Download - 1-s2.0-S0048969714016027-main.pdf
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 113
Health impact assessment of traf 1047297c-related air pollution at the urbanproject scale In1047298uence of variability and uncertainty
Chidsanuphong Chart-asa a Jacqueline MacDonald Gibson b
a Institute for the Study of Natural Resources and Environmental Management Mae Fah Luang University Chiang Rai Thailandb Department of Environmental Sciences and Engineering Gillings School of Global Public Health University of NC Chapel Hill USA
a b s t r a c ta r t i c l e i n f o
Article history
Received 2 December 2013Received in revised form 30 October 2014
Accepted 5 November 2014
Available online 26 November 2014
Editor Lidia Morawska
Keywords
PM25
Traf 1047297c
Health impact assessment
Variability
Uncertainty
Thispaper developsand then demonstratesa new approach for quantifyinghealth impacts of traf 1047297c-related par-
ticulate matter airpollution at the urban project scale that includesvariability and uncertainty in the analysis We
focus on primary particulate matter having a diameter less than 25 μ m (PM25) The new approach accounts for
variability in vehicle emissions due to temperature roadgrade and traf 1047297c behavior variability seasonal variabil-
ity in concentrationndashresponse coef 1047297cients demographicvariability at a 1047297ne spatial scale uncertaintyin air qual-
ity model accuracy and uncertainty in concentrationndashresponse coef 1047297cients We demonstrate the approach for a
case study roadway corridor with a population of 16000 where a new extension of the University of North
Carolina (UNC) at Chapel Hill campus is slated for construction The results indicate that at this case study site
health impact estimates increased by factors of 4ndash9 depending on the health impact considered compared to
using a conventional health impact assessment approach that overlooks these variability and uncertainty
sources In addition we demonstrate how the method can be used to assess health disparities For example in
the case study corridor our method demonstrates the existence of statistically signi1047297cant racial disparities in ex-
posure to traf 1047297c-related PM25under present-day traf 1047297c conditions thecorrelationbetween percentblackand an-
nual attributable deaths in each census block is 037 (t (114) = 42 p b 00001) Overall our results show that the
proposed newcampuswill cause only a small incremental increase in healthrisks (annual risk 6 times 10minus10 lifetime
risk 4 times 10minus8) compared to if the campus is not built Nonetheless the approach we illustrate could be useful for
improving the quality of information to support decision-making for other urban development projectscopy 2014 Elsevier BV All rights reserved
1 Introduction
In the United States nonpro1047297t organizations and public health prac-
titioners increasingly advocate for formal health impact assessments
(HIAs) to inform regional and local land-use and transportation plan-
ning decisions (Wernham 2011 Bhatia and Corburn 2011) Signaling
the heightened interest in HIAs the US National Academy of Sciences
in 2011 published a report Improving Health in the United States The
Role of Health Impact Assessment concluding that ldquoHIA is a particularly
promising approach for integrating health implications into decision-
makingrdquo (National Research Council 2011) The report offered the fol-
lowing formal de1047297nition of HIA
HIA is a systematic process that usesan arrayof data sources and an-
alytic methods and considers input from stakeholders to determine
the potential effects of a proposed policy plan program or project
on the health of a population and the distribution of those effects
within the population HIA provides recommendations on monitor-
ing and managing those effects
As the National Academies report explainsthe increasing demand for
HIAs in the United States is driven by thegrowing recognition that reduc-
ing obesity and chronic disease rates will require substantial changes to
decision-making processes in arenas outside the traditional healthcare
sector For example decisions by transportation and municipal planning
organizations can promote or limit opportunities for physical activity
and can exacerbate or decrease exposure to ambient air pollution
While HIAs of local and regional decisions have been used in Europe
Australia Canada and Thailand for decades the 1047297rst US HIA of a local
project was completed in 1999 by the San FranciscoDepartment of Public
Health (Bhatia and Corburn 2011 National Research Council 2011) By
the end of 2012 however at least 115 HIAs of local or regional US pro-
jects had been completed and another 64 were under way (Singleton-
Baldrey 2012) Of the completed HIAs 70 (more than 60) focused on
proposed local or regional changes to the built environment andor trans-
portation networks (Singleton-Baldrey 2012 Dannenberg et al 2008)
Science of the Total Environment 506ndash507 (2015) 409ndash421
Corresponding author at Department of Environmental Sciences and Engineering
Gillings School of Global Public Health University of North Carolina Campus Box 7431
Chapel Hill NC 27599-7431 USA Tel +1 919 966 7892
E-mail addresses chidsanuphonggmailcom (C Chart-asa) jackiemacdonaldunc
edu (JM Gibson)
httpdxdoiorg101016jscitotenv201411020
0048-9697copy 2014 Elsevier BV All rights reserved
Contents lists available at ScienceDirect
Science of the Total Environment
j o u r n a l h o m e p a g e w w w e l s e v i e r c o m l o c a t e s c i t o t e n v
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 213
To facilitate comparison of alternatives and guide decision-making
HIAs ideally would provide quantitative estimates of the health out-
comes of the decision options under consideration That is they would
estimate the number of deaths and illnesses prevented or caused by
each alternative This information could be used to quantify the health
costs (positive or negative) of each alternative Quanti1047297cation can pres-
ent health impacts more concisely (as numerical summaries) than
lengthy qualitative discussions In addition (rightly or wrongly) quanti-
tative assessment can lend legitimacy to the analysis Furthermoresome federal and state regulations require quantitative costndashbene1047297t
analyses (Federal Highway Administration 2003) However only 5 of
the 70 US HIAs focusing on local or regional transportation projects
carried out prior to 2013 quanti1047297ed the expected health impacts
(Singleton-Baldrey 2012 Bhatia and Seto 2011) Table 1 summarizes
these HIAs The remaining HIAs expressed qualitative conclusions
The Aerotropolis Atlanta Brown 1047297eld Redevelopment HIA (Ross et al
2011) illustrates the qualitative approach used by most previous local
and regional US HIAs This HIA evaluated a plan to convert a former
Ford assembly plant near Atlanta Georgia to a new community called
ldquoAerotropolis Atlantardquo The HIAs analysis of air quality impacts was
based on a review of previous studies (not associated with this project)
of traf 1047297c impacts on air quality and health It concluded ldquoAerotropolis
may lead to a change in traf 1047297c volume around the site hellip potentially
impacting people who live work or visit within the air-shed of the af-
fected streetsrdquo The HIA recommended several mitigation measures in-
cluding congestion pricing increased public transit zoning of sensitive
uses away from roadways and vegetation buffers around roadways
However the HIA did not quantify the air quality or health impacts of
the proposed new development or these mitigation alternatives
While the above-mentioned 1047297ve previous quantitative HIAs esti-
mate the magnitude of air quality and related health impacts none con-
siders the potential variability and uncertainty of the estimates Rather
these HIAs each provide a single deterministic prediction of health im-
pacts for each decision option (see Table 1) In so doing these HIAs not
only convey a potentially misleading degree of certainty but also ne-
glect to provide decision-makers with information about the plausible
range of impacts US Environmental Protection Agency guidance
documents indicate that health risk assessments of national and statepolicies should include sensitivity and uncertainty analyses (US
EnvironmentalProtection Agency 2001) Indeed sensitivity and uncer-
tainty analyses are cornerstones of health impact estimates the agency
prepares to inform national policy decisions such as changes to air pol-
lution standards (US Environmental Protection Agency 2010) None-
theless current US local-level HIAs do not report variability and
uncertainty in their health impact estimates
The reliance of local HIA practitioners on deterministic estimates is a
major limitation for several reasons First it fails to consider the full
range of potential risksmdashthat is the potential for risks at the tails of
the risk distribution For example vulnerable populations are often at
the upper tails not the centers of the exposure and effect distributions
(Fann et al 2011) Second risk estimates relying only on central ten-
dencies of each input variable may differ from those considering the
full distributions of each input variable Except in special cases the ex-
pected value of a function of random variables is not the same as the
function applied to the expected values of each variable Third deter-
ministic approaches ignore the potential dependencies among model
inputvariables (for example dependencies in meteorological character-
istics used to estimate pollutant dispersion) Fourth deterministic
Table 1
Previous quantitative transportation-related HIAs in the United States
Title Project scenario
analyzed
Traf 1047297c-related
air pollutants
considered
Study area
population
Estimated annual health impactsa
Pittsburg Railroad Avenue speci1047297c plan
HIA (Human Impact Partners 2008)
Construction of new Bay Area Rapid
Transit (BART) station and mixed-usevillage in Pittsburg CA including 1600
housing units and 450000 sq ft of retail
commercial and public service spaces
PM25 4770 bull 6 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)bull 5 hospital admissions for asthma (age b 65)
from short-term exposures β = 00025
(00015 00036)
bull 12 lower respiratory symptom days (ages 7ndash14)
from short-term exposures β = 00182
(00124 00241)
Evaluating the healthfulness of
affordable housing opportunity
sites along the San
Pablo Avenue Corridor using
HIA (Human Impact Partners 2009)
Construction of affordable housing sites
in El Cerrito and Richmond CA
PM25 1000000 bull 33ndash41 deaths (all ages) from long-term exposures
RR = 1014 (no report of 95 con1047297dence interval)
Oak to Ninth Avenue HIA (UC Berkeley
Health Impact Group 2006)
Development of new waterfront community
in Oakland CA including 3100 housing units
and 200000 sq ft of retail commercial
and public service spaces
PM10 10000 bull 08 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)
bull 04 chronic bronchitis cases (age ge 27) from
long-term exposures β = 00132 (00064 00200)
bull 106 emergency room visits for asthma (age b 65)
from short-term exposures β = 00037
(00024 00049)
MacArthur BART Transit Village HIA
(UC Berkeley Health Impact Group
2007)
Redevelopment of parking lot into a
mixed-use village in Oakland CA
including 625 housing units and
30000 sq ft of retail commercial
and public service spaces
PM25 100000 bull 27 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)
bull 10 chronic bronchitis cases (age ge 27) from
long-term exposures β = 00132 (00064 00200)
bull 342 acute bronchitis cases (ages 8-12) from
short-term exposures β = 00272 (00101 00443)
bull 01 hospital admissions for asthma (age b 65)
from short-term exposures β = 00025
(00015 00036)
bull 269 lower respiratory symptom days (ages 7ndash14)
from short-term exposures β = 00182 (00124
00241)
Health impact assessment of the Port of
Oakland (UC Berkeley Health Impact
Group 2010)
Ongoing growth of port operations in West
Oakland CA
PM25 22000 bull 13 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)
a
β = concentrationndashresponse coef 1047297cient used to estimate health impacts RR = relative risk used to estimate health impacts
410 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 313
estimates fail to provide information to decision-makers about the de-
gree of certainty in the estimated risks For example decision-makers
may be more concerned about a risk factor with a relatively low central
riskestimate (for example 1 in10000) ifthere isa good chancethatthe
risk could be much higher than thecentral estimate (for example a 10
chance of the risk exceeding 1 in 100) than they would be if presented
only with the central estimate of risk
Variability and uncertainty in estimated risks of traf 1047297c-related air
pollution can arise from multiple sources Variability arises naturallydue to differences in members of a population weather patterns traf 1047297c
geographic features and so on it is a property of nature ldquousually not re-
ducible through further measurement or studyrdquo (Frey and Burmaster
1999) On theother hand uncertainty arises due to thelack of informa-
tion or knowledge including limited data on a population partial igno-
rance of phenomena in1047298uencing a particular risk and disagreements
between models and the reality they are intended to represent (Frey
and Burmaster 1999) Example sources of uncertainty include the
mathematical form used to predict the effects of changes in pollution
exposure on public health the parameters in such mathematical equa-
tions and the accuracy of models predicting air pollution levels under
different traf 1047297c scenarios Theoretically uncertainty can be decreased
through further studies
This paper aims to strengthen the knowledge base and toolset avail-
able to HIA practitioners wishing to incorporate variability and uncer-
tainty in quantitative transportation-related HIAs Like the HIAs listed
in Table 1 this analysis focuses on a potential new land development
expected to increase future traf 1047297c on a major municipal road corridor
(see the ldquoCase study siterdquo section) The potential for increased traf 1047297c
has raised concerns about increases in air pollution and its associated
adverse health effects including increased risks of cardiovascular and
respiratory diseases Like four of the1047297ve HIAs in Table 1 the analysis fo-
cuses on airborne particulate matter having a diameter less than 25μ m
(denoted as PM25) as an indicator of traf 1047297c-related air pollution Like
the other HIAs this analysis is restricted to primary PM 25 (that is
PM25 emitted directly by vehicle operations rather than that formed
by chemical reactions in the atmosphere) This study considers the ef-
fects of short-term exposure to traf 1047297c-related PM25 on cardiovascular
and respiratory mortality (all ages) and unscheduled hospital admis-sions (age 65 and over) These health outcomes were previously select-
ed for the core analysis in the US Environmental Protection Agencys
(EPAs) quantitative health risk assessment for supporting the review
of the US National Ambient Air Quality Standards for PM (US
Environmental Protection Agency 2010)
We use the case study road corridor to explore the effects on health
impact estimates of PM25 from roadway traf 1047297c when including or ex-
cluding various sources of variability and uncertainty We 1047297rst use a
portion of the road corridor to explore the question ldquoWhich variability
and uncertainty sources have the greatest effects on the mean values
and upper con1047297dence limits of estimated health risksrdquo Then we dem-
onstrate a method for incorporating the key variability and uncertainty
sources in a comprehensive assessment of potential air pollution-
related health risks for the entire case study roadway corridor undercurrent conditions and future conditions with and without the pro-
posed new development
2 Case study site
We demonstrate the suggested new assessment process to explore
some of the potential health impacts arising from a planned new cam-
pus extension for the University of North Carolina (UNC) at Chapel
Hill The new campus called ldquoCarolina Northrdquo is intended to increase
the universitys capability to translate research into applications It will
be located about 3 km (2 miles) north of the existing campus ( Fig 1)
If constructed it is expected to increase the number of trips to the
area by 10000 per day by 2015 with most of the increases expected
to occur along MartinLuther King Jr Blvd themain link to the existing
campus and the major northndashsouth road corridor in Chapel Hill (Vanasse
Hangen Brustlin Inc 2009) By 2025 the number of additional daily trips
to the campus is expected to increase by as many as 40000 We consider
the potential impacts of the expected additional traf 1047297c-related air pollu-
tion among residents living in census blocks within 500 m of Martin Lu-
ther King Jr Blvd In all this area encompasses 160 US census blocks
(see Fig 1) and has a total population of about 16000mdashmore than one-
quarter of Chapel Hills total population of 57000
We analyze the effects of primary emissions from traf 1047297
c along Mar-tin Luther King Jr Blvd on ambient PM25 concentrations and popula-
tion health under three different scenarios (1) the year 2009
(2) 2025 assuming the new campus is not built and (3) 2025 with
the new campus The baseline comparison year is 2009 because the
most comprehensive transportation analysis of the study corridor was
conducted using 2009 data (Vanasse Hangen Brustlin Inc 2009)
Table 2 provides summary information about the population size and
traf 1047297c volumes under these three scenarios
3 Methods and data sources
This analysis has two main parts
1) Analyze the effects of including variability and uncertainty in the HIA
First we investigate in the effects on health impact estimates of in-cludingseveral different uncertainty and variability sources as com-
pared to results obtained using the conventional deterministic
approach For computational ef 1047297ciency we focus on the 12 census
blocks highlighted in Fig 1B which our prior air quality modeling in-
dicated are more vulnerable to traf 1047297c-related PM25 than most other
census blocks in the corridor (Chart-asa et al 2013) The total pop-
ulation in the 12 blocks is 1117 (about 7 of the total population in
the study corridor)
2) Quantify the health impacts of traf 1047297c from the proposed new campus in
the study corridor Second we quantify traf 1047297c-related air quality and
health outcomes along the entire study corridor for the three devel-
opment scenarios in Table 2 This analysis includes the variability
and uncertainty sources identi1047297ed in part 1 as having an in1047298uence
on the central estimates or upper con1047297
dence estimates of themodeled risks
Both analyses use the same modeling framework describedin detail
in the followingsections However the1047297rst analysis introducesvariabil-
ity and uncertainty sources one at a time in order to explore their po-
tential in1047298uence on the computed health risks while the second
analysis includes all key variability and uncertainty sources
31 Modeling framework overview
Quantifying the health impacts of traf 1047297c-related air pollution re-
quires three categories of information (1) estimates of the excess
PM25 concentrations to which the population is exposed as a result of
primary emissions from traf 1047297c (2) concentrationndashresponse functions
relating exposure concentrations to probabilities of adverse health out-comes and (3) incidence rates of the health outcomes of concern (from
all causes) in the exposed population (Ostro 2004 Ostro and Chestnut
1998 Cohen et al 2005 Li et al 2010 ) Fig 2 summarizes how this
analysis combines these three information categories (shows as shaded
boxes)to estimate health impacts Theunshaded boxes show variability
and uncertainty sources considered in this study The subscript notation
indicatesthat theanalysisis conductedat thecensus block scalewhere i
represents an individual block That is health risks are characterized
separately for each census block considering variability in traf 1047297c-
related PM25 exposure concentrations and population demographic
characteristics within each block The subscripts j k and l indicate dif-
ferences in baseline health status by age ( j) gender (k) and race (l)
In addition this analysis considers seasonal (subscript m) variability
because epidemiologic evidence suggests seasonal differences in
411C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 413
dosendashresponse functions (Zanobetti and Schwartz 2009 Bell et al
2008) The following sections provide details on our methods for esti-
mating PM25 exposure concentrations (left-most shaded diamond in
Fig 2) selecting concentrationndashresponse functions (central diamond)
estimating baseline incidence rates of adverse health outcomes in the
study population (right-most diamond) and incorporating variability
and uncertainty (white rectangles) into the analysis
32 PM 25 concentrations attributable to primary emissions from traf 1047297c
The 24-hour exposures to PM25 arising from primary emissions
from traf 1047297c along the case study roadway corridor were estimated
using an integrated air quality modeling approach described in Chart-
asa et al (2013) In brief the approach employs standard traf 1047297c emis-
sions and air quality dispersion modeling tools but it adds a novel ap-
proach for modeling variability in vehicle emissions due to variability
in hourly temperature roadgrade and traf 1047297c behavior (including cruis-
ing speed and percent time spent idling decelerating and accelerating)
The exposure modeling approach links a novel application of MOVES2010b commonly used in the United States to estimate vehicle emis-
sions factors (gvehicle-mile) and CAL3QHCR which characterizes
PM25 dispersion away from roadways By linking these models and
employing a new approach for characterizing variabilityin emission fac-
tors we simulated probability distributions of the average 24-hour
Fig 1 (A)The study corridorbetween the intersection of Martin LutherKing JrBlvdand Whit1047297eld Rdand theintersection of South ColumbiaSt andMt Carmel Church Rd Chapel Hill
NC and the census blocks located within 500 m from the study corridor (B) The road segment and census blocks for simulations to demonstrate differences in health burden estimates
when including variability and the uncertainty in the modeling approach Dots represent census block centroids
Table 2
Population size and traf 1047297c volumes under three scenarios considered
Scenario Traf 1047297c volumes of road segments on study corridor (vehh) a Total population of 160 census blocks located within 500 m from study corridor
2009 4ndash1758 16042
2025 without the new campus 5ndash2443 19140b
2025 with the new campus 5ndash2832 19140b
a Ranges indicate variability in traf 1047297c 1047298ow by road segment day of week and time of dayb
Computed from growth rates forecasted by the North Carolina Capital Area Metropolitan Planning Organization (2005)
412 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 513
PM25 concentration in each season (winter spring summer and fall) at
the centroid of each of the 160 census blocks in the study corridorOur analysis considers variability in vehicle emission factors by di-
viding the 82-km roadway corridor into 1200 links and estimating sep-
arate emission factors for each link for each hour of each simulation day
(Chart-asa et al 2013) Unlike previous studies linking MOVES and
CAL3QHCR our analysis considers hourly variability in temperature
and link-speci1047297c variability in road grade and vehicle behavior Hourly
meteorological pro1047297les for 2006ndash2012 were obtained from the national
weather stations in ChapelHill andGreensboroNorthCarolinaand me-
teorological pro1047297les for input to CAL3QHCR were generated from EPAs
Meteorological Processor for Regulatory Models (NCDC 2013 NOAA
2013) The meteorological pro1047297les contained a total of 2100 days with
complete required data (525 days for winter 560 days for spring
532 days for summer and 483 days for fall) For each census block we
used CAL3QHCR to estimate the PM25 concentration (averaged over24 h) attributable to primary traf 1047297c emissions from each of the 1200
roadway links for each of the 2100 days for which meteorological data
were available Separate estimates were prepared for 2009 and 2025
using emission factors from MOVES modeling and simulated traf 1047297c
data for 2009 and 2025 scenarios with or without Carolina North from
the Transportation Impact Analysis (TIA) for the Carolina North Devel-
opment (Vanasse Hangen Brustlin Inc 2009) Then for each develop-
ment scenario seven separate mean estimatesmdashone for each of the
seven years for which meteorological data were availablemdashof the sea-
sonal mean value of the24-hour average PM25exposure concentrations
were computed for each season
For each scenario (year 2009 and year 2025 with and without con-
structingCarolina North) andeach season we then computed bootstrap
estimatesof the mean value andstandarddeviation of theseasonal daily
average PM25 exposure concentrations by randomly selecting one of
the seven years assigning the associated seasonal mean concentrationsas computed using that years data to each census block and then re-
peating the process 1999 times For each of the 160 census blocks the
result was an estimated mean value and standard deviation of the sea-
sonal 24-hour-average concentration of PM25 attributable to primary
emissions from traf 1047297c along the roadway corridor under each scenario
Within each census block and for each scenario the seasonal average
traf 1047297c-related 24-hour PM25 concentration then was represented as a
normaldistribution(left-truncatedat zero) with themean andstandard
deviation estimated from the corresponding 2000 bootstrap simula-
tions The TIA estimated hourly traf 1047297c counts for each scenario along
each roadway link only for weekdays we assumed traf 1047297c counts on
weekends would be the same and hence may have slightly over-
estimated exposure concentrations
In addition to considering variability in PM25exposures arising fromprimary traf 1047297c emissions we assessed the effects of uncertainty in the
accuracy of the air quality model predictions Our previous research
on the integrated air quality modeling approach as well as previous
work by others suggests that the combined MOVESndashCAL3QHCR
model generally predicts PM25 concentration within a factor of two of
measured concentrations (although accuracy varies with local condi-
tions and the quality of data available to support the model) ( Chart-
asa etal2013 Yura etal2007) FollowingMorganand Henrionsguid-
ance (Morgan et al 1990) we represented model uncertainty with an
uncertainty factor (UF ) parameterized by a triangular probability distri-
bution with lower limit = 05 upper limit = 20 and mode = 10
(spanning the expected factor-of-two uncertainty in the model) Ac-
cording to Morgan and Henrion the triangular distribution is especially
appropriate for situations in which ldquothe distributions of variables in a
Fig 2 Overview of framework for incorporating variability and uncertainty into assessment of the health impacts of traf 1047297c-related PM25 The rectangles show sources of variability and
uncertainty The shaded diamonds show the three major information categories needed for quantitative health impact assessment
413C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 613
model arenot preciselyknownrdquo but in which ldquovaluestowardthe middle
of the range of possible values are considered more likely to occur than
values near either extremerdquo Based on previous evaluations of the per-
formance of near-roadway air pollutant dispersion models the exact
form of the distribution representing model uncertainty is not known
making the triangular distribution an appropriate choice for character-
izing model uncertainty Correspondingly in each census block the ex-
cess PM25 24-hour average exposure concentration attributable to
primary emissions from traf 1047297
c on the case study roadway was estimat-ed for each season as
PM exposureimfrac14 U F PM modelim
eth1THORN
where PM exposureimrepresents the average 24-hour PM25 concentration
in census block i (i = 1ndash160) and season m (m = winter spring sum-
mer fall) attributable to primary traf 1047297c emissions on the case study
roadway UF is the model uncertainty factor and PM modelimis thecorre-
sponding model-predicted seasonal daily average PM25 concentration
arising from primary emissions from traf 1047297c
321 Concentrationndashresponse functions
As recommended by the World Health Organization and others
(Ostro and Chestnut 1998 Li et al 2010 Aunan 1996) we use
the following relationship to describe the link between seasonal
daily average PM25 concentrations and the relative risk of cardio-
vascular and respiratory health outcomes
RRimn frac14 e β mn P M exposureim eth2THORN
where β mn is the concentrationndashresponse coef 1047297cient describing the
effects of PM on health outcome n during season m and RRimn is the
relative risk of health outcome n during season m in census block i
The number of adverse health cases in the population attributable
to traf 1047297c-related PM25 then can be determined from the following
relationship
Δ yi jklmn frac14 y0i jklmn A F i jklmn 3a
frac14 y0i jklmn
RRimnminus1
RRimn
3b
frac14 y0i jklmn
e β mn P M exposureimminus1
e β mn P M exposureim
3c
frac14 y0i jklmn 1minuse
minus β mn P M exposureim
3d
where AF i jklmn and Δ yi jklmn are the fraction and number of casesof adverse health event n attributable to traf 1047297c-related PM25 in sea-
son m in census block i for age group j gender k and race l and
where yi jklmn0 is the observed total number of cases in the same lo-
cation and among the same population group Eqs (2) (3a) (3b)
(3c) and (3d) are the standard equations used in analyses by the
WHO and other organizations to attribute observed cases of ad-
verse health events to speci1047297c risk factors ( Ostro and Chestnut
1998 Murray et al 2003 Mathers et al 2001 Pruumlss-uumlstuumln et al
2003)
The β values in Eqs (2) (3c) and (3d) (known as dosendashresponse co-
ef 1047297cients) were drawn from the US Environmental Protection Agency
guidance document Quantitative Health Risk Assessment for Particulate
Matter (US Environmental Protection Agency 2010 Zanobetti and
Schwartz 2009 Bell et al 2008) Table 3 shows the coef 1047297cient values
used in this analysis EPA retrieved these coef 1047297cients from peer-
reviewed epidemiologic studies that met certain quality-assurance
criteria including for example the estimation of exposure from mea-
sured rather than modeled PM25 data For mortality effects the coef 1047297-
cients are speci1047297c to 15 US metropolitan areas For morbidity effects
coef 1047297cients are speci1047297c to region (Northeast Southeast Northwest
and Southwest) This study employed mortality coef 1047297cients developed
from studies in Atlanta since Atlanta is climatologically the most similar
to Chapel Hill among the 15 cities studied We used morbiditycoef 1047297cients for the Southeast region in which Chapel Hill is located
All concentrationndashresponse coef 1047297cients were represented as normal
distributions with all negative valuestruncated at zero (to avoid associ-
ating PM exposure with positive health effects) Standard deviationsfor
each season and health outcome were estimated from the con1047297dence
intervals in Table 3
33 Baseline incidence rates of adverse health outcomes
Data on baseline incidence rates of health outcomes were obtained
from North Carolina public health databases Annual mortality rates
for each age group (Table 4) were calculated by dividing thetotal num-
ber of deaths in Orange County (where Chapel Hill is located) in 2010
(North Carolina State Center for Health Statistics 2012) bythe 2010Or-
ange County census population (Minnesota Population Center 2011)
Annual unscheduled hospital admission rates (Table 5) were obtained
from 2009 emergency department visit data reported by the North Car-
olina Disease Event Tracking and Epidemiologic Collection Tool (NC DE-
TECT) (University of North Carolina at Chapel Hill 2011) We were
unable to obtain data on incidence rates by gender and race so we as-
sume that incidence rates are the same for both genders and all races
(which is a limitation of this analysis) It should be noted as well that
the ICD codes speci1047297c to the concentrationndashresponse coef 1047297cients
might not be entirely matched to the ICD codes speci1047297c to the incidence
rates used in this study depending on reported data Moreover emer-
gency department visits may not result in hospital admissions and
some hospital admissions may occur without 1047297rst visiting the emergen-
cy department
Tore1047298ect seasonal variation we adjusted the incidence rates for car-diovascular and respiratory mortality and unscheduled hospital admis-
sions using data on temporal variability in cardiovascular and
respiratory deaths in Orange County during 1999ndash2010 from the CDC
WONDER database (Centers for Disease Control and Prevention
2013) The fractions for cardiovascular events are 025 031 020 and
024 for winter spring summer and fall respectively while the frac-
tions for respiratory events are 030 026 021 and 023 for winter
spring summer and fall respectively
To determine the total number of cases in any given season (ie
yi jklmn0 in Eqs (3a) (3b) (3c) and (3d)) we multiplied the given inci-
dence rateby the corresponding size of each demographic group in each
census block
34 Testing the effects of variability and uncertainty on health impact estimates
Five simulations of 2000 iterations each were run using Analytica
version 45 (Lumina Decision Systems Los Gatos California) to demon-
strate differences in health burden estimates when including variability
and uncertainty Table 6 lists the1047297ve simulations and the variability and
uncertainty considered in each The1047297rst simulation (1a) follows the de-
terministic approach of previous HIAs usingaverage traf 1047297c volumes and
a constant traf 1047297c emission factor corresponding to traf 1047297c cruising at
35 mph on a 1047298at roadway under a constant ambient temperature of
70 degF Like previous HIAs simulation 1a accounts for neither uncertainty
in the concentrationndashresponse coef 1047297cient (usingthe meanvalue asa de-
terministic estimate) nor seasonal variability Simulation 1b is identical
to simulation 1a except that it uses seasonal concentrationndashresponse
414 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 713
coef 1047297cients (also deterministic) Simulations 2ndash4 systematically include
(one at a time) variability in vehicle emissions rates (simulations 2ndash4)
uncertainty in concentrationndashresponse coef 1047297cients (simulations 3ndash4)
and air quality model prediction error (simulation 4)
35 Comparing health impacts under alternative scenarios
As noted previously we simulated health impacts for the full study
corridor (160 census blocks) for three different scenarios (Table 2)
year 2009 and year 2025 with and without the new campus For each
scenario 2000 simulations were run in Analytica Traf 1047297c patterns (traf-
1047297c volumes along each roadway link in the corridor) for each scenario
were taken from a previous traf 1047297c impact analysis conducted for theTownof Chapel Hill(Vanasse Hangen Brustlin Inc2009)The 2025cen-
sus block populations were obtained from forecasts by the North
Carolina Capital Area Metropolitan Planning Organization (2005)
These growth rates account for demographic changes expected to
occur if the Carolina North campus is built
4 Results and discussion
This analysis explored the effects of variability and uncertainty on
health impact estimates of near-roadway air pollution arising from traf-1047297c attracted by newsuburban development projects Most previous US
HIAs of such projects have provided qualitative rather than quantitative
assessments of health impacts the few quantitative HIAs have not
systematically represented variability and uncertainty in the variables
used to estimate health impacts or in the resulting health outcome pre-
dictions We explored whether including variability and uncertainty
makes a difference in centralestimates of health impacts and we exam-
ined the magnitude of uncertainty in the resulting estimates We then
employed an approach that accounts for variability and uncertainty to
model the expected health impacts in the year 2025 of new traf 1047297c gen-
erated by a new research campus development along a busy roadway
corridor in Chapel Hill North Carolina
41 Effect of including variability and uncertainty
Our results suggest that the conventional deterministic HIA ap-
proach may systematically under-estimate potential health impacts of
traf 1047297c-related PM25 exposure (Fig 3)
Incorporating traf 1047297c emission variability into the analysis (as in sim-
ulation 2) caused the mean value of estimated health impacts to in-
crease by more than a factor of two compared to estimates that
exclude such variability (simulation 1b) This increase occurred because
neglecting the effects on vehicle emissions of variability in temperature
road grade vehicle speed and traf 1047297c behavior (idling accelerating de-
celerating or cruising) resulted in under-estimates of PM25 exposure
Table 3
Concentrationndashresponse coef 1047297cients used in this study
Health outcome Disease category ICD-9 or ICD-10 codea Age group Season Mean concentration-response
coef 1047297cient (95 CI) per 10 μ gm3b
Mortality Cardiovascular I01ndashI59 All ages All-yearc 066 (minus066 198)
Winter 135 (minus193 462)
Spring 076 (minus273 425)
Summer 062 (minus222 347)
Fall minus018 (minus293 257)
Respiratory J00ndash J99 All ages All-yearc
121 (minus048 290)Winter 093 (minus144 329)
Spring 035 (minus205 275)
Summer 077 (minus155 310)
Fall 096 (minus134 325)
Unscheduled hospital admissions Cardiovascular 410ndash414 426ndash429 430ndash438 and 440ndash4 49 6 5 a nd over A ll-yea rc 029 (minus019 077)
Winter 105 (minus007 219)
Spring 075 (minus026 176)
Summer minus067 (minus161 026)
Fall 017 (minus072 106)
Respiratory 464ndash466 480ndash487 and 490ndash492 65 and over All-yearc 035 (minus044 113)
Winter 040 (minus146 224)
Spring 075 (minus082 231)
Summer minus052 (minus209 105)
Fall 014 (minus130 158)
a ICD-10 for mortality and ICD-9 for unscheduled hospital admissionsb Coef 1047297cients were originally from Zanobetti and Schwartz (2009) and Bell et al (2008) respectively
c Used only in simulation 1
Table 4Annual mortality rates by race gender and age group for Orange County (per 1000 people)
Cause of death ICD-10 code Age group Race and gender
White male Black male Other male White female Black female Other female
Cardiovascular disease I05ndashI09 I10ndashI15 I20ndashI25 I26ndashI28 and I30ndashI52 0 to 34 000 000 000 000 000 000
35 to 44 017 000 000 000 095 000
45 to 54 029 272 000 013 153 000
55 to 64 179 235 257 091 105 000
65 to 74 343 725 000 244 000 000
75 to 84 1647 1724 000 770 2065 000
85+ 5251 2632 12500 3034 2299 000
Respiratory disease J00ndash J99 0 to 54 000 000 000 000 000 000
55 to 64 049 000 000 045 105 000
65 to 74 206 483 000 183 345 000
75 to 84 524 575 116 495 590 000
85+ 3580 1316 000 787 1149 000
415C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 213
To facilitate comparison of alternatives and guide decision-making
HIAs ideally would provide quantitative estimates of the health out-
comes of the decision options under consideration That is they would
estimate the number of deaths and illnesses prevented or caused by
each alternative This information could be used to quantify the health
costs (positive or negative) of each alternative Quanti1047297cation can pres-
ent health impacts more concisely (as numerical summaries) than
lengthy qualitative discussions In addition (rightly or wrongly) quanti-
tative assessment can lend legitimacy to the analysis Furthermoresome federal and state regulations require quantitative costndashbene1047297t
analyses (Federal Highway Administration 2003) However only 5 of
the 70 US HIAs focusing on local or regional transportation projects
carried out prior to 2013 quanti1047297ed the expected health impacts
(Singleton-Baldrey 2012 Bhatia and Seto 2011) Table 1 summarizes
these HIAs The remaining HIAs expressed qualitative conclusions
The Aerotropolis Atlanta Brown 1047297eld Redevelopment HIA (Ross et al
2011) illustrates the qualitative approach used by most previous local
and regional US HIAs This HIA evaluated a plan to convert a former
Ford assembly plant near Atlanta Georgia to a new community called
ldquoAerotropolis Atlantardquo The HIAs analysis of air quality impacts was
based on a review of previous studies (not associated with this project)
of traf 1047297c impacts on air quality and health It concluded ldquoAerotropolis
may lead to a change in traf 1047297c volume around the site hellip potentially
impacting people who live work or visit within the air-shed of the af-
fected streetsrdquo The HIA recommended several mitigation measures in-
cluding congestion pricing increased public transit zoning of sensitive
uses away from roadways and vegetation buffers around roadways
However the HIA did not quantify the air quality or health impacts of
the proposed new development or these mitigation alternatives
While the above-mentioned 1047297ve previous quantitative HIAs esti-
mate the magnitude of air quality and related health impacts none con-
siders the potential variability and uncertainty of the estimates Rather
these HIAs each provide a single deterministic prediction of health im-
pacts for each decision option (see Table 1) In so doing these HIAs not
only convey a potentially misleading degree of certainty but also ne-
glect to provide decision-makers with information about the plausible
range of impacts US Environmental Protection Agency guidance
documents indicate that health risk assessments of national and statepolicies should include sensitivity and uncertainty analyses (US
EnvironmentalProtection Agency 2001) Indeed sensitivity and uncer-
tainty analyses are cornerstones of health impact estimates the agency
prepares to inform national policy decisions such as changes to air pol-
lution standards (US Environmental Protection Agency 2010) None-
theless current US local-level HIAs do not report variability and
uncertainty in their health impact estimates
The reliance of local HIA practitioners on deterministic estimates is a
major limitation for several reasons First it fails to consider the full
range of potential risksmdashthat is the potential for risks at the tails of
the risk distribution For example vulnerable populations are often at
the upper tails not the centers of the exposure and effect distributions
(Fann et al 2011) Second risk estimates relying only on central ten-
dencies of each input variable may differ from those considering the
full distributions of each input variable Except in special cases the ex-
pected value of a function of random variables is not the same as the
function applied to the expected values of each variable Third deter-
ministic approaches ignore the potential dependencies among model
inputvariables (for example dependencies in meteorological character-
istics used to estimate pollutant dispersion) Fourth deterministic
Table 1
Previous quantitative transportation-related HIAs in the United States
Title Project scenario
analyzed
Traf 1047297c-related
air pollutants
considered
Study area
population
Estimated annual health impactsa
Pittsburg Railroad Avenue speci1047297c plan
HIA (Human Impact Partners 2008)
Construction of new Bay Area Rapid
Transit (BART) station and mixed-usevillage in Pittsburg CA including 1600
housing units and 450000 sq ft of retail
commercial and public service spaces
PM25 4770 bull 6 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)bull 5 hospital admissions for asthma (age b 65)
from short-term exposures β = 00025
(00015 00036)
bull 12 lower respiratory symptom days (ages 7ndash14)
from short-term exposures β = 00182
(00124 00241)
Evaluating the healthfulness of
affordable housing opportunity
sites along the San
Pablo Avenue Corridor using
HIA (Human Impact Partners 2009)
Construction of affordable housing sites
in El Cerrito and Richmond CA
PM25 1000000 bull 33ndash41 deaths (all ages) from long-term exposures
RR = 1014 (no report of 95 con1047297dence interval)
Oak to Ninth Avenue HIA (UC Berkeley
Health Impact Group 2006)
Development of new waterfront community
in Oakland CA including 3100 housing units
and 200000 sq ft of retail commercial
and public service spaces
PM10 10000 bull 08 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)
bull 04 chronic bronchitis cases (age ge 27) from
long-term exposures β = 00132 (00064 00200)
bull 106 emergency room visits for asthma (age b 65)
from short-term exposures β = 00037
(00024 00049)
MacArthur BART Transit Village HIA
(UC Berkeley Health Impact Group
2007)
Redevelopment of parking lot into a
mixed-use village in Oakland CA
including 625 housing units and
30000 sq ft of retail commercial
and public service spaces
PM25 100000 bull 27 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)
bull 10 chronic bronchitis cases (age ge 27) from
long-term exposures β = 00132 (00064 00200)
bull 342 acute bronchitis cases (ages 8-12) from
short-term exposures β = 00272 (00101 00443)
bull 01 hospital admissions for asthma (age b 65)
from short-term exposures β = 00025
(00015 00036)
bull 269 lower respiratory symptom days (ages 7ndash14)
from short-term exposures β = 00182 (00124
00241)
Health impact assessment of the Port of
Oakland (UC Berkeley Health Impact
Group 2010)
Ongoing growth of port operations in West
Oakland CA
PM25 22000 bull 13 deaths (age ge 30) from long-term exposures
β = 00046 (00034 00058)
a
β = concentrationndashresponse coef 1047297cient used to estimate health impacts RR = relative risk used to estimate health impacts
410 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 313
estimates fail to provide information to decision-makers about the de-
gree of certainty in the estimated risks For example decision-makers
may be more concerned about a risk factor with a relatively low central
riskestimate (for example 1 in10000) ifthere isa good chancethatthe
risk could be much higher than thecentral estimate (for example a 10
chance of the risk exceeding 1 in 100) than they would be if presented
only with the central estimate of risk
Variability and uncertainty in estimated risks of traf 1047297c-related air
pollution can arise from multiple sources Variability arises naturallydue to differences in members of a population weather patterns traf 1047297c
geographic features and so on it is a property of nature ldquousually not re-
ducible through further measurement or studyrdquo (Frey and Burmaster
1999) On theother hand uncertainty arises due to thelack of informa-
tion or knowledge including limited data on a population partial igno-
rance of phenomena in1047298uencing a particular risk and disagreements
between models and the reality they are intended to represent (Frey
and Burmaster 1999) Example sources of uncertainty include the
mathematical form used to predict the effects of changes in pollution
exposure on public health the parameters in such mathematical equa-
tions and the accuracy of models predicting air pollution levels under
different traf 1047297c scenarios Theoretically uncertainty can be decreased
through further studies
This paper aims to strengthen the knowledge base and toolset avail-
able to HIA practitioners wishing to incorporate variability and uncer-
tainty in quantitative transportation-related HIAs Like the HIAs listed
in Table 1 this analysis focuses on a potential new land development
expected to increase future traf 1047297c on a major municipal road corridor
(see the ldquoCase study siterdquo section) The potential for increased traf 1047297c
has raised concerns about increases in air pollution and its associated
adverse health effects including increased risks of cardiovascular and
respiratory diseases Like four of the1047297ve HIAs in Table 1 the analysis fo-
cuses on airborne particulate matter having a diameter less than 25μ m
(denoted as PM25) as an indicator of traf 1047297c-related air pollution Like
the other HIAs this analysis is restricted to primary PM 25 (that is
PM25 emitted directly by vehicle operations rather than that formed
by chemical reactions in the atmosphere) This study considers the ef-
fects of short-term exposure to traf 1047297c-related PM25 on cardiovascular
and respiratory mortality (all ages) and unscheduled hospital admis-sions (age 65 and over) These health outcomes were previously select-
ed for the core analysis in the US Environmental Protection Agencys
(EPAs) quantitative health risk assessment for supporting the review
of the US National Ambient Air Quality Standards for PM (US
Environmental Protection Agency 2010)
We use the case study road corridor to explore the effects on health
impact estimates of PM25 from roadway traf 1047297c when including or ex-
cluding various sources of variability and uncertainty We 1047297rst use a
portion of the road corridor to explore the question ldquoWhich variability
and uncertainty sources have the greatest effects on the mean values
and upper con1047297dence limits of estimated health risksrdquo Then we dem-
onstrate a method for incorporating the key variability and uncertainty
sources in a comprehensive assessment of potential air pollution-
related health risks for the entire case study roadway corridor undercurrent conditions and future conditions with and without the pro-
posed new development
2 Case study site
We demonstrate the suggested new assessment process to explore
some of the potential health impacts arising from a planned new cam-
pus extension for the University of North Carolina (UNC) at Chapel
Hill The new campus called ldquoCarolina Northrdquo is intended to increase
the universitys capability to translate research into applications It will
be located about 3 km (2 miles) north of the existing campus ( Fig 1)
If constructed it is expected to increase the number of trips to the
area by 10000 per day by 2015 with most of the increases expected
to occur along MartinLuther King Jr Blvd themain link to the existing
campus and the major northndashsouth road corridor in Chapel Hill (Vanasse
Hangen Brustlin Inc 2009) By 2025 the number of additional daily trips
to the campus is expected to increase by as many as 40000 We consider
the potential impacts of the expected additional traf 1047297c-related air pollu-
tion among residents living in census blocks within 500 m of Martin Lu-
ther King Jr Blvd In all this area encompasses 160 US census blocks
(see Fig 1) and has a total population of about 16000mdashmore than one-
quarter of Chapel Hills total population of 57000
We analyze the effects of primary emissions from traf 1047297
c along Mar-tin Luther King Jr Blvd on ambient PM25 concentrations and popula-
tion health under three different scenarios (1) the year 2009
(2) 2025 assuming the new campus is not built and (3) 2025 with
the new campus The baseline comparison year is 2009 because the
most comprehensive transportation analysis of the study corridor was
conducted using 2009 data (Vanasse Hangen Brustlin Inc 2009)
Table 2 provides summary information about the population size and
traf 1047297c volumes under these three scenarios
3 Methods and data sources
This analysis has two main parts
1) Analyze the effects of including variability and uncertainty in the HIA
First we investigate in the effects on health impact estimates of in-cludingseveral different uncertainty and variability sources as com-
pared to results obtained using the conventional deterministic
approach For computational ef 1047297ciency we focus on the 12 census
blocks highlighted in Fig 1B which our prior air quality modeling in-
dicated are more vulnerable to traf 1047297c-related PM25 than most other
census blocks in the corridor (Chart-asa et al 2013) The total pop-
ulation in the 12 blocks is 1117 (about 7 of the total population in
the study corridor)
2) Quantify the health impacts of traf 1047297c from the proposed new campus in
the study corridor Second we quantify traf 1047297c-related air quality and
health outcomes along the entire study corridor for the three devel-
opment scenarios in Table 2 This analysis includes the variability
and uncertainty sources identi1047297ed in part 1 as having an in1047298uence
on the central estimates or upper con1047297
dence estimates of themodeled risks
Both analyses use the same modeling framework describedin detail
in the followingsections However the1047297rst analysis introducesvariabil-
ity and uncertainty sources one at a time in order to explore their po-
tential in1047298uence on the computed health risks while the second
analysis includes all key variability and uncertainty sources
31 Modeling framework overview
Quantifying the health impacts of traf 1047297c-related air pollution re-
quires three categories of information (1) estimates of the excess
PM25 concentrations to which the population is exposed as a result of
primary emissions from traf 1047297c (2) concentrationndashresponse functions
relating exposure concentrations to probabilities of adverse health out-comes and (3) incidence rates of the health outcomes of concern (from
all causes) in the exposed population (Ostro 2004 Ostro and Chestnut
1998 Cohen et al 2005 Li et al 2010 ) Fig 2 summarizes how this
analysis combines these three information categories (shows as shaded
boxes)to estimate health impacts Theunshaded boxes show variability
and uncertainty sources considered in this study The subscript notation
indicatesthat theanalysisis conductedat thecensus block scalewhere i
represents an individual block That is health risks are characterized
separately for each census block considering variability in traf 1047297c-
related PM25 exposure concentrations and population demographic
characteristics within each block The subscripts j k and l indicate dif-
ferences in baseline health status by age ( j) gender (k) and race (l)
In addition this analysis considers seasonal (subscript m) variability
because epidemiologic evidence suggests seasonal differences in
411C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 413
dosendashresponse functions (Zanobetti and Schwartz 2009 Bell et al
2008) The following sections provide details on our methods for esti-
mating PM25 exposure concentrations (left-most shaded diamond in
Fig 2) selecting concentrationndashresponse functions (central diamond)
estimating baseline incidence rates of adverse health outcomes in the
study population (right-most diamond) and incorporating variability
and uncertainty (white rectangles) into the analysis
32 PM 25 concentrations attributable to primary emissions from traf 1047297c
The 24-hour exposures to PM25 arising from primary emissions
from traf 1047297c along the case study roadway corridor were estimated
using an integrated air quality modeling approach described in Chart-
asa et al (2013) In brief the approach employs standard traf 1047297c emis-
sions and air quality dispersion modeling tools but it adds a novel ap-
proach for modeling variability in vehicle emissions due to variability
in hourly temperature roadgrade and traf 1047297c behavior (including cruis-
ing speed and percent time spent idling decelerating and accelerating)
The exposure modeling approach links a novel application of MOVES2010b commonly used in the United States to estimate vehicle emis-
sions factors (gvehicle-mile) and CAL3QHCR which characterizes
PM25 dispersion away from roadways By linking these models and
employing a new approach for characterizing variabilityin emission fac-
tors we simulated probability distributions of the average 24-hour
Fig 1 (A)The study corridorbetween the intersection of Martin LutherKing JrBlvdand Whit1047297eld Rdand theintersection of South ColumbiaSt andMt Carmel Church Rd Chapel Hill
NC and the census blocks located within 500 m from the study corridor (B) The road segment and census blocks for simulations to demonstrate differences in health burden estimates
when including variability and the uncertainty in the modeling approach Dots represent census block centroids
Table 2
Population size and traf 1047297c volumes under three scenarios considered
Scenario Traf 1047297c volumes of road segments on study corridor (vehh) a Total population of 160 census blocks located within 500 m from study corridor
2009 4ndash1758 16042
2025 without the new campus 5ndash2443 19140b
2025 with the new campus 5ndash2832 19140b
a Ranges indicate variability in traf 1047297c 1047298ow by road segment day of week and time of dayb
Computed from growth rates forecasted by the North Carolina Capital Area Metropolitan Planning Organization (2005)
412 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 513
PM25 concentration in each season (winter spring summer and fall) at
the centroid of each of the 160 census blocks in the study corridorOur analysis considers variability in vehicle emission factors by di-
viding the 82-km roadway corridor into 1200 links and estimating sep-
arate emission factors for each link for each hour of each simulation day
(Chart-asa et al 2013) Unlike previous studies linking MOVES and
CAL3QHCR our analysis considers hourly variability in temperature
and link-speci1047297c variability in road grade and vehicle behavior Hourly
meteorological pro1047297les for 2006ndash2012 were obtained from the national
weather stations in ChapelHill andGreensboroNorthCarolinaand me-
teorological pro1047297les for input to CAL3QHCR were generated from EPAs
Meteorological Processor for Regulatory Models (NCDC 2013 NOAA
2013) The meteorological pro1047297les contained a total of 2100 days with
complete required data (525 days for winter 560 days for spring
532 days for summer and 483 days for fall) For each census block we
used CAL3QHCR to estimate the PM25 concentration (averaged over24 h) attributable to primary traf 1047297c emissions from each of the 1200
roadway links for each of the 2100 days for which meteorological data
were available Separate estimates were prepared for 2009 and 2025
using emission factors from MOVES modeling and simulated traf 1047297c
data for 2009 and 2025 scenarios with or without Carolina North from
the Transportation Impact Analysis (TIA) for the Carolina North Devel-
opment (Vanasse Hangen Brustlin Inc 2009) Then for each develop-
ment scenario seven separate mean estimatesmdashone for each of the
seven years for which meteorological data were availablemdashof the sea-
sonal mean value of the24-hour average PM25exposure concentrations
were computed for each season
For each scenario (year 2009 and year 2025 with and without con-
structingCarolina North) andeach season we then computed bootstrap
estimatesof the mean value andstandarddeviation of theseasonal daily
average PM25 exposure concentrations by randomly selecting one of
the seven years assigning the associated seasonal mean concentrationsas computed using that years data to each census block and then re-
peating the process 1999 times For each of the 160 census blocks the
result was an estimated mean value and standard deviation of the sea-
sonal 24-hour-average concentration of PM25 attributable to primary
emissions from traf 1047297c along the roadway corridor under each scenario
Within each census block and for each scenario the seasonal average
traf 1047297c-related 24-hour PM25 concentration then was represented as a
normaldistribution(left-truncatedat zero) with themean andstandard
deviation estimated from the corresponding 2000 bootstrap simula-
tions The TIA estimated hourly traf 1047297c counts for each scenario along
each roadway link only for weekdays we assumed traf 1047297c counts on
weekends would be the same and hence may have slightly over-
estimated exposure concentrations
In addition to considering variability in PM25exposures arising fromprimary traf 1047297c emissions we assessed the effects of uncertainty in the
accuracy of the air quality model predictions Our previous research
on the integrated air quality modeling approach as well as previous
work by others suggests that the combined MOVESndashCAL3QHCR
model generally predicts PM25 concentration within a factor of two of
measured concentrations (although accuracy varies with local condi-
tions and the quality of data available to support the model) ( Chart-
asa etal2013 Yura etal2007) FollowingMorganand Henrionsguid-
ance (Morgan et al 1990) we represented model uncertainty with an
uncertainty factor (UF ) parameterized by a triangular probability distri-
bution with lower limit = 05 upper limit = 20 and mode = 10
(spanning the expected factor-of-two uncertainty in the model) Ac-
cording to Morgan and Henrion the triangular distribution is especially
appropriate for situations in which ldquothe distributions of variables in a
Fig 2 Overview of framework for incorporating variability and uncertainty into assessment of the health impacts of traf 1047297c-related PM25 The rectangles show sources of variability and
uncertainty The shaded diamonds show the three major information categories needed for quantitative health impact assessment
413C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 613
model arenot preciselyknownrdquo but in which ldquovaluestowardthe middle
of the range of possible values are considered more likely to occur than
values near either extremerdquo Based on previous evaluations of the per-
formance of near-roadway air pollutant dispersion models the exact
form of the distribution representing model uncertainty is not known
making the triangular distribution an appropriate choice for character-
izing model uncertainty Correspondingly in each census block the ex-
cess PM25 24-hour average exposure concentration attributable to
primary emissions from traf 1047297
c on the case study roadway was estimat-ed for each season as
PM exposureimfrac14 U F PM modelim
eth1THORN
where PM exposureimrepresents the average 24-hour PM25 concentration
in census block i (i = 1ndash160) and season m (m = winter spring sum-
mer fall) attributable to primary traf 1047297c emissions on the case study
roadway UF is the model uncertainty factor and PM modelimis thecorre-
sponding model-predicted seasonal daily average PM25 concentration
arising from primary emissions from traf 1047297c
321 Concentrationndashresponse functions
As recommended by the World Health Organization and others
(Ostro and Chestnut 1998 Li et al 2010 Aunan 1996) we use
the following relationship to describe the link between seasonal
daily average PM25 concentrations and the relative risk of cardio-
vascular and respiratory health outcomes
RRimn frac14 e β mn P M exposureim eth2THORN
where β mn is the concentrationndashresponse coef 1047297cient describing the
effects of PM on health outcome n during season m and RRimn is the
relative risk of health outcome n during season m in census block i
The number of adverse health cases in the population attributable
to traf 1047297c-related PM25 then can be determined from the following
relationship
Δ yi jklmn frac14 y0i jklmn A F i jklmn 3a
frac14 y0i jklmn
RRimnminus1
RRimn
3b
frac14 y0i jklmn
e β mn P M exposureimminus1
e β mn P M exposureim
3c
frac14 y0i jklmn 1minuse
minus β mn P M exposureim
3d
where AF i jklmn and Δ yi jklmn are the fraction and number of casesof adverse health event n attributable to traf 1047297c-related PM25 in sea-
son m in census block i for age group j gender k and race l and
where yi jklmn0 is the observed total number of cases in the same lo-
cation and among the same population group Eqs (2) (3a) (3b)
(3c) and (3d) are the standard equations used in analyses by the
WHO and other organizations to attribute observed cases of ad-
verse health events to speci1047297c risk factors ( Ostro and Chestnut
1998 Murray et al 2003 Mathers et al 2001 Pruumlss-uumlstuumln et al
2003)
The β values in Eqs (2) (3c) and (3d) (known as dosendashresponse co-
ef 1047297cients) were drawn from the US Environmental Protection Agency
guidance document Quantitative Health Risk Assessment for Particulate
Matter (US Environmental Protection Agency 2010 Zanobetti and
Schwartz 2009 Bell et al 2008) Table 3 shows the coef 1047297cient values
used in this analysis EPA retrieved these coef 1047297cients from peer-
reviewed epidemiologic studies that met certain quality-assurance
criteria including for example the estimation of exposure from mea-
sured rather than modeled PM25 data For mortality effects the coef 1047297-
cients are speci1047297c to 15 US metropolitan areas For morbidity effects
coef 1047297cients are speci1047297c to region (Northeast Southeast Northwest
and Southwest) This study employed mortality coef 1047297cients developed
from studies in Atlanta since Atlanta is climatologically the most similar
to Chapel Hill among the 15 cities studied We used morbiditycoef 1047297cients for the Southeast region in which Chapel Hill is located
All concentrationndashresponse coef 1047297cients were represented as normal
distributions with all negative valuestruncated at zero (to avoid associ-
ating PM exposure with positive health effects) Standard deviationsfor
each season and health outcome were estimated from the con1047297dence
intervals in Table 3
33 Baseline incidence rates of adverse health outcomes
Data on baseline incidence rates of health outcomes were obtained
from North Carolina public health databases Annual mortality rates
for each age group (Table 4) were calculated by dividing thetotal num-
ber of deaths in Orange County (where Chapel Hill is located) in 2010
(North Carolina State Center for Health Statistics 2012) bythe 2010Or-
ange County census population (Minnesota Population Center 2011)
Annual unscheduled hospital admission rates (Table 5) were obtained
from 2009 emergency department visit data reported by the North Car-
olina Disease Event Tracking and Epidemiologic Collection Tool (NC DE-
TECT) (University of North Carolina at Chapel Hill 2011) We were
unable to obtain data on incidence rates by gender and race so we as-
sume that incidence rates are the same for both genders and all races
(which is a limitation of this analysis) It should be noted as well that
the ICD codes speci1047297c to the concentrationndashresponse coef 1047297cients
might not be entirely matched to the ICD codes speci1047297c to the incidence
rates used in this study depending on reported data Moreover emer-
gency department visits may not result in hospital admissions and
some hospital admissions may occur without 1047297rst visiting the emergen-
cy department
Tore1047298ect seasonal variation we adjusted the incidence rates for car-diovascular and respiratory mortality and unscheduled hospital admis-
sions using data on temporal variability in cardiovascular and
respiratory deaths in Orange County during 1999ndash2010 from the CDC
WONDER database (Centers for Disease Control and Prevention
2013) The fractions for cardiovascular events are 025 031 020 and
024 for winter spring summer and fall respectively while the frac-
tions for respiratory events are 030 026 021 and 023 for winter
spring summer and fall respectively
To determine the total number of cases in any given season (ie
yi jklmn0 in Eqs (3a) (3b) (3c) and (3d)) we multiplied the given inci-
dence rateby the corresponding size of each demographic group in each
census block
34 Testing the effects of variability and uncertainty on health impact estimates
Five simulations of 2000 iterations each were run using Analytica
version 45 (Lumina Decision Systems Los Gatos California) to demon-
strate differences in health burden estimates when including variability
and uncertainty Table 6 lists the1047297ve simulations and the variability and
uncertainty considered in each The1047297rst simulation (1a) follows the de-
terministic approach of previous HIAs usingaverage traf 1047297c volumes and
a constant traf 1047297c emission factor corresponding to traf 1047297c cruising at
35 mph on a 1047298at roadway under a constant ambient temperature of
70 degF Like previous HIAs simulation 1a accounts for neither uncertainty
in the concentrationndashresponse coef 1047297cient (usingthe meanvalue asa de-
terministic estimate) nor seasonal variability Simulation 1b is identical
to simulation 1a except that it uses seasonal concentrationndashresponse
414 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 713
coef 1047297cients (also deterministic) Simulations 2ndash4 systematically include
(one at a time) variability in vehicle emissions rates (simulations 2ndash4)
uncertainty in concentrationndashresponse coef 1047297cients (simulations 3ndash4)
and air quality model prediction error (simulation 4)
35 Comparing health impacts under alternative scenarios
As noted previously we simulated health impacts for the full study
corridor (160 census blocks) for three different scenarios (Table 2)
year 2009 and year 2025 with and without the new campus For each
scenario 2000 simulations were run in Analytica Traf 1047297c patterns (traf-
1047297c volumes along each roadway link in the corridor) for each scenario
were taken from a previous traf 1047297c impact analysis conducted for theTownof Chapel Hill(Vanasse Hangen Brustlin Inc2009)The 2025cen-
sus block populations were obtained from forecasts by the North
Carolina Capital Area Metropolitan Planning Organization (2005)
These growth rates account for demographic changes expected to
occur if the Carolina North campus is built
4 Results and discussion
This analysis explored the effects of variability and uncertainty on
health impact estimates of near-roadway air pollution arising from traf-1047297c attracted by newsuburban development projects Most previous US
HIAs of such projects have provided qualitative rather than quantitative
assessments of health impacts the few quantitative HIAs have not
systematically represented variability and uncertainty in the variables
used to estimate health impacts or in the resulting health outcome pre-
dictions We explored whether including variability and uncertainty
makes a difference in centralestimates of health impacts and we exam-
ined the magnitude of uncertainty in the resulting estimates We then
employed an approach that accounts for variability and uncertainty to
model the expected health impacts in the year 2025 of new traf 1047297c gen-
erated by a new research campus development along a busy roadway
corridor in Chapel Hill North Carolina
41 Effect of including variability and uncertainty
Our results suggest that the conventional deterministic HIA ap-
proach may systematically under-estimate potential health impacts of
traf 1047297c-related PM25 exposure (Fig 3)
Incorporating traf 1047297c emission variability into the analysis (as in sim-
ulation 2) caused the mean value of estimated health impacts to in-
crease by more than a factor of two compared to estimates that
exclude such variability (simulation 1b) This increase occurred because
neglecting the effects on vehicle emissions of variability in temperature
road grade vehicle speed and traf 1047297c behavior (idling accelerating de-
celerating or cruising) resulted in under-estimates of PM25 exposure
Table 3
Concentrationndashresponse coef 1047297cients used in this study
Health outcome Disease category ICD-9 or ICD-10 codea Age group Season Mean concentration-response
coef 1047297cient (95 CI) per 10 μ gm3b
Mortality Cardiovascular I01ndashI59 All ages All-yearc 066 (minus066 198)
Winter 135 (minus193 462)
Spring 076 (minus273 425)
Summer 062 (minus222 347)
Fall minus018 (minus293 257)
Respiratory J00ndash J99 All ages All-yearc
121 (minus048 290)Winter 093 (minus144 329)
Spring 035 (minus205 275)
Summer 077 (minus155 310)
Fall 096 (minus134 325)
Unscheduled hospital admissions Cardiovascular 410ndash414 426ndash429 430ndash438 and 440ndash4 49 6 5 a nd over A ll-yea rc 029 (minus019 077)
Winter 105 (minus007 219)
Spring 075 (minus026 176)
Summer minus067 (minus161 026)
Fall 017 (minus072 106)
Respiratory 464ndash466 480ndash487 and 490ndash492 65 and over All-yearc 035 (minus044 113)
Winter 040 (minus146 224)
Spring 075 (minus082 231)
Summer minus052 (minus209 105)
Fall 014 (minus130 158)
a ICD-10 for mortality and ICD-9 for unscheduled hospital admissionsb Coef 1047297cients were originally from Zanobetti and Schwartz (2009) and Bell et al (2008) respectively
c Used only in simulation 1
Table 4Annual mortality rates by race gender and age group for Orange County (per 1000 people)
Cause of death ICD-10 code Age group Race and gender
White male Black male Other male White female Black female Other female
Cardiovascular disease I05ndashI09 I10ndashI15 I20ndashI25 I26ndashI28 and I30ndashI52 0 to 34 000 000 000 000 000 000
35 to 44 017 000 000 000 095 000
45 to 54 029 272 000 013 153 000
55 to 64 179 235 257 091 105 000
65 to 74 343 725 000 244 000 000
75 to 84 1647 1724 000 770 2065 000
85+ 5251 2632 12500 3034 2299 000
Respiratory disease J00ndash J99 0 to 54 000 000 000 000 000 000
55 to 64 049 000 000 045 105 000
65 to 74 206 483 000 183 345 000
75 to 84 524 575 116 495 590 000
85+ 3580 1316 000 787 1149 000
415C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 313
estimates fail to provide information to decision-makers about the de-
gree of certainty in the estimated risks For example decision-makers
may be more concerned about a risk factor with a relatively low central
riskestimate (for example 1 in10000) ifthere isa good chancethatthe
risk could be much higher than thecentral estimate (for example a 10
chance of the risk exceeding 1 in 100) than they would be if presented
only with the central estimate of risk
Variability and uncertainty in estimated risks of traf 1047297c-related air
pollution can arise from multiple sources Variability arises naturallydue to differences in members of a population weather patterns traf 1047297c
geographic features and so on it is a property of nature ldquousually not re-
ducible through further measurement or studyrdquo (Frey and Burmaster
1999) On theother hand uncertainty arises due to thelack of informa-
tion or knowledge including limited data on a population partial igno-
rance of phenomena in1047298uencing a particular risk and disagreements
between models and the reality they are intended to represent (Frey
and Burmaster 1999) Example sources of uncertainty include the
mathematical form used to predict the effects of changes in pollution
exposure on public health the parameters in such mathematical equa-
tions and the accuracy of models predicting air pollution levels under
different traf 1047297c scenarios Theoretically uncertainty can be decreased
through further studies
This paper aims to strengthen the knowledge base and toolset avail-
able to HIA practitioners wishing to incorporate variability and uncer-
tainty in quantitative transportation-related HIAs Like the HIAs listed
in Table 1 this analysis focuses on a potential new land development
expected to increase future traf 1047297c on a major municipal road corridor
(see the ldquoCase study siterdquo section) The potential for increased traf 1047297c
has raised concerns about increases in air pollution and its associated
adverse health effects including increased risks of cardiovascular and
respiratory diseases Like four of the1047297ve HIAs in Table 1 the analysis fo-
cuses on airborne particulate matter having a diameter less than 25μ m
(denoted as PM25) as an indicator of traf 1047297c-related air pollution Like
the other HIAs this analysis is restricted to primary PM 25 (that is
PM25 emitted directly by vehicle operations rather than that formed
by chemical reactions in the atmosphere) This study considers the ef-
fects of short-term exposure to traf 1047297c-related PM25 on cardiovascular
and respiratory mortality (all ages) and unscheduled hospital admis-sions (age 65 and over) These health outcomes were previously select-
ed for the core analysis in the US Environmental Protection Agencys
(EPAs) quantitative health risk assessment for supporting the review
of the US National Ambient Air Quality Standards for PM (US
Environmental Protection Agency 2010)
We use the case study road corridor to explore the effects on health
impact estimates of PM25 from roadway traf 1047297c when including or ex-
cluding various sources of variability and uncertainty We 1047297rst use a
portion of the road corridor to explore the question ldquoWhich variability
and uncertainty sources have the greatest effects on the mean values
and upper con1047297dence limits of estimated health risksrdquo Then we dem-
onstrate a method for incorporating the key variability and uncertainty
sources in a comprehensive assessment of potential air pollution-
related health risks for the entire case study roadway corridor undercurrent conditions and future conditions with and without the pro-
posed new development
2 Case study site
We demonstrate the suggested new assessment process to explore
some of the potential health impacts arising from a planned new cam-
pus extension for the University of North Carolina (UNC) at Chapel
Hill The new campus called ldquoCarolina Northrdquo is intended to increase
the universitys capability to translate research into applications It will
be located about 3 km (2 miles) north of the existing campus ( Fig 1)
If constructed it is expected to increase the number of trips to the
area by 10000 per day by 2015 with most of the increases expected
to occur along MartinLuther King Jr Blvd themain link to the existing
campus and the major northndashsouth road corridor in Chapel Hill (Vanasse
Hangen Brustlin Inc 2009) By 2025 the number of additional daily trips
to the campus is expected to increase by as many as 40000 We consider
the potential impacts of the expected additional traf 1047297c-related air pollu-
tion among residents living in census blocks within 500 m of Martin Lu-
ther King Jr Blvd In all this area encompasses 160 US census blocks
(see Fig 1) and has a total population of about 16000mdashmore than one-
quarter of Chapel Hills total population of 57000
We analyze the effects of primary emissions from traf 1047297
c along Mar-tin Luther King Jr Blvd on ambient PM25 concentrations and popula-
tion health under three different scenarios (1) the year 2009
(2) 2025 assuming the new campus is not built and (3) 2025 with
the new campus The baseline comparison year is 2009 because the
most comprehensive transportation analysis of the study corridor was
conducted using 2009 data (Vanasse Hangen Brustlin Inc 2009)
Table 2 provides summary information about the population size and
traf 1047297c volumes under these three scenarios
3 Methods and data sources
This analysis has two main parts
1) Analyze the effects of including variability and uncertainty in the HIA
First we investigate in the effects on health impact estimates of in-cludingseveral different uncertainty and variability sources as com-
pared to results obtained using the conventional deterministic
approach For computational ef 1047297ciency we focus on the 12 census
blocks highlighted in Fig 1B which our prior air quality modeling in-
dicated are more vulnerable to traf 1047297c-related PM25 than most other
census blocks in the corridor (Chart-asa et al 2013) The total pop-
ulation in the 12 blocks is 1117 (about 7 of the total population in
the study corridor)
2) Quantify the health impacts of traf 1047297c from the proposed new campus in
the study corridor Second we quantify traf 1047297c-related air quality and
health outcomes along the entire study corridor for the three devel-
opment scenarios in Table 2 This analysis includes the variability
and uncertainty sources identi1047297ed in part 1 as having an in1047298uence
on the central estimates or upper con1047297
dence estimates of themodeled risks
Both analyses use the same modeling framework describedin detail
in the followingsections However the1047297rst analysis introducesvariabil-
ity and uncertainty sources one at a time in order to explore their po-
tential in1047298uence on the computed health risks while the second
analysis includes all key variability and uncertainty sources
31 Modeling framework overview
Quantifying the health impacts of traf 1047297c-related air pollution re-
quires three categories of information (1) estimates of the excess
PM25 concentrations to which the population is exposed as a result of
primary emissions from traf 1047297c (2) concentrationndashresponse functions
relating exposure concentrations to probabilities of adverse health out-comes and (3) incidence rates of the health outcomes of concern (from
all causes) in the exposed population (Ostro 2004 Ostro and Chestnut
1998 Cohen et al 2005 Li et al 2010 ) Fig 2 summarizes how this
analysis combines these three information categories (shows as shaded
boxes)to estimate health impacts Theunshaded boxes show variability
and uncertainty sources considered in this study The subscript notation
indicatesthat theanalysisis conductedat thecensus block scalewhere i
represents an individual block That is health risks are characterized
separately for each census block considering variability in traf 1047297c-
related PM25 exposure concentrations and population demographic
characteristics within each block The subscripts j k and l indicate dif-
ferences in baseline health status by age ( j) gender (k) and race (l)
In addition this analysis considers seasonal (subscript m) variability
because epidemiologic evidence suggests seasonal differences in
411C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 413
dosendashresponse functions (Zanobetti and Schwartz 2009 Bell et al
2008) The following sections provide details on our methods for esti-
mating PM25 exposure concentrations (left-most shaded diamond in
Fig 2) selecting concentrationndashresponse functions (central diamond)
estimating baseline incidence rates of adverse health outcomes in the
study population (right-most diamond) and incorporating variability
and uncertainty (white rectangles) into the analysis
32 PM 25 concentrations attributable to primary emissions from traf 1047297c
The 24-hour exposures to PM25 arising from primary emissions
from traf 1047297c along the case study roadway corridor were estimated
using an integrated air quality modeling approach described in Chart-
asa et al (2013) In brief the approach employs standard traf 1047297c emis-
sions and air quality dispersion modeling tools but it adds a novel ap-
proach for modeling variability in vehicle emissions due to variability
in hourly temperature roadgrade and traf 1047297c behavior (including cruis-
ing speed and percent time spent idling decelerating and accelerating)
The exposure modeling approach links a novel application of MOVES2010b commonly used in the United States to estimate vehicle emis-
sions factors (gvehicle-mile) and CAL3QHCR which characterizes
PM25 dispersion away from roadways By linking these models and
employing a new approach for characterizing variabilityin emission fac-
tors we simulated probability distributions of the average 24-hour
Fig 1 (A)The study corridorbetween the intersection of Martin LutherKing JrBlvdand Whit1047297eld Rdand theintersection of South ColumbiaSt andMt Carmel Church Rd Chapel Hill
NC and the census blocks located within 500 m from the study corridor (B) The road segment and census blocks for simulations to demonstrate differences in health burden estimates
when including variability and the uncertainty in the modeling approach Dots represent census block centroids
Table 2
Population size and traf 1047297c volumes under three scenarios considered
Scenario Traf 1047297c volumes of road segments on study corridor (vehh) a Total population of 160 census blocks located within 500 m from study corridor
2009 4ndash1758 16042
2025 without the new campus 5ndash2443 19140b
2025 with the new campus 5ndash2832 19140b
a Ranges indicate variability in traf 1047297c 1047298ow by road segment day of week and time of dayb
Computed from growth rates forecasted by the North Carolina Capital Area Metropolitan Planning Organization (2005)
412 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 513
PM25 concentration in each season (winter spring summer and fall) at
the centroid of each of the 160 census blocks in the study corridorOur analysis considers variability in vehicle emission factors by di-
viding the 82-km roadway corridor into 1200 links and estimating sep-
arate emission factors for each link for each hour of each simulation day
(Chart-asa et al 2013) Unlike previous studies linking MOVES and
CAL3QHCR our analysis considers hourly variability in temperature
and link-speci1047297c variability in road grade and vehicle behavior Hourly
meteorological pro1047297les for 2006ndash2012 were obtained from the national
weather stations in ChapelHill andGreensboroNorthCarolinaand me-
teorological pro1047297les for input to CAL3QHCR were generated from EPAs
Meteorological Processor for Regulatory Models (NCDC 2013 NOAA
2013) The meteorological pro1047297les contained a total of 2100 days with
complete required data (525 days for winter 560 days for spring
532 days for summer and 483 days for fall) For each census block we
used CAL3QHCR to estimate the PM25 concentration (averaged over24 h) attributable to primary traf 1047297c emissions from each of the 1200
roadway links for each of the 2100 days for which meteorological data
were available Separate estimates were prepared for 2009 and 2025
using emission factors from MOVES modeling and simulated traf 1047297c
data for 2009 and 2025 scenarios with or without Carolina North from
the Transportation Impact Analysis (TIA) for the Carolina North Devel-
opment (Vanasse Hangen Brustlin Inc 2009) Then for each develop-
ment scenario seven separate mean estimatesmdashone for each of the
seven years for which meteorological data were availablemdashof the sea-
sonal mean value of the24-hour average PM25exposure concentrations
were computed for each season
For each scenario (year 2009 and year 2025 with and without con-
structingCarolina North) andeach season we then computed bootstrap
estimatesof the mean value andstandarddeviation of theseasonal daily
average PM25 exposure concentrations by randomly selecting one of
the seven years assigning the associated seasonal mean concentrationsas computed using that years data to each census block and then re-
peating the process 1999 times For each of the 160 census blocks the
result was an estimated mean value and standard deviation of the sea-
sonal 24-hour-average concentration of PM25 attributable to primary
emissions from traf 1047297c along the roadway corridor under each scenario
Within each census block and for each scenario the seasonal average
traf 1047297c-related 24-hour PM25 concentration then was represented as a
normaldistribution(left-truncatedat zero) with themean andstandard
deviation estimated from the corresponding 2000 bootstrap simula-
tions The TIA estimated hourly traf 1047297c counts for each scenario along
each roadway link only for weekdays we assumed traf 1047297c counts on
weekends would be the same and hence may have slightly over-
estimated exposure concentrations
In addition to considering variability in PM25exposures arising fromprimary traf 1047297c emissions we assessed the effects of uncertainty in the
accuracy of the air quality model predictions Our previous research
on the integrated air quality modeling approach as well as previous
work by others suggests that the combined MOVESndashCAL3QHCR
model generally predicts PM25 concentration within a factor of two of
measured concentrations (although accuracy varies with local condi-
tions and the quality of data available to support the model) ( Chart-
asa etal2013 Yura etal2007) FollowingMorganand Henrionsguid-
ance (Morgan et al 1990) we represented model uncertainty with an
uncertainty factor (UF ) parameterized by a triangular probability distri-
bution with lower limit = 05 upper limit = 20 and mode = 10
(spanning the expected factor-of-two uncertainty in the model) Ac-
cording to Morgan and Henrion the triangular distribution is especially
appropriate for situations in which ldquothe distributions of variables in a
Fig 2 Overview of framework for incorporating variability and uncertainty into assessment of the health impacts of traf 1047297c-related PM25 The rectangles show sources of variability and
uncertainty The shaded diamonds show the three major information categories needed for quantitative health impact assessment
413C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 613
model arenot preciselyknownrdquo but in which ldquovaluestowardthe middle
of the range of possible values are considered more likely to occur than
values near either extremerdquo Based on previous evaluations of the per-
formance of near-roadway air pollutant dispersion models the exact
form of the distribution representing model uncertainty is not known
making the triangular distribution an appropriate choice for character-
izing model uncertainty Correspondingly in each census block the ex-
cess PM25 24-hour average exposure concentration attributable to
primary emissions from traf 1047297
c on the case study roadway was estimat-ed for each season as
PM exposureimfrac14 U F PM modelim
eth1THORN
where PM exposureimrepresents the average 24-hour PM25 concentration
in census block i (i = 1ndash160) and season m (m = winter spring sum-
mer fall) attributable to primary traf 1047297c emissions on the case study
roadway UF is the model uncertainty factor and PM modelimis thecorre-
sponding model-predicted seasonal daily average PM25 concentration
arising from primary emissions from traf 1047297c
321 Concentrationndashresponse functions
As recommended by the World Health Organization and others
(Ostro and Chestnut 1998 Li et al 2010 Aunan 1996) we use
the following relationship to describe the link between seasonal
daily average PM25 concentrations and the relative risk of cardio-
vascular and respiratory health outcomes
RRimn frac14 e β mn P M exposureim eth2THORN
where β mn is the concentrationndashresponse coef 1047297cient describing the
effects of PM on health outcome n during season m and RRimn is the
relative risk of health outcome n during season m in census block i
The number of adverse health cases in the population attributable
to traf 1047297c-related PM25 then can be determined from the following
relationship
Δ yi jklmn frac14 y0i jklmn A F i jklmn 3a
frac14 y0i jklmn
RRimnminus1
RRimn
3b
frac14 y0i jklmn
e β mn P M exposureimminus1
e β mn P M exposureim
3c
frac14 y0i jklmn 1minuse
minus β mn P M exposureim
3d
where AF i jklmn and Δ yi jklmn are the fraction and number of casesof adverse health event n attributable to traf 1047297c-related PM25 in sea-
son m in census block i for age group j gender k and race l and
where yi jklmn0 is the observed total number of cases in the same lo-
cation and among the same population group Eqs (2) (3a) (3b)
(3c) and (3d) are the standard equations used in analyses by the
WHO and other organizations to attribute observed cases of ad-
verse health events to speci1047297c risk factors ( Ostro and Chestnut
1998 Murray et al 2003 Mathers et al 2001 Pruumlss-uumlstuumln et al
2003)
The β values in Eqs (2) (3c) and (3d) (known as dosendashresponse co-
ef 1047297cients) were drawn from the US Environmental Protection Agency
guidance document Quantitative Health Risk Assessment for Particulate
Matter (US Environmental Protection Agency 2010 Zanobetti and
Schwartz 2009 Bell et al 2008) Table 3 shows the coef 1047297cient values
used in this analysis EPA retrieved these coef 1047297cients from peer-
reviewed epidemiologic studies that met certain quality-assurance
criteria including for example the estimation of exposure from mea-
sured rather than modeled PM25 data For mortality effects the coef 1047297-
cients are speci1047297c to 15 US metropolitan areas For morbidity effects
coef 1047297cients are speci1047297c to region (Northeast Southeast Northwest
and Southwest) This study employed mortality coef 1047297cients developed
from studies in Atlanta since Atlanta is climatologically the most similar
to Chapel Hill among the 15 cities studied We used morbiditycoef 1047297cients for the Southeast region in which Chapel Hill is located
All concentrationndashresponse coef 1047297cients were represented as normal
distributions with all negative valuestruncated at zero (to avoid associ-
ating PM exposure with positive health effects) Standard deviationsfor
each season and health outcome were estimated from the con1047297dence
intervals in Table 3
33 Baseline incidence rates of adverse health outcomes
Data on baseline incidence rates of health outcomes were obtained
from North Carolina public health databases Annual mortality rates
for each age group (Table 4) were calculated by dividing thetotal num-
ber of deaths in Orange County (where Chapel Hill is located) in 2010
(North Carolina State Center for Health Statistics 2012) bythe 2010Or-
ange County census population (Minnesota Population Center 2011)
Annual unscheduled hospital admission rates (Table 5) were obtained
from 2009 emergency department visit data reported by the North Car-
olina Disease Event Tracking and Epidemiologic Collection Tool (NC DE-
TECT) (University of North Carolina at Chapel Hill 2011) We were
unable to obtain data on incidence rates by gender and race so we as-
sume that incidence rates are the same for both genders and all races
(which is a limitation of this analysis) It should be noted as well that
the ICD codes speci1047297c to the concentrationndashresponse coef 1047297cients
might not be entirely matched to the ICD codes speci1047297c to the incidence
rates used in this study depending on reported data Moreover emer-
gency department visits may not result in hospital admissions and
some hospital admissions may occur without 1047297rst visiting the emergen-
cy department
Tore1047298ect seasonal variation we adjusted the incidence rates for car-diovascular and respiratory mortality and unscheduled hospital admis-
sions using data on temporal variability in cardiovascular and
respiratory deaths in Orange County during 1999ndash2010 from the CDC
WONDER database (Centers for Disease Control and Prevention
2013) The fractions for cardiovascular events are 025 031 020 and
024 for winter spring summer and fall respectively while the frac-
tions for respiratory events are 030 026 021 and 023 for winter
spring summer and fall respectively
To determine the total number of cases in any given season (ie
yi jklmn0 in Eqs (3a) (3b) (3c) and (3d)) we multiplied the given inci-
dence rateby the corresponding size of each demographic group in each
census block
34 Testing the effects of variability and uncertainty on health impact estimates
Five simulations of 2000 iterations each were run using Analytica
version 45 (Lumina Decision Systems Los Gatos California) to demon-
strate differences in health burden estimates when including variability
and uncertainty Table 6 lists the1047297ve simulations and the variability and
uncertainty considered in each The1047297rst simulation (1a) follows the de-
terministic approach of previous HIAs usingaverage traf 1047297c volumes and
a constant traf 1047297c emission factor corresponding to traf 1047297c cruising at
35 mph on a 1047298at roadway under a constant ambient temperature of
70 degF Like previous HIAs simulation 1a accounts for neither uncertainty
in the concentrationndashresponse coef 1047297cient (usingthe meanvalue asa de-
terministic estimate) nor seasonal variability Simulation 1b is identical
to simulation 1a except that it uses seasonal concentrationndashresponse
414 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 713
coef 1047297cients (also deterministic) Simulations 2ndash4 systematically include
(one at a time) variability in vehicle emissions rates (simulations 2ndash4)
uncertainty in concentrationndashresponse coef 1047297cients (simulations 3ndash4)
and air quality model prediction error (simulation 4)
35 Comparing health impacts under alternative scenarios
As noted previously we simulated health impacts for the full study
corridor (160 census blocks) for three different scenarios (Table 2)
year 2009 and year 2025 with and without the new campus For each
scenario 2000 simulations were run in Analytica Traf 1047297c patterns (traf-
1047297c volumes along each roadway link in the corridor) for each scenario
were taken from a previous traf 1047297c impact analysis conducted for theTownof Chapel Hill(Vanasse Hangen Brustlin Inc2009)The 2025cen-
sus block populations were obtained from forecasts by the North
Carolina Capital Area Metropolitan Planning Organization (2005)
These growth rates account for demographic changes expected to
occur if the Carolina North campus is built
4 Results and discussion
This analysis explored the effects of variability and uncertainty on
health impact estimates of near-roadway air pollution arising from traf-1047297c attracted by newsuburban development projects Most previous US
HIAs of such projects have provided qualitative rather than quantitative
assessments of health impacts the few quantitative HIAs have not
systematically represented variability and uncertainty in the variables
used to estimate health impacts or in the resulting health outcome pre-
dictions We explored whether including variability and uncertainty
makes a difference in centralestimates of health impacts and we exam-
ined the magnitude of uncertainty in the resulting estimates We then
employed an approach that accounts for variability and uncertainty to
model the expected health impacts in the year 2025 of new traf 1047297c gen-
erated by a new research campus development along a busy roadway
corridor in Chapel Hill North Carolina
41 Effect of including variability and uncertainty
Our results suggest that the conventional deterministic HIA ap-
proach may systematically under-estimate potential health impacts of
traf 1047297c-related PM25 exposure (Fig 3)
Incorporating traf 1047297c emission variability into the analysis (as in sim-
ulation 2) caused the mean value of estimated health impacts to in-
crease by more than a factor of two compared to estimates that
exclude such variability (simulation 1b) This increase occurred because
neglecting the effects on vehicle emissions of variability in temperature
road grade vehicle speed and traf 1047297c behavior (idling accelerating de-
celerating or cruising) resulted in under-estimates of PM25 exposure
Table 3
Concentrationndashresponse coef 1047297cients used in this study
Health outcome Disease category ICD-9 or ICD-10 codea Age group Season Mean concentration-response
coef 1047297cient (95 CI) per 10 μ gm3b
Mortality Cardiovascular I01ndashI59 All ages All-yearc 066 (minus066 198)
Winter 135 (minus193 462)
Spring 076 (minus273 425)
Summer 062 (minus222 347)
Fall minus018 (minus293 257)
Respiratory J00ndash J99 All ages All-yearc
121 (minus048 290)Winter 093 (minus144 329)
Spring 035 (minus205 275)
Summer 077 (minus155 310)
Fall 096 (minus134 325)
Unscheduled hospital admissions Cardiovascular 410ndash414 426ndash429 430ndash438 and 440ndash4 49 6 5 a nd over A ll-yea rc 029 (minus019 077)
Winter 105 (minus007 219)
Spring 075 (minus026 176)
Summer minus067 (minus161 026)
Fall 017 (minus072 106)
Respiratory 464ndash466 480ndash487 and 490ndash492 65 and over All-yearc 035 (minus044 113)
Winter 040 (minus146 224)
Spring 075 (minus082 231)
Summer minus052 (minus209 105)
Fall 014 (minus130 158)
a ICD-10 for mortality and ICD-9 for unscheduled hospital admissionsb Coef 1047297cients were originally from Zanobetti and Schwartz (2009) and Bell et al (2008) respectively
c Used only in simulation 1
Table 4Annual mortality rates by race gender and age group for Orange County (per 1000 people)
Cause of death ICD-10 code Age group Race and gender
White male Black male Other male White female Black female Other female
Cardiovascular disease I05ndashI09 I10ndashI15 I20ndashI25 I26ndashI28 and I30ndashI52 0 to 34 000 000 000 000 000 000
35 to 44 017 000 000 000 095 000
45 to 54 029 272 000 013 153 000
55 to 64 179 235 257 091 105 000
65 to 74 343 725 000 244 000 000
75 to 84 1647 1724 000 770 2065 000
85+ 5251 2632 12500 3034 2299 000
Respiratory disease J00ndash J99 0 to 54 000 000 000 000 000 000
55 to 64 049 000 000 045 105 000
65 to 74 206 483 000 183 345 000
75 to 84 524 575 116 495 590 000
85+ 3580 1316 000 787 1149 000
415C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 413
dosendashresponse functions (Zanobetti and Schwartz 2009 Bell et al
2008) The following sections provide details on our methods for esti-
mating PM25 exposure concentrations (left-most shaded diamond in
Fig 2) selecting concentrationndashresponse functions (central diamond)
estimating baseline incidence rates of adverse health outcomes in the
study population (right-most diamond) and incorporating variability
and uncertainty (white rectangles) into the analysis
32 PM 25 concentrations attributable to primary emissions from traf 1047297c
The 24-hour exposures to PM25 arising from primary emissions
from traf 1047297c along the case study roadway corridor were estimated
using an integrated air quality modeling approach described in Chart-
asa et al (2013) In brief the approach employs standard traf 1047297c emis-
sions and air quality dispersion modeling tools but it adds a novel ap-
proach for modeling variability in vehicle emissions due to variability
in hourly temperature roadgrade and traf 1047297c behavior (including cruis-
ing speed and percent time spent idling decelerating and accelerating)
The exposure modeling approach links a novel application of MOVES2010b commonly used in the United States to estimate vehicle emis-
sions factors (gvehicle-mile) and CAL3QHCR which characterizes
PM25 dispersion away from roadways By linking these models and
employing a new approach for characterizing variabilityin emission fac-
tors we simulated probability distributions of the average 24-hour
Fig 1 (A)The study corridorbetween the intersection of Martin LutherKing JrBlvdand Whit1047297eld Rdand theintersection of South ColumbiaSt andMt Carmel Church Rd Chapel Hill
NC and the census blocks located within 500 m from the study corridor (B) The road segment and census blocks for simulations to demonstrate differences in health burden estimates
when including variability and the uncertainty in the modeling approach Dots represent census block centroids
Table 2
Population size and traf 1047297c volumes under three scenarios considered
Scenario Traf 1047297c volumes of road segments on study corridor (vehh) a Total population of 160 census blocks located within 500 m from study corridor
2009 4ndash1758 16042
2025 without the new campus 5ndash2443 19140b
2025 with the new campus 5ndash2832 19140b
a Ranges indicate variability in traf 1047297c 1047298ow by road segment day of week and time of dayb
Computed from growth rates forecasted by the North Carolina Capital Area Metropolitan Planning Organization (2005)
412 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 513
PM25 concentration in each season (winter spring summer and fall) at
the centroid of each of the 160 census blocks in the study corridorOur analysis considers variability in vehicle emission factors by di-
viding the 82-km roadway corridor into 1200 links and estimating sep-
arate emission factors for each link for each hour of each simulation day
(Chart-asa et al 2013) Unlike previous studies linking MOVES and
CAL3QHCR our analysis considers hourly variability in temperature
and link-speci1047297c variability in road grade and vehicle behavior Hourly
meteorological pro1047297les for 2006ndash2012 were obtained from the national
weather stations in ChapelHill andGreensboroNorthCarolinaand me-
teorological pro1047297les for input to CAL3QHCR were generated from EPAs
Meteorological Processor for Regulatory Models (NCDC 2013 NOAA
2013) The meteorological pro1047297les contained a total of 2100 days with
complete required data (525 days for winter 560 days for spring
532 days for summer and 483 days for fall) For each census block we
used CAL3QHCR to estimate the PM25 concentration (averaged over24 h) attributable to primary traf 1047297c emissions from each of the 1200
roadway links for each of the 2100 days for which meteorological data
were available Separate estimates were prepared for 2009 and 2025
using emission factors from MOVES modeling and simulated traf 1047297c
data for 2009 and 2025 scenarios with or without Carolina North from
the Transportation Impact Analysis (TIA) for the Carolina North Devel-
opment (Vanasse Hangen Brustlin Inc 2009) Then for each develop-
ment scenario seven separate mean estimatesmdashone for each of the
seven years for which meteorological data were availablemdashof the sea-
sonal mean value of the24-hour average PM25exposure concentrations
were computed for each season
For each scenario (year 2009 and year 2025 with and without con-
structingCarolina North) andeach season we then computed bootstrap
estimatesof the mean value andstandarddeviation of theseasonal daily
average PM25 exposure concentrations by randomly selecting one of
the seven years assigning the associated seasonal mean concentrationsas computed using that years data to each census block and then re-
peating the process 1999 times For each of the 160 census blocks the
result was an estimated mean value and standard deviation of the sea-
sonal 24-hour-average concentration of PM25 attributable to primary
emissions from traf 1047297c along the roadway corridor under each scenario
Within each census block and for each scenario the seasonal average
traf 1047297c-related 24-hour PM25 concentration then was represented as a
normaldistribution(left-truncatedat zero) with themean andstandard
deviation estimated from the corresponding 2000 bootstrap simula-
tions The TIA estimated hourly traf 1047297c counts for each scenario along
each roadway link only for weekdays we assumed traf 1047297c counts on
weekends would be the same and hence may have slightly over-
estimated exposure concentrations
In addition to considering variability in PM25exposures arising fromprimary traf 1047297c emissions we assessed the effects of uncertainty in the
accuracy of the air quality model predictions Our previous research
on the integrated air quality modeling approach as well as previous
work by others suggests that the combined MOVESndashCAL3QHCR
model generally predicts PM25 concentration within a factor of two of
measured concentrations (although accuracy varies with local condi-
tions and the quality of data available to support the model) ( Chart-
asa etal2013 Yura etal2007) FollowingMorganand Henrionsguid-
ance (Morgan et al 1990) we represented model uncertainty with an
uncertainty factor (UF ) parameterized by a triangular probability distri-
bution with lower limit = 05 upper limit = 20 and mode = 10
(spanning the expected factor-of-two uncertainty in the model) Ac-
cording to Morgan and Henrion the triangular distribution is especially
appropriate for situations in which ldquothe distributions of variables in a
Fig 2 Overview of framework for incorporating variability and uncertainty into assessment of the health impacts of traf 1047297c-related PM25 The rectangles show sources of variability and
uncertainty The shaded diamonds show the three major information categories needed for quantitative health impact assessment
413C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 613
model arenot preciselyknownrdquo but in which ldquovaluestowardthe middle
of the range of possible values are considered more likely to occur than
values near either extremerdquo Based on previous evaluations of the per-
formance of near-roadway air pollutant dispersion models the exact
form of the distribution representing model uncertainty is not known
making the triangular distribution an appropriate choice for character-
izing model uncertainty Correspondingly in each census block the ex-
cess PM25 24-hour average exposure concentration attributable to
primary emissions from traf 1047297
c on the case study roadway was estimat-ed for each season as
PM exposureimfrac14 U F PM modelim
eth1THORN
where PM exposureimrepresents the average 24-hour PM25 concentration
in census block i (i = 1ndash160) and season m (m = winter spring sum-
mer fall) attributable to primary traf 1047297c emissions on the case study
roadway UF is the model uncertainty factor and PM modelimis thecorre-
sponding model-predicted seasonal daily average PM25 concentration
arising from primary emissions from traf 1047297c
321 Concentrationndashresponse functions
As recommended by the World Health Organization and others
(Ostro and Chestnut 1998 Li et al 2010 Aunan 1996) we use
the following relationship to describe the link between seasonal
daily average PM25 concentrations and the relative risk of cardio-
vascular and respiratory health outcomes
RRimn frac14 e β mn P M exposureim eth2THORN
where β mn is the concentrationndashresponse coef 1047297cient describing the
effects of PM on health outcome n during season m and RRimn is the
relative risk of health outcome n during season m in census block i
The number of adverse health cases in the population attributable
to traf 1047297c-related PM25 then can be determined from the following
relationship
Δ yi jklmn frac14 y0i jklmn A F i jklmn 3a
frac14 y0i jklmn
RRimnminus1
RRimn
3b
frac14 y0i jklmn
e β mn P M exposureimminus1
e β mn P M exposureim
3c
frac14 y0i jklmn 1minuse
minus β mn P M exposureim
3d
where AF i jklmn and Δ yi jklmn are the fraction and number of casesof adverse health event n attributable to traf 1047297c-related PM25 in sea-
son m in census block i for age group j gender k and race l and
where yi jklmn0 is the observed total number of cases in the same lo-
cation and among the same population group Eqs (2) (3a) (3b)
(3c) and (3d) are the standard equations used in analyses by the
WHO and other organizations to attribute observed cases of ad-
verse health events to speci1047297c risk factors ( Ostro and Chestnut
1998 Murray et al 2003 Mathers et al 2001 Pruumlss-uumlstuumln et al
2003)
The β values in Eqs (2) (3c) and (3d) (known as dosendashresponse co-
ef 1047297cients) were drawn from the US Environmental Protection Agency
guidance document Quantitative Health Risk Assessment for Particulate
Matter (US Environmental Protection Agency 2010 Zanobetti and
Schwartz 2009 Bell et al 2008) Table 3 shows the coef 1047297cient values
used in this analysis EPA retrieved these coef 1047297cients from peer-
reviewed epidemiologic studies that met certain quality-assurance
criteria including for example the estimation of exposure from mea-
sured rather than modeled PM25 data For mortality effects the coef 1047297-
cients are speci1047297c to 15 US metropolitan areas For morbidity effects
coef 1047297cients are speci1047297c to region (Northeast Southeast Northwest
and Southwest) This study employed mortality coef 1047297cients developed
from studies in Atlanta since Atlanta is climatologically the most similar
to Chapel Hill among the 15 cities studied We used morbiditycoef 1047297cients for the Southeast region in which Chapel Hill is located
All concentrationndashresponse coef 1047297cients were represented as normal
distributions with all negative valuestruncated at zero (to avoid associ-
ating PM exposure with positive health effects) Standard deviationsfor
each season and health outcome were estimated from the con1047297dence
intervals in Table 3
33 Baseline incidence rates of adverse health outcomes
Data on baseline incidence rates of health outcomes were obtained
from North Carolina public health databases Annual mortality rates
for each age group (Table 4) were calculated by dividing thetotal num-
ber of deaths in Orange County (where Chapel Hill is located) in 2010
(North Carolina State Center for Health Statistics 2012) bythe 2010Or-
ange County census population (Minnesota Population Center 2011)
Annual unscheduled hospital admission rates (Table 5) were obtained
from 2009 emergency department visit data reported by the North Car-
olina Disease Event Tracking and Epidemiologic Collection Tool (NC DE-
TECT) (University of North Carolina at Chapel Hill 2011) We were
unable to obtain data on incidence rates by gender and race so we as-
sume that incidence rates are the same for both genders and all races
(which is a limitation of this analysis) It should be noted as well that
the ICD codes speci1047297c to the concentrationndashresponse coef 1047297cients
might not be entirely matched to the ICD codes speci1047297c to the incidence
rates used in this study depending on reported data Moreover emer-
gency department visits may not result in hospital admissions and
some hospital admissions may occur without 1047297rst visiting the emergen-
cy department
Tore1047298ect seasonal variation we adjusted the incidence rates for car-diovascular and respiratory mortality and unscheduled hospital admis-
sions using data on temporal variability in cardiovascular and
respiratory deaths in Orange County during 1999ndash2010 from the CDC
WONDER database (Centers for Disease Control and Prevention
2013) The fractions for cardiovascular events are 025 031 020 and
024 for winter spring summer and fall respectively while the frac-
tions for respiratory events are 030 026 021 and 023 for winter
spring summer and fall respectively
To determine the total number of cases in any given season (ie
yi jklmn0 in Eqs (3a) (3b) (3c) and (3d)) we multiplied the given inci-
dence rateby the corresponding size of each demographic group in each
census block
34 Testing the effects of variability and uncertainty on health impact estimates
Five simulations of 2000 iterations each were run using Analytica
version 45 (Lumina Decision Systems Los Gatos California) to demon-
strate differences in health burden estimates when including variability
and uncertainty Table 6 lists the1047297ve simulations and the variability and
uncertainty considered in each The1047297rst simulation (1a) follows the de-
terministic approach of previous HIAs usingaverage traf 1047297c volumes and
a constant traf 1047297c emission factor corresponding to traf 1047297c cruising at
35 mph on a 1047298at roadway under a constant ambient temperature of
70 degF Like previous HIAs simulation 1a accounts for neither uncertainty
in the concentrationndashresponse coef 1047297cient (usingthe meanvalue asa de-
terministic estimate) nor seasonal variability Simulation 1b is identical
to simulation 1a except that it uses seasonal concentrationndashresponse
414 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 713
coef 1047297cients (also deterministic) Simulations 2ndash4 systematically include
(one at a time) variability in vehicle emissions rates (simulations 2ndash4)
uncertainty in concentrationndashresponse coef 1047297cients (simulations 3ndash4)
and air quality model prediction error (simulation 4)
35 Comparing health impacts under alternative scenarios
As noted previously we simulated health impacts for the full study
corridor (160 census blocks) for three different scenarios (Table 2)
year 2009 and year 2025 with and without the new campus For each
scenario 2000 simulations were run in Analytica Traf 1047297c patterns (traf-
1047297c volumes along each roadway link in the corridor) for each scenario
were taken from a previous traf 1047297c impact analysis conducted for theTownof Chapel Hill(Vanasse Hangen Brustlin Inc2009)The 2025cen-
sus block populations were obtained from forecasts by the North
Carolina Capital Area Metropolitan Planning Organization (2005)
These growth rates account for demographic changes expected to
occur if the Carolina North campus is built
4 Results and discussion
This analysis explored the effects of variability and uncertainty on
health impact estimates of near-roadway air pollution arising from traf-1047297c attracted by newsuburban development projects Most previous US
HIAs of such projects have provided qualitative rather than quantitative
assessments of health impacts the few quantitative HIAs have not
systematically represented variability and uncertainty in the variables
used to estimate health impacts or in the resulting health outcome pre-
dictions We explored whether including variability and uncertainty
makes a difference in centralestimates of health impacts and we exam-
ined the magnitude of uncertainty in the resulting estimates We then
employed an approach that accounts for variability and uncertainty to
model the expected health impacts in the year 2025 of new traf 1047297c gen-
erated by a new research campus development along a busy roadway
corridor in Chapel Hill North Carolina
41 Effect of including variability and uncertainty
Our results suggest that the conventional deterministic HIA ap-
proach may systematically under-estimate potential health impacts of
traf 1047297c-related PM25 exposure (Fig 3)
Incorporating traf 1047297c emission variability into the analysis (as in sim-
ulation 2) caused the mean value of estimated health impacts to in-
crease by more than a factor of two compared to estimates that
exclude such variability (simulation 1b) This increase occurred because
neglecting the effects on vehicle emissions of variability in temperature
road grade vehicle speed and traf 1047297c behavior (idling accelerating de-
celerating or cruising) resulted in under-estimates of PM25 exposure
Table 3
Concentrationndashresponse coef 1047297cients used in this study
Health outcome Disease category ICD-9 or ICD-10 codea Age group Season Mean concentration-response
coef 1047297cient (95 CI) per 10 μ gm3b
Mortality Cardiovascular I01ndashI59 All ages All-yearc 066 (minus066 198)
Winter 135 (minus193 462)
Spring 076 (minus273 425)
Summer 062 (minus222 347)
Fall minus018 (minus293 257)
Respiratory J00ndash J99 All ages All-yearc
121 (minus048 290)Winter 093 (minus144 329)
Spring 035 (minus205 275)
Summer 077 (minus155 310)
Fall 096 (minus134 325)
Unscheduled hospital admissions Cardiovascular 410ndash414 426ndash429 430ndash438 and 440ndash4 49 6 5 a nd over A ll-yea rc 029 (minus019 077)
Winter 105 (minus007 219)
Spring 075 (minus026 176)
Summer minus067 (minus161 026)
Fall 017 (minus072 106)
Respiratory 464ndash466 480ndash487 and 490ndash492 65 and over All-yearc 035 (minus044 113)
Winter 040 (minus146 224)
Spring 075 (minus082 231)
Summer minus052 (minus209 105)
Fall 014 (minus130 158)
a ICD-10 for mortality and ICD-9 for unscheduled hospital admissionsb Coef 1047297cients were originally from Zanobetti and Schwartz (2009) and Bell et al (2008) respectively
c Used only in simulation 1
Table 4Annual mortality rates by race gender and age group for Orange County (per 1000 people)
Cause of death ICD-10 code Age group Race and gender
White male Black male Other male White female Black female Other female
Cardiovascular disease I05ndashI09 I10ndashI15 I20ndashI25 I26ndashI28 and I30ndashI52 0 to 34 000 000 000 000 000 000
35 to 44 017 000 000 000 095 000
45 to 54 029 272 000 013 153 000
55 to 64 179 235 257 091 105 000
65 to 74 343 725 000 244 000 000
75 to 84 1647 1724 000 770 2065 000
85+ 5251 2632 12500 3034 2299 000
Respiratory disease J00ndash J99 0 to 54 000 000 000 000 000 000
55 to 64 049 000 000 045 105 000
65 to 74 206 483 000 183 345 000
75 to 84 524 575 116 495 590 000
85+ 3580 1316 000 787 1149 000
415C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 513
PM25 concentration in each season (winter spring summer and fall) at
the centroid of each of the 160 census blocks in the study corridorOur analysis considers variability in vehicle emission factors by di-
viding the 82-km roadway corridor into 1200 links and estimating sep-
arate emission factors for each link for each hour of each simulation day
(Chart-asa et al 2013) Unlike previous studies linking MOVES and
CAL3QHCR our analysis considers hourly variability in temperature
and link-speci1047297c variability in road grade and vehicle behavior Hourly
meteorological pro1047297les for 2006ndash2012 were obtained from the national
weather stations in ChapelHill andGreensboroNorthCarolinaand me-
teorological pro1047297les for input to CAL3QHCR were generated from EPAs
Meteorological Processor for Regulatory Models (NCDC 2013 NOAA
2013) The meteorological pro1047297les contained a total of 2100 days with
complete required data (525 days for winter 560 days for spring
532 days for summer and 483 days for fall) For each census block we
used CAL3QHCR to estimate the PM25 concentration (averaged over24 h) attributable to primary traf 1047297c emissions from each of the 1200
roadway links for each of the 2100 days for which meteorological data
were available Separate estimates were prepared for 2009 and 2025
using emission factors from MOVES modeling and simulated traf 1047297c
data for 2009 and 2025 scenarios with or without Carolina North from
the Transportation Impact Analysis (TIA) for the Carolina North Devel-
opment (Vanasse Hangen Brustlin Inc 2009) Then for each develop-
ment scenario seven separate mean estimatesmdashone for each of the
seven years for which meteorological data were availablemdashof the sea-
sonal mean value of the24-hour average PM25exposure concentrations
were computed for each season
For each scenario (year 2009 and year 2025 with and without con-
structingCarolina North) andeach season we then computed bootstrap
estimatesof the mean value andstandarddeviation of theseasonal daily
average PM25 exposure concentrations by randomly selecting one of
the seven years assigning the associated seasonal mean concentrationsas computed using that years data to each census block and then re-
peating the process 1999 times For each of the 160 census blocks the
result was an estimated mean value and standard deviation of the sea-
sonal 24-hour-average concentration of PM25 attributable to primary
emissions from traf 1047297c along the roadway corridor under each scenario
Within each census block and for each scenario the seasonal average
traf 1047297c-related 24-hour PM25 concentration then was represented as a
normaldistribution(left-truncatedat zero) with themean andstandard
deviation estimated from the corresponding 2000 bootstrap simula-
tions The TIA estimated hourly traf 1047297c counts for each scenario along
each roadway link only for weekdays we assumed traf 1047297c counts on
weekends would be the same and hence may have slightly over-
estimated exposure concentrations
In addition to considering variability in PM25exposures arising fromprimary traf 1047297c emissions we assessed the effects of uncertainty in the
accuracy of the air quality model predictions Our previous research
on the integrated air quality modeling approach as well as previous
work by others suggests that the combined MOVESndashCAL3QHCR
model generally predicts PM25 concentration within a factor of two of
measured concentrations (although accuracy varies with local condi-
tions and the quality of data available to support the model) ( Chart-
asa etal2013 Yura etal2007) FollowingMorganand Henrionsguid-
ance (Morgan et al 1990) we represented model uncertainty with an
uncertainty factor (UF ) parameterized by a triangular probability distri-
bution with lower limit = 05 upper limit = 20 and mode = 10
(spanning the expected factor-of-two uncertainty in the model) Ac-
cording to Morgan and Henrion the triangular distribution is especially
appropriate for situations in which ldquothe distributions of variables in a
Fig 2 Overview of framework for incorporating variability and uncertainty into assessment of the health impacts of traf 1047297c-related PM25 The rectangles show sources of variability and
uncertainty The shaded diamonds show the three major information categories needed for quantitative health impact assessment
413C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 613
model arenot preciselyknownrdquo but in which ldquovaluestowardthe middle
of the range of possible values are considered more likely to occur than
values near either extremerdquo Based on previous evaluations of the per-
formance of near-roadway air pollutant dispersion models the exact
form of the distribution representing model uncertainty is not known
making the triangular distribution an appropriate choice for character-
izing model uncertainty Correspondingly in each census block the ex-
cess PM25 24-hour average exposure concentration attributable to
primary emissions from traf 1047297
c on the case study roadway was estimat-ed for each season as
PM exposureimfrac14 U F PM modelim
eth1THORN
where PM exposureimrepresents the average 24-hour PM25 concentration
in census block i (i = 1ndash160) and season m (m = winter spring sum-
mer fall) attributable to primary traf 1047297c emissions on the case study
roadway UF is the model uncertainty factor and PM modelimis thecorre-
sponding model-predicted seasonal daily average PM25 concentration
arising from primary emissions from traf 1047297c
321 Concentrationndashresponse functions
As recommended by the World Health Organization and others
(Ostro and Chestnut 1998 Li et al 2010 Aunan 1996) we use
the following relationship to describe the link between seasonal
daily average PM25 concentrations and the relative risk of cardio-
vascular and respiratory health outcomes
RRimn frac14 e β mn P M exposureim eth2THORN
where β mn is the concentrationndashresponse coef 1047297cient describing the
effects of PM on health outcome n during season m and RRimn is the
relative risk of health outcome n during season m in census block i
The number of adverse health cases in the population attributable
to traf 1047297c-related PM25 then can be determined from the following
relationship
Δ yi jklmn frac14 y0i jklmn A F i jklmn 3a
frac14 y0i jklmn
RRimnminus1
RRimn
3b
frac14 y0i jklmn
e β mn P M exposureimminus1
e β mn P M exposureim
3c
frac14 y0i jklmn 1minuse
minus β mn P M exposureim
3d
where AF i jklmn and Δ yi jklmn are the fraction and number of casesof adverse health event n attributable to traf 1047297c-related PM25 in sea-
son m in census block i for age group j gender k and race l and
where yi jklmn0 is the observed total number of cases in the same lo-
cation and among the same population group Eqs (2) (3a) (3b)
(3c) and (3d) are the standard equations used in analyses by the
WHO and other organizations to attribute observed cases of ad-
verse health events to speci1047297c risk factors ( Ostro and Chestnut
1998 Murray et al 2003 Mathers et al 2001 Pruumlss-uumlstuumln et al
2003)
The β values in Eqs (2) (3c) and (3d) (known as dosendashresponse co-
ef 1047297cients) were drawn from the US Environmental Protection Agency
guidance document Quantitative Health Risk Assessment for Particulate
Matter (US Environmental Protection Agency 2010 Zanobetti and
Schwartz 2009 Bell et al 2008) Table 3 shows the coef 1047297cient values
used in this analysis EPA retrieved these coef 1047297cients from peer-
reviewed epidemiologic studies that met certain quality-assurance
criteria including for example the estimation of exposure from mea-
sured rather than modeled PM25 data For mortality effects the coef 1047297-
cients are speci1047297c to 15 US metropolitan areas For morbidity effects
coef 1047297cients are speci1047297c to region (Northeast Southeast Northwest
and Southwest) This study employed mortality coef 1047297cients developed
from studies in Atlanta since Atlanta is climatologically the most similar
to Chapel Hill among the 15 cities studied We used morbiditycoef 1047297cients for the Southeast region in which Chapel Hill is located
All concentrationndashresponse coef 1047297cients were represented as normal
distributions with all negative valuestruncated at zero (to avoid associ-
ating PM exposure with positive health effects) Standard deviationsfor
each season and health outcome were estimated from the con1047297dence
intervals in Table 3
33 Baseline incidence rates of adverse health outcomes
Data on baseline incidence rates of health outcomes were obtained
from North Carolina public health databases Annual mortality rates
for each age group (Table 4) were calculated by dividing thetotal num-
ber of deaths in Orange County (where Chapel Hill is located) in 2010
(North Carolina State Center for Health Statistics 2012) bythe 2010Or-
ange County census population (Minnesota Population Center 2011)
Annual unscheduled hospital admission rates (Table 5) were obtained
from 2009 emergency department visit data reported by the North Car-
olina Disease Event Tracking and Epidemiologic Collection Tool (NC DE-
TECT) (University of North Carolina at Chapel Hill 2011) We were
unable to obtain data on incidence rates by gender and race so we as-
sume that incidence rates are the same for both genders and all races
(which is a limitation of this analysis) It should be noted as well that
the ICD codes speci1047297c to the concentrationndashresponse coef 1047297cients
might not be entirely matched to the ICD codes speci1047297c to the incidence
rates used in this study depending on reported data Moreover emer-
gency department visits may not result in hospital admissions and
some hospital admissions may occur without 1047297rst visiting the emergen-
cy department
Tore1047298ect seasonal variation we adjusted the incidence rates for car-diovascular and respiratory mortality and unscheduled hospital admis-
sions using data on temporal variability in cardiovascular and
respiratory deaths in Orange County during 1999ndash2010 from the CDC
WONDER database (Centers for Disease Control and Prevention
2013) The fractions for cardiovascular events are 025 031 020 and
024 for winter spring summer and fall respectively while the frac-
tions for respiratory events are 030 026 021 and 023 for winter
spring summer and fall respectively
To determine the total number of cases in any given season (ie
yi jklmn0 in Eqs (3a) (3b) (3c) and (3d)) we multiplied the given inci-
dence rateby the corresponding size of each demographic group in each
census block
34 Testing the effects of variability and uncertainty on health impact estimates
Five simulations of 2000 iterations each were run using Analytica
version 45 (Lumina Decision Systems Los Gatos California) to demon-
strate differences in health burden estimates when including variability
and uncertainty Table 6 lists the1047297ve simulations and the variability and
uncertainty considered in each The1047297rst simulation (1a) follows the de-
terministic approach of previous HIAs usingaverage traf 1047297c volumes and
a constant traf 1047297c emission factor corresponding to traf 1047297c cruising at
35 mph on a 1047298at roadway under a constant ambient temperature of
70 degF Like previous HIAs simulation 1a accounts for neither uncertainty
in the concentrationndashresponse coef 1047297cient (usingthe meanvalue asa de-
terministic estimate) nor seasonal variability Simulation 1b is identical
to simulation 1a except that it uses seasonal concentrationndashresponse
414 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 713
coef 1047297cients (also deterministic) Simulations 2ndash4 systematically include
(one at a time) variability in vehicle emissions rates (simulations 2ndash4)
uncertainty in concentrationndashresponse coef 1047297cients (simulations 3ndash4)
and air quality model prediction error (simulation 4)
35 Comparing health impacts under alternative scenarios
As noted previously we simulated health impacts for the full study
corridor (160 census blocks) for three different scenarios (Table 2)
year 2009 and year 2025 with and without the new campus For each
scenario 2000 simulations were run in Analytica Traf 1047297c patterns (traf-
1047297c volumes along each roadway link in the corridor) for each scenario
were taken from a previous traf 1047297c impact analysis conducted for theTownof Chapel Hill(Vanasse Hangen Brustlin Inc2009)The 2025cen-
sus block populations were obtained from forecasts by the North
Carolina Capital Area Metropolitan Planning Organization (2005)
These growth rates account for demographic changes expected to
occur if the Carolina North campus is built
4 Results and discussion
This analysis explored the effects of variability and uncertainty on
health impact estimates of near-roadway air pollution arising from traf-1047297c attracted by newsuburban development projects Most previous US
HIAs of such projects have provided qualitative rather than quantitative
assessments of health impacts the few quantitative HIAs have not
systematically represented variability and uncertainty in the variables
used to estimate health impacts or in the resulting health outcome pre-
dictions We explored whether including variability and uncertainty
makes a difference in centralestimates of health impacts and we exam-
ined the magnitude of uncertainty in the resulting estimates We then
employed an approach that accounts for variability and uncertainty to
model the expected health impacts in the year 2025 of new traf 1047297c gen-
erated by a new research campus development along a busy roadway
corridor in Chapel Hill North Carolina
41 Effect of including variability and uncertainty
Our results suggest that the conventional deterministic HIA ap-
proach may systematically under-estimate potential health impacts of
traf 1047297c-related PM25 exposure (Fig 3)
Incorporating traf 1047297c emission variability into the analysis (as in sim-
ulation 2) caused the mean value of estimated health impacts to in-
crease by more than a factor of two compared to estimates that
exclude such variability (simulation 1b) This increase occurred because
neglecting the effects on vehicle emissions of variability in temperature
road grade vehicle speed and traf 1047297c behavior (idling accelerating de-
celerating or cruising) resulted in under-estimates of PM25 exposure
Table 3
Concentrationndashresponse coef 1047297cients used in this study
Health outcome Disease category ICD-9 or ICD-10 codea Age group Season Mean concentration-response
coef 1047297cient (95 CI) per 10 μ gm3b
Mortality Cardiovascular I01ndashI59 All ages All-yearc 066 (minus066 198)
Winter 135 (minus193 462)
Spring 076 (minus273 425)
Summer 062 (minus222 347)
Fall minus018 (minus293 257)
Respiratory J00ndash J99 All ages All-yearc
121 (minus048 290)Winter 093 (minus144 329)
Spring 035 (minus205 275)
Summer 077 (minus155 310)
Fall 096 (minus134 325)
Unscheduled hospital admissions Cardiovascular 410ndash414 426ndash429 430ndash438 and 440ndash4 49 6 5 a nd over A ll-yea rc 029 (minus019 077)
Winter 105 (minus007 219)
Spring 075 (minus026 176)
Summer minus067 (minus161 026)
Fall 017 (minus072 106)
Respiratory 464ndash466 480ndash487 and 490ndash492 65 and over All-yearc 035 (minus044 113)
Winter 040 (minus146 224)
Spring 075 (minus082 231)
Summer minus052 (minus209 105)
Fall 014 (minus130 158)
a ICD-10 for mortality and ICD-9 for unscheduled hospital admissionsb Coef 1047297cients were originally from Zanobetti and Schwartz (2009) and Bell et al (2008) respectively
c Used only in simulation 1
Table 4Annual mortality rates by race gender and age group for Orange County (per 1000 people)
Cause of death ICD-10 code Age group Race and gender
White male Black male Other male White female Black female Other female
Cardiovascular disease I05ndashI09 I10ndashI15 I20ndashI25 I26ndashI28 and I30ndashI52 0 to 34 000 000 000 000 000 000
35 to 44 017 000 000 000 095 000
45 to 54 029 272 000 013 153 000
55 to 64 179 235 257 091 105 000
65 to 74 343 725 000 244 000 000
75 to 84 1647 1724 000 770 2065 000
85+ 5251 2632 12500 3034 2299 000
Respiratory disease J00ndash J99 0 to 54 000 000 000 000 000 000
55 to 64 049 000 000 045 105 000
65 to 74 206 483 000 183 345 000
75 to 84 524 575 116 495 590 000
85+ 3580 1316 000 787 1149 000
415C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 613
model arenot preciselyknownrdquo but in which ldquovaluestowardthe middle
of the range of possible values are considered more likely to occur than
values near either extremerdquo Based on previous evaluations of the per-
formance of near-roadway air pollutant dispersion models the exact
form of the distribution representing model uncertainty is not known
making the triangular distribution an appropriate choice for character-
izing model uncertainty Correspondingly in each census block the ex-
cess PM25 24-hour average exposure concentration attributable to
primary emissions from traf 1047297
c on the case study roadway was estimat-ed for each season as
PM exposureimfrac14 U F PM modelim
eth1THORN
where PM exposureimrepresents the average 24-hour PM25 concentration
in census block i (i = 1ndash160) and season m (m = winter spring sum-
mer fall) attributable to primary traf 1047297c emissions on the case study
roadway UF is the model uncertainty factor and PM modelimis thecorre-
sponding model-predicted seasonal daily average PM25 concentration
arising from primary emissions from traf 1047297c
321 Concentrationndashresponse functions
As recommended by the World Health Organization and others
(Ostro and Chestnut 1998 Li et al 2010 Aunan 1996) we use
the following relationship to describe the link between seasonal
daily average PM25 concentrations and the relative risk of cardio-
vascular and respiratory health outcomes
RRimn frac14 e β mn P M exposureim eth2THORN
where β mn is the concentrationndashresponse coef 1047297cient describing the
effects of PM on health outcome n during season m and RRimn is the
relative risk of health outcome n during season m in census block i
The number of adverse health cases in the population attributable
to traf 1047297c-related PM25 then can be determined from the following
relationship
Δ yi jklmn frac14 y0i jklmn A F i jklmn 3a
frac14 y0i jklmn
RRimnminus1
RRimn
3b
frac14 y0i jklmn
e β mn P M exposureimminus1
e β mn P M exposureim
3c
frac14 y0i jklmn 1minuse
minus β mn P M exposureim
3d
where AF i jklmn and Δ yi jklmn are the fraction and number of casesof adverse health event n attributable to traf 1047297c-related PM25 in sea-
son m in census block i for age group j gender k and race l and
where yi jklmn0 is the observed total number of cases in the same lo-
cation and among the same population group Eqs (2) (3a) (3b)
(3c) and (3d) are the standard equations used in analyses by the
WHO and other organizations to attribute observed cases of ad-
verse health events to speci1047297c risk factors ( Ostro and Chestnut
1998 Murray et al 2003 Mathers et al 2001 Pruumlss-uumlstuumln et al
2003)
The β values in Eqs (2) (3c) and (3d) (known as dosendashresponse co-
ef 1047297cients) were drawn from the US Environmental Protection Agency
guidance document Quantitative Health Risk Assessment for Particulate
Matter (US Environmental Protection Agency 2010 Zanobetti and
Schwartz 2009 Bell et al 2008) Table 3 shows the coef 1047297cient values
used in this analysis EPA retrieved these coef 1047297cients from peer-
reviewed epidemiologic studies that met certain quality-assurance
criteria including for example the estimation of exposure from mea-
sured rather than modeled PM25 data For mortality effects the coef 1047297-
cients are speci1047297c to 15 US metropolitan areas For morbidity effects
coef 1047297cients are speci1047297c to region (Northeast Southeast Northwest
and Southwest) This study employed mortality coef 1047297cients developed
from studies in Atlanta since Atlanta is climatologically the most similar
to Chapel Hill among the 15 cities studied We used morbiditycoef 1047297cients for the Southeast region in which Chapel Hill is located
All concentrationndashresponse coef 1047297cients were represented as normal
distributions with all negative valuestruncated at zero (to avoid associ-
ating PM exposure with positive health effects) Standard deviationsfor
each season and health outcome were estimated from the con1047297dence
intervals in Table 3
33 Baseline incidence rates of adverse health outcomes
Data on baseline incidence rates of health outcomes were obtained
from North Carolina public health databases Annual mortality rates
for each age group (Table 4) were calculated by dividing thetotal num-
ber of deaths in Orange County (where Chapel Hill is located) in 2010
(North Carolina State Center for Health Statistics 2012) bythe 2010Or-
ange County census population (Minnesota Population Center 2011)
Annual unscheduled hospital admission rates (Table 5) were obtained
from 2009 emergency department visit data reported by the North Car-
olina Disease Event Tracking and Epidemiologic Collection Tool (NC DE-
TECT) (University of North Carolina at Chapel Hill 2011) We were
unable to obtain data on incidence rates by gender and race so we as-
sume that incidence rates are the same for both genders and all races
(which is a limitation of this analysis) It should be noted as well that
the ICD codes speci1047297c to the concentrationndashresponse coef 1047297cients
might not be entirely matched to the ICD codes speci1047297c to the incidence
rates used in this study depending on reported data Moreover emer-
gency department visits may not result in hospital admissions and
some hospital admissions may occur without 1047297rst visiting the emergen-
cy department
Tore1047298ect seasonal variation we adjusted the incidence rates for car-diovascular and respiratory mortality and unscheduled hospital admis-
sions using data on temporal variability in cardiovascular and
respiratory deaths in Orange County during 1999ndash2010 from the CDC
WONDER database (Centers for Disease Control and Prevention
2013) The fractions for cardiovascular events are 025 031 020 and
024 for winter spring summer and fall respectively while the frac-
tions for respiratory events are 030 026 021 and 023 for winter
spring summer and fall respectively
To determine the total number of cases in any given season (ie
yi jklmn0 in Eqs (3a) (3b) (3c) and (3d)) we multiplied the given inci-
dence rateby the corresponding size of each demographic group in each
census block
34 Testing the effects of variability and uncertainty on health impact estimates
Five simulations of 2000 iterations each were run using Analytica
version 45 (Lumina Decision Systems Los Gatos California) to demon-
strate differences in health burden estimates when including variability
and uncertainty Table 6 lists the1047297ve simulations and the variability and
uncertainty considered in each The1047297rst simulation (1a) follows the de-
terministic approach of previous HIAs usingaverage traf 1047297c volumes and
a constant traf 1047297c emission factor corresponding to traf 1047297c cruising at
35 mph on a 1047298at roadway under a constant ambient temperature of
70 degF Like previous HIAs simulation 1a accounts for neither uncertainty
in the concentrationndashresponse coef 1047297cient (usingthe meanvalue asa de-
terministic estimate) nor seasonal variability Simulation 1b is identical
to simulation 1a except that it uses seasonal concentrationndashresponse
414 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 713
coef 1047297cients (also deterministic) Simulations 2ndash4 systematically include
(one at a time) variability in vehicle emissions rates (simulations 2ndash4)
uncertainty in concentrationndashresponse coef 1047297cients (simulations 3ndash4)
and air quality model prediction error (simulation 4)
35 Comparing health impacts under alternative scenarios
As noted previously we simulated health impacts for the full study
corridor (160 census blocks) for three different scenarios (Table 2)
year 2009 and year 2025 with and without the new campus For each
scenario 2000 simulations were run in Analytica Traf 1047297c patterns (traf-
1047297c volumes along each roadway link in the corridor) for each scenario
were taken from a previous traf 1047297c impact analysis conducted for theTownof Chapel Hill(Vanasse Hangen Brustlin Inc2009)The 2025cen-
sus block populations were obtained from forecasts by the North
Carolina Capital Area Metropolitan Planning Organization (2005)
These growth rates account for demographic changes expected to
occur if the Carolina North campus is built
4 Results and discussion
This analysis explored the effects of variability and uncertainty on
health impact estimates of near-roadway air pollution arising from traf-1047297c attracted by newsuburban development projects Most previous US
HIAs of such projects have provided qualitative rather than quantitative
assessments of health impacts the few quantitative HIAs have not
systematically represented variability and uncertainty in the variables
used to estimate health impacts or in the resulting health outcome pre-
dictions We explored whether including variability and uncertainty
makes a difference in centralestimates of health impacts and we exam-
ined the magnitude of uncertainty in the resulting estimates We then
employed an approach that accounts for variability and uncertainty to
model the expected health impacts in the year 2025 of new traf 1047297c gen-
erated by a new research campus development along a busy roadway
corridor in Chapel Hill North Carolina
41 Effect of including variability and uncertainty
Our results suggest that the conventional deterministic HIA ap-
proach may systematically under-estimate potential health impacts of
traf 1047297c-related PM25 exposure (Fig 3)
Incorporating traf 1047297c emission variability into the analysis (as in sim-
ulation 2) caused the mean value of estimated health impacts to in-
crease by more than a factor of two compared to estimates that
exclude such variability (simulation 1b) This increase occurred because
neglecting the effects on vehicle emissions of variability in temperature
road grade vehicle speed and traf 1047297c behavior (idling accelerating de-
celerating or cruising) resulted in under-estimates of PM25 exposure
Table 3
Concentrationndashresponse coef 1047297cients used in this study
Health outcome Disease category ICD-9 or ICD-10 codea Age group Season Mean concentration-response
coef 1047297cient (95 CI) per 10 μ gm3b
Mortality Cardiovascular I01ndashI59 All ages All-yearc 066 (minus066 198)
Winter 135 (minus193 462)
Spring 076 (minus273 425)
Summer 062 (minus222 347)
Fall minus018 (minus293 257)
Respiratory J00ndash J99 All ages All-yearc
121 (minus048 290)Winter 093 (minus144 329)
Spring 035 (minus205 275)
Summer 077 (minus155 310)
Fall 096 (minus134 325)
Unscheduled hospital admissions Cardiovascular 410ndash414 426ndash429 430ndash438 and 440ndash4 49 6 5 a nd over A ll-yea rc 029 (minus019 077)
Winter 105 (minus007 219)
Spring 075 (minus026 176)
Summer minus067 (minus161 026)
Fall 017 (minus072 106)
Respiratory 464ndash466 480ndash487 and 490ndash492 65 and over All-yearc 035 (minus044 113)
Winter 040 (minus146 224)
Spring 075 (minus082 231)
Summer minus052 (minus209 105)
Fall 014 (minus130 158)
a ICD-10 for mortality and ICD-9 for unscheduled hospital admissionsb Coef 1047297cients were originally from Zanobetti and Schwartz (2009) and Bell et al (2008) respectively
c Used only in simulation 1
Table 4Annual mortality rates by race gender and age group for Orange County (per 1000 people)
Cause of death ICD-10 code Age group Race and gender
White male Black male Other male White female Black female Other female
Cardiovascular disease I05ndashI09 I10ndashI15 I20ndashI25 I26ndashI28 and I30ndashI52 0 to 34 000 000 000 000 000 000
35 to 44 017 000 000 000 095 000
45 to 54 029 272 000 013 153 000
55 to 64 179 235 257 091 105 000
65 to 74 343 725 000 244 000 000
75 to 84 1647 1724 000 770 2065 000
85+ 5251 2632 12500 3034 2299 000
Respiratory disease J00ndash J99 0 to 54 000 000 000 000 000 000
55 to 64 049 000 000 045 105 000
65 to 74 206 483 000 183 345 000
75 to 84 524 575 116 495 590 000
85+ 3580 1316 000 787 1149 000
415C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 713
coef 1047297cients (also deterministic) Simulations 2ndash4 systematically include
(one at a time) variability in vehicle emissions rates (simulations 2ndash4)
uncertainty in concentrationndashresponse coef 1047297cients (simulations 3ndash4)
and air quality model prediction error (simulation 4)
35 Comparing health impacts under alternative scenarios
As noted previously we simulated health impacts for the full study
corridor (160 census blocks) for three different scenarios (Table 2)
year 2009 and year 2025 with and without the new campus For each
scenario 2000 simulations were run in Analytica Traf 1047297c patterns (traf-
1047297c volumes along each roadway link in the corridor) for each scenario
were taken from a previous traf 1047297c impact analysis conducted for theTownof Chapel Hill(Vanasse Hangen Brustlin Inc2009)The 2025cen-
sus block populations were obtained from forecasts by the North
Carolina Capital Area Metropolitan Planning Organization (2005)
These growth rates account for demographic changes expected to
occur if the Carolina North campus is built
4 Results and discussion
This analysis explored the effects of variability and uncertainty on
health impact estimates of near-roadway air pollution arising from traf-1047297c attracted by newsuburban development projects Most previous US
HIAs of such projects have provided qualitative rather than quantitative
assessments of health impacts the few quantitative HIAs have not
systematically represented variability and uncertainty in the variables
used to estimate health impacts or in the resulting health outcome pre-
dictions We explored whether including variability and uncertainty
makes a difference in centralestimates of health impacts and we exam-
ined the magnitude of uncertainty in the resulting estimates We then
employed an approach that accounts for variability and uncertainty to
model the expected health impacts in the year 2025 of new traf 1047297c gen-
erated by a new research campus development along a busy roadway
corridor in Chapel Hill North Carolina
41 Effect of including variability and uncertainty
Our results suggest that the conventional deterministic HIA ap-
proach may systematically under-estimate potential health impacts of
traf 1047297c-related PM25 exposure (Fig 3)
Incorporating traf 1047297c emission variability into the analysis (as in sim-
ulation 2) caused the mean value of estimated health impacts to in-
crease by more than a factor of two compared to estimates that
exclude such variability (simulation 1b) This increase occurred because
neglecting the effects on vehicle emissions of variability in temperature
road grade vehicle speed and traf 1047297c behavior (idling accelerating de-
celerating or cruising) resulted in under-estimates of PM25 exposure
Table 3
Concentrationndashresponse coef 1047297cients used in this study
Health outcome Disease category ICD-9 or ICD-10 codea Age group Season Mean concentration-response
coef 1047297cient (95 CI) per 10 μ gm3b
Mortality Cardiovascular I01ndashI59 All ages All-yearc 066 (minus066 198)
Winter 135 (minus193 462)
Spring 076 (minus273 425)
Summer 062 (minus222 347)
Fall minus018 (minus293 257)
Respiratory J00ndash J99 All ages All-yearc
121 (minus048 290)Winter 093 (minus144 329)
Spring 035 (minus205 275)
Summer 077 (minus155 310)
Fall 096 (minus134 325)
Unscheduled hospital admissions Cardiovascular 410ndash414 426ndash429 430ndash438 and 440ndash4 49 6 5 a nd over A ll-yea rc 029 (minus019 077)
Winter 105 (minus007 219)
Spring 075 (minus026 176)
Summer minus067 (minus161 026)
Fall 017 (minus072 106)
Respiratory 464ndash466 480ndash487 and 490ndash492 65 and over All-yearc 035 (minus044 113)
Winter 040 (minus146 224)
Spring 075 (minus082 231)
Summer minus052 (minus209 105)
Fall 014 (minus130 158)
a ICD-10 for mortality and ICD-9 for unscheduled hospital admissionsb Coef 1047297cients were originally from Zanobetti and Schwartz (2009) and Bell et al (2008) respectively
c Used only in simulation 1
Table 4Annual mortality rates by race gender and age group for Orange County (per 1000 people)
Cause of death ICD-10 code Age group Race and gender
White male Black male Other male White female Black female Other female
Cardiovascular disease I05ndashI09 I10ndashI15 I20ndashI25 I26ndashI28 and I30ndashI52 0 to 34 000 000 000 000 000 000
35 to 44 017 000 000 000 095 000
45 to 54 029 272 000 013 153 000
55 to 64 179 235 257 091 105 000
65 to 74 343 725 000 244 000 000
75 to 84 1647 1724 000 770 2065 000
85+ 5251 2632 12500 3034 2299 000
Respiratory disease J00ndash J99 0 to 54 000 000 000 000 000 000
55 to 64 049 000 000 045 105 000
65 to 74 206 483 000 183 345 000
75 to 84 524 575 116 495 590 000
85+ 3580 1316 000 787 1149 000
415C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 813
concentrations on average For example vehicle emission rates nearly
tripled when the road grade changed from 0 (as assumed in theconventional modeling approach) to 10 (Chart-asa et al 2013) Simi-
larly emissions doubled when the temperature decreased from 70 degF
(the default assumption under the conventional assessment approach)
to10degF (Chart-asa et al 2013) For 1047298at roadways with traf 1047297c moving at
constant speeds in climates with minimal temperature 1047298uctuation var-
iability in emissions factors is expected to be small but for most cases it
is clear that emissions factor variability is an important consideration
when predicting health impacts
When additionally including the uncertainty in concentrationndashre-
sponse coef 1047297cients into the modeling approach (simulation 3) the
mean estimate of health impacts increased still further the estimated
number of attributable CVD deaths and respiratory hospital admissions
more than doubled while respiratory deaths and CVD hospital admis-
sions increased by 69 and 11 respectively This result occurred be-
cause we represented concentrationndashresponse coef 1047297cients as right-
skewed probability distributions (normal distributions left-truncated
at zero) This representation is appropriate because of the constraint
that the coef 1047297cients must be non-negative (since PM25 exposure does
not bene1047297t public health) The result is that the mean value of the con-
centrationndashresponse coef 1047297cients is greater than the median value
which in turn increased the mean estimated health impacts compared
to when such uncertainty was excluded
When additionally including the uncertainty in model prediction
error (simulation 4) the mean estimates increased by another 16ndash17
compared to simulation 3 This result occurred because of the right
skew in the triangular distribution used to represent model uncertainty
and the interactions of this distribution with that used to represent the
concentrationndashresponse coef 1047297cient As previously explained the trian-
gular distribution re1047298ects previous research on the performance of theCAL3QHCR model (Chart-asa et al 2013)
In summary incorporating variability and uncertainty into the
model predictions increased the mean value of estimated health im-
pacts compared to predictions that excluded variability and uncertain-
ty The health impact estimates increased by factors of 7 8 4 and 9 for
CVD deaths CVD hospital admissions respiratory deaths and respirato-
ry hospital admissions respectively The estimates that excluded vari-
ability and uncertainty are biased so low that they are outside the 95
con1047297dence intervals of estimates including variability and uncertainty
These biased predictions could have important implications fordecision-making For example it is possible that excluding variability
and uncertainty and hence producing unrealistically low estimates of
health impacts could result in a decision not to pursue mitigation mea-
sures that would have been determined cost-effective had the full im-
pacts of variability and uncertainty been considered
42 Overall population health impacts at the case study site
This analysis predicted that by 2025 the total number of adverse
health cases attributable to traf 1047297c-related PM25 on the case study road-
way will decreaserelative to 2009 with or without theCarolina North De-
velopment (although the decrease is lower with the development)
(Table 7) This decrease in the number of adverse health outcomes is
predicted to occur despite an expected 20 increase in the population
by 2025 Overall the numberof cases of CVD mortality CVD hospital ad-
missions respiratory mortality and respiratory hospital admissions are
expected to decrease by 42 38 47 and 42 respectively The de-
creased risks arise from the built-in assumptions of MOVES that future
vehicles will be cleaner than todays 1047298eet resulting in traf 1047297c emissions
that decline by about 50 on average compared to todays vehicles
However the increased traf 1047297c associated with the new campus will off-
set even greater decreases in near-roadway PM25 expected to occur in
2025 in the absence of the new campus the number of adverse health
outcomes is expected to be about 30 lower if the new campus is not
built compared to if it is built (results not shown)
The health risks of primary PM25 from traf 1047297c on the case study
roadway vary considerably by season and location (Fig 4) For CVD
mortality effects arehighest in winterdue to the in1047298uencesof high con-
centrationndashresponse coef 1047297cients seasonal incidence variations andtraf 1047297c emission factors during low temperatures The spatial variability
in risk is especially pronounced in winter as illustrated by the grada-
tions by censusblock illustrated in Fig 4 Similar seasonal and spatialef-
fects are observed for the other three health outcomes (not shown)
To investigate the potential factors explaining the spatial distribu-
tion of risk we calculated correlations between several potential ex-
planatory variables and the total excess mortality and morbidity
attributable to PM25 from the roadway in each census block for the
year 2009 Variables included distance from the roadway to the census
block centroid total census block population population over age 64
percent of the population identifying as black and mean PM25 concen-
trationattributable to theroadway acrossall seasons Forexcess mortal-
itythe correlations are largest for mean PM25 concentration (r = 042 t
(158) = 58 p = 14 times 10minus8
) and percentage of the population identi-fying as black (r = 037 t (114) = 42 p =25times10minus5) The correlations
are smaller fordistance to theroadway (r =minus022 t (158) =minus28 p =
00028) total population (r = 015 t (158) = 20 p = 0025) and pop-
ulation over age 64 (r = 016 t (158) = 20 p = 0025) The results are
similar for excess morbidity The spatial distribution in risk arises from
complex interactionsamong a variety of factors including factors affect-
ing population susceptibility (potentially including age and race) and
factors affecting exposure concentration Factors that affect the spatial
distribution of exposure concentrations include not only distance from
the roadway but also roadgrade vehicle typesvehicle speedtraf 1047297c vol-
ume the presence of intersections and wind speed and direction The
effects of such factors are described in detail in Chart-asa et al (2013)
The above-noted correlation between mortality risk associated with
traf 1047297c-related PM25 exposures and the percentage of the census block
Table 5
Annual emergency department visits rates for North Carolina
Cause of Visit ICD 9 code Age group Annual rate
Cardiovascular disease 4275 428 and 5184 (excluding failure due to fumes and vapors) 430ndash435 and 4370ndash4371 65 and over 00856
Respiratory diseas e 466 and 480ndash486 65 and over 00355
Table 6
Sources of uncertainty and variability included in the 1047297ve simulations
Uncertainty and variability sources Simulation
number
1a 1 b 2 3 4
Sources of uncertainty
PM25 exposure concentration
bull Air quality model prediction accuracy x
Dosendashresponse function
bull Dosendashresponse coef