysis of the effects of alternative hypothetical thresholds on
estimated risks associated with "as is" levels of O3 above background levels and
risk reductions associated with just meeting alternative standards
27
-------
SCHEDULE AND MILESTONES
Table 4 below includes the key milestones for the exposure analysis and health risk
assessment that will be conducted as part of the current O3 NAAQS review. A consultation with
the CASAC 03 Panel is planned for May 5, 2005 to obtain input on this draft Scope and Methods
Plan. Staff will then proceed to develop exposure and health risk estimates associated with
recent O3 levels and levels adjusted to just meet the current 8-hour O3 standard. These estimates
and the methodology used to develop them will be discussed in the first draft O3 Staff Paper and
in separate exposure analysis and risk assessment technical support documents. These draft
reports will be released for CASAC and public review in conjunction with the release of the first
draft O3 Staff Paper by the end of September 2005. EPA will receive comments on these draft
documents from the CASAC O3 Panel and general public at a meeting in December 2005. As
noted earlier in section 3.7.1, staff anticipates including a fuller treatment of uncertainty and
variability in the revised exposure analysis and health risk assessment that will be prepared
following the December 2005 CASAC review meeting. The revised exposure analysis and risk
assessment reports will also include estimates associated with just meeting any alternative
standards that may be recommended by staff for consideration. The revised analyses will be
released in April 2006 in conjunction with a second draft O3 Staff Paper for review by CASAC
and public at a meeting to be held in July 2006. Staff will consider these review comments and
prepare final exposure analysis and risk assessment reports by September 2006.
Table 4. Key Milestones for the Exposure Analysis and Health Risk Assessment for the Oj
NAAQS review
Milestone
Release 1st draft O3 AQCD
Release draft Scope and Methods Plan
CASAC/public review and meeting on 1st draft O3 AQCD
CASAC consultation on draft Scope and Methods Plan
Release 2nd draft 03 AQCD
Release 1st drafts of the O3 Staff Paper and the Exposure Analysis
and Risk Assessment reports
CASAC/public review and meeting on 2nd draft 03 AQCD and 1st
drafts of the O3 Staff Paper and the Exposure Analysis and Risk
Assessment reports
Final O3 AQCD
Release 2nd drafts of the O3 Staff Paper and the Exposure Analysis
and Risk Assessment reports
CASAC/public review and meeting on 2nd drafts of the O3 Staff
Paper and the Exposure Analysis and Risk Assessment reports
Final O3 Staff Paper, Exposure Analysis, and Risk Assessment
Date
January 3 1,2005
April 2005
May 4-5, 2005
May 5, 2005
August/September 2005
September 2005
December 2005
February 28, 2006
April 2006
July 2006
September 2006
28
-------
6 REFERENCES
Abt Associates Inc. (2005). Paniculate Matter Health Risk Assessment for Selected Urban
Areas: Second Draft Report, January. Prepared for Office of Air Quality Planning and
Standards, U.S. Environmental Protection Agency, Research Triangle Park, NC. Available
electronically on the internet at: http://www.epa.gov/ttn/'naaqs/stajndards/pnv's_pin__cr_td.hlnil.
Bell, M,A. McDermott, S.L. Zeger, J.M. Samet, and F. Dominici (2004). "Ozone and short-term
mortality in 95 US urban communities, 1987-2000." JAMA 292(19):2372-2378.
EPA (1996a). Review of National Ambient Air Quality Standards for Ozone: Assessment of
Scientific and Technical Information - OAQPS Staff Paper. EPA/452/R-96-007. Office of Air
Quality Planning and Standards, Research Triangle Park, NC. Available electronically on the
internet at: hup://www.epa.gov/ltt)/naaqs/standards/o/.one/s_o3_pr_sp.html.
EPA (1996b). Air Quality Criteria for Ozone and Related Photochemical Oxidants.
EPA/600/P-93/004aF-cF. Office of Research and Development, National Center for
Environmental Assessment, Research Triangle Park, NC. Available electronically on the
internet at: littp://'cfpub.epa.g.o.v/ncea/cfm/recordisplay.cfm?deid=2831..
EPA (2002). Consolidated Human Activities Database Users Guide. The database and
documentation are available electronically on the internet at: htlp:/Avww.epa.aov/chadnet J /.
EPA (2003). Total Risk Integrated Methodology TRIM. Expo/Inhalation User's Document
Volume I: Air Pollutants Exposure Model (APEX, version 3) User's Guide. Office of Air Quality
Planning and Standards, Research Triangle Park, NC. Available electronically on the internet at:
http://www.epa.gov/ttn/fera/human apex.html.
EPA (2004). Air Quality Criteria for Paniculate Matter. EPA 600/P-99/002bF, 2v. National
Center for Environmental Assessment, Research Triangle Park, NC. Available electronically on
the internet at: htcp://\vww.epa.&ov/ttn/naaqs/standards/ptn/s pm crcd.html
EPA (2005a). Plan for Review of the National Ambient Air Quality Standards for Ozone.
Office of Air Quality Planning and Standards, Research Triangle Park, NC. March. Available
electronically on the internet at
http://vvvvw.epa.gov/tln/r)aaqi;/!>tandard.s/o^one/s_o3_cr_pd.htrnl.
EPA (2005b). Air Quality Criteria for Ozone and Other Related Photochemical Oxidants. First
External Review Draft. National Center for Environmental Assessment, Research Triangle Park,
NC. Available electronically on the internet at:
http://cfB.ub .epa.aov/ucea^fm/recordisplay .cfm?deid= 114523.
29
-------
EPA (2005c). Review of National Ambient Air Quality Standards for Paniculate Matter:
Assessment of Scientific and Technical Information - OAOPS Staff Paper (second draft). EPA-
452/D-05-001. Office of Air Quality Planning and Standards, Research Triangle Park, NC.
Available electronically on the internet at:
htlp://u^v\v.epa.gov/lrn/naaqs/standards/prri/s_pm_cr_sp.htnil.
Fiore, A.M., D.J. Jacob, I. Bey, R.M. Yantosca, B.D. Field, A.C. Fusco, and J.G. Wilkinson
(2002a). "Background ozone over the United States in summer: Origin, trend, and contribution
to pollution episodes."! Geophys. Res., 107(D15), 4275.
Fiore, A.M., D.J. Jacob, B.D. Field, D.G. Streets, S.D. Fernandes, and C. Jang (2002b).
"Linking ozone pollution with climate change: The case for controlling methane." Geophys. Res.
Lett., 29(19), 1919.
Fiore, A.M., D.J. Jacob, H. Liu, R.M. Yantosca, T.D. Fairlie, and Q. Li (2003). "Variability in
surface ozone background over the United States: Implications for air quality policy." Journal
Of Geophysical Research Vol. 108(D24), 4787.
Friedman, M.S., K.E. Powell, L. Hutwagner, L.M. Graham, and W.G. Teague (2001). "Impact
of changes in transportation and commuting behaviors during the 1996 summer Olympic games
in Atlanta on air quality and childhood asthma" JAMA 285:897-905.
Fusco, A. C., and J. A. Logan (2003). "Analysis of 1970- 1995 trends in tropospheric ozone at
Northern Hemisphere midlatitudes with the GEOSCHEM model." J. Geophys. Res., 108(D15),
4449.
Gent, J.F., E.W. Triche, T.R. Holford, K. Belanger, M.B. Bracken, W.S. Beckett, B.P. Leaderer
(2003). "Association of low-level ozone and fine particles with respiratory symptoms in children
with asthma." JAMA 290(14):1859-1867.
Graham, S. and T. McCurdy (2004). "Developing meaningful cohorts for human exposure
models. " Journal of Exposure Analysis and Environmental Epidemiology 14:23-43.
Gilliland, F.D., K. Berhane, E.B. Rappaport, D.C. Thomas, E. Avol, W.J. Gauderman, S.J.
London, H.G. Margolis, R. McConnell, K.T. Islam and J.M. Peters (2001). "The effects of
ambient air pollution on school absenteeism due to respiratory illnesses." Epidemiology
12(l):43-54.
Horstman, D.H. et al. (1990). "Ozone concentration and pulmonary response relationships for
6.6-hour exposures with five hours of moderate exercise to 0.08, 0.10, and 0.12 ppm." American
Review of Respiratory Disease 142:1158-1163.
Huang, Y., F. Dominici, M.L. Bell (2004). "Bayesian hierarchical distributed lag models for
summer ozone exposure and cardio-respiratory mortality." John Hopkins University,
Department of Biostatistics Working Paper. 46.
30
-------
Ito, K. (2003). Associations of particulate matter components with daily mortality and morbidity
in Detroit, Michigan. In: "Revised Analyses of Time-Series Studies of Air Pollution and
Health," Health Effects Institute Special Report, May.
Jaffe, D.H., Singer, M.E., and Rimm, A.A. (2003). "Air pollution and emergency department
visits for asthma among Ohio Medicaid recipients, 1991-1996." Environmental Research 91:21-
28.
Johnson, T., Capel, J., and McCoy, M. (1996a). Estimation of Ozone Exposures Experienced by
Urban Residents Using a Probabilistic Version ofNEMand 1990 Population Data. Prepared by
International Technology Air Quality Services for Office of Air Qualliry Planning and Standards,
EPA, Research Triangle Park, NC. Available electronically on the internet at:
http://www.epa.ao\'/ttn/naaqs/standards/ozone/s o3 pr td.htmi.
Johnson, T, Capel, J., McCoy, M., and Wamasch, J. (1996b). Estimation of Ozone Exposures
Experienced by Outdoor Children in Nine Urban Areas Using a Probabilistic Version ofNEM.
Prepared by International Technology Air Quality Services for Office of Air Quallity Planning
and Standards, EPA, Research Triangle Park, NC. Available electronically on the internet at:
http://www.epa.gov/un/naaqs/standards/o/one/s_o3j)r td.htmi.
Johnson, T. (1997). "Sensitivity of Exposure Estimates to Air Quality Adjustment Procedure,"
Letter to Harvey Richmond, Office of Air Quality Planning and Standards, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina.
Kann, L. et al. (2000). "Youth risk behavior surveillance-United States, 1999." Mortality and
Morbidity Weekly Report 49(S805): 1-96.
Linn, W., Y. Szlachcis, H.J. Gong, P. Kinney, K. Berhane (2000). "Air pollution and daily
hospital admissions in metropolitan Los Angeles." Environmental Health Perspective 108:427-
434.
McCurdy, T. (2000). "Conceptual basis for multi-route intake dose modeling using an energy
expenditure approach." J. Exposure Anal. Environ. Epidemiol. 10:1-12.
McCurdy, T., Glen, G., Smith, L., Lakkadi, Y. (2000). "The National Exposure Research
Laboratory's Consolidated Human Activity Database."/ Exposure Anal. Environ. Epidemiol.
10:566-578.
McDonnell, W.F. et al. (1985). "Reproducibility of individual responses to ozone exposure."
American Review of Respirtory Disease 131:36-40.
McDonnell, W.F. et al. (1991). "Respiratory response of humans exposed to low levels of ozone
for 6.6 hours." American Review of Respirtory Disease 147:804-810.
31
-------
Moolgavkar, S. H.; Luebeck, E. G.; Hall, T. A.; Anderson, E. L. (1995). "Air pollution and daily
mortality in Philadelphia." Epidemiology 6: 476-484.
Morgan and Henrion (1990). Uncertainty: A Guide To Dealing with Uncertainty in Qualitative
Risk and Policy Analysis. Cambridge University Press.
Mortimer, K.M., L.M. Neas, D.W. Dockery, S. Redline, and I.B. Tager (2002). "The effects on
air pollution on inner-city children with asthma." European Respiratory Journal. 19:699-705.
Peel, J.L., P.E. Tolbert, M. Klein, K.B. Metzger, W.D. Flanders, K. Todd, J.M. Mulholland, P.B.
Ryan, and H. Frumkin, (2005). "Ambient air pollution and respirator)' emergency department
visits." Epidemiology 16(2): 164-174.
Post, E., D. Hoaglin, L. Deck, and K. Lamtz (2001). "An Empirical Bayes approach to
estimating the relation of mortality to exposure to paniculate matter," Risk Analysis 21(5): 837-
842
Richmond H., T. Palma, J. Langstaff, T. McCurdy, G. Glenn, and L. Smith (2002). "Further
refinements and testing of APEX (3.0): EPA's population exposure model for criteria and air
toxic inhalation exposures." Poster presentation. Joint meeting of the International Society of
Exposure Analysis and International Society of Environmental Epidemiology, August 11-15,
2002, Vancouver, Canada.
Schwartz, J. (2000). "The distributed lag between air pollution and daily deaths." Epidemiology
ll(3):320-326.
Schwartz, J. (2004). "How sensitive is the association between ozone and daily deaths to control
for temperature?" Am. J. Resp. Crit. Care Med.
Schwartz, J., C. Spix, G. Touloumi, L. Bacharova, T. Barumamdzadeh, A. le Tertre, T.
Piekarksi, A. Ponce de Leon, A. Ponka, G. Rossi, M. Saez, J.P. Schouten (1996).
"Methodological issues in studies of air pollution and daily counts of deaths or hospital
admissions.'V. Epid. andComm. Health 50(Suppl 1):S3-S11.
Thurston, G.D., K. Ito, P.L. Kinney, M. Lippmann (1992). "A multi-year study of air pollution
and respiratory hospital admission in three New York State metropolitan areas: Results for 1988
and 1989 summers.''/ Exposure Anal. Environ. Epidemiol. 2(4):429-450.
Thurston, G.D. and Ito, K. (2001). "Epidemiological studies of acute ozone exposures and
mortality.",/. Exposure Anal. Environ. Epidemiol. 11:286.
Tolbert, P.E., J.A. Mulholland, D.L. Macintosh, F. Xu, et al. (2000). "Air quality and Pediatric
Emergency Room Visits for Asthma in Atlanta, GA, USA." American Journal of Epidemiology
151(8):798-810.
32
-------
Whitfield, R., Biller, W., Jusko, M., and Keisler, J. (1996). A Probabilistic Assessment of
Health Risks Associated with Short- and Long-Term Exposure to Tropospheric Ozone. Argonne
National Laboratory, Argonne, IL.
Whitfield, R. (1997). A Probabilistic Assessment of Health Risks Associated with Short-term
Exposure to Tropospheric Ozone: A Supplement. Argonne National Laboratory, Argonne, IL.
33
-------
Figure 1. Overview of the APEX Model
1. Characterize study area
2. Characterize study population
3. Generate N number of
simulated individuals (profiles)
2000 Census tract-level data for the entire U.S. (sectors=tracts for the NAAOS ozone exposure application')
Sector location data I
(latitude, longitude) 1
Defined study area (sectors
within a city radius and with air
quality and meteorological data
within their radii of influence)
Locations of air quality and
meteorological measurements;
radii of influence
Sector population data
(age/gender/race)
Commuting flow data
(origin/destination sectors)
-/
- "v,
1
>
t
(^
Population within
the study area
Age/gender/tract-specific
employment probabilities
Age/gender-specific
physiological
distribution data (body
weight, height, etc)
Stochastic
profile generator
Distribution functions for
profile variables
(e.g, probability of air
conditioning)
Distribution functions
for seasonal and daily
varying profile variables
(e.g., window status, car
speed)
(
c_
(^ ) - National
database
J) - Simulation <^
step ^"""^
- Area-specific
input data
j> - Data processor
^^ \ .
- Intermediate step
or data
^___]JJ - Output data
A simulated individual with the
following profile:
• Home sector
• Work sector (if employed)
•Age
• Gender
• Race
• Employment status
• Home gas stove
• Home gas pilot
• Home air conditioner
• Car air conditioner
• Physiological parameters
(height, weight, etc.)
-------
Figure 1. Overview of the APEX Model, continued
4. Construct sequence of activity events
for each simulated individual
Diary events/activities
and personal information
(e.g., from CHAD)
Activity diary pools by day
type/temperature category
Profile for an
individual
Stochastic diary
selector using
age, gender, and
employment
Selected diary records for each day in the simulation
period, resulting in a sequence of events
(microenvironments visited, minutes spent, and
activity) in the simulation period, for an individual
Each day in the simulation
period is assigned to an
activity pool based on day type
and temperature category
Maximum/mean daily
temperature data
Stochastic
calculation of energy
expended per event (adjusted for
hysiological limits and EPOQ
and ventilation
rates
Physiological
parameters from
profile
Sequence of events for an
individual
-------
Figure 1. Overview of the APEX Model, concluded
5. Calculate concentrations in
microenvironments for all events for
each simulated individual
6. Calculate hourly
exposures for each
simulated individual
7. Calculate
population exposure
statistics
Microenvironments defined by
grouping of CHAD location codes
UJ
ON
Hourly air quality
data for all sectors
Select calculation method for
each microenvironment:
• Factors
• Mass balance
Calculate
concentrations in all
microenvironments
Concentrations for all events
for each simulated individual
Sequence of events for
each simulated individual
I
1
Average exposures
for simulated person,
stratified by ventilation
rate:
• Hourly
• Daily 1-hour max
• Daily 8-hour max
Daily...
I
Population exposure
indicators for:
• Total population
• Children
• Asthmatic children
Hourly concentrations and
minutes spent in each
microenvironment visited by
the simulated individual
Calculate hourly
concentrations in
microenvironments
visited
-------
Figure 2. Major Components of Ozone Health Risk Assessment
Based on Controlled Human Exposure Studies
Modeled
Background
Ozone Levels
Ambient Population-
Oriented Monitoring
for Selected Urban
Areas
Modeled Hour-by-Hour
Exposures Resulting From
(1) "As is" Ambient Ozone
Levels and (2) Background
Ozone Levels
Estimates of Lung
Function Risk
Associated with "As
is" Ozone Levels Over
Background
Controlled Human
Exposure Studies
(various lung
function endpoints)
Exposure -Response
Relationships
Using Exposure
Metrics Based on
Hourly Ozone
Exposures
Air Quality Adjustment
Modeled Hour-by-Hour
Exposures Resulting From
(1) "As is" Ambient Ozone
Levels and (2) When
Standards are Met
Estimates of Lung
Function Risk
Reduction Associated
with Meeting
Standards
Current or Alternative
Standards
-------
Figure 3. Major Components of Ozone Health Risk Assessment
Based on Epidemiology and Field Studies
oo
Ambient Population-
Oriented Monitoring
for Selected Urban
Areas and Estimated
Background Levels
"As is" Ozone Levels
over Background
Epidemiological and
Field Studies (various
health endpoints)
City-Specific (or
National) Baseline
Health Effects Rates
(various health
endpoints)
Air Quality Adjustment
Current or Alternative
Standards
Estimates of Health
Risk Associated with
"as is" Ozone Levels
Over Background
Concentration-
Response
Relationships
Using Ambient
Ozone
Concentrations
Ozone
Reductions To
Meet Standards
Estimates of Health
Risk Reduction
Associated with
Meeting Standards
------- |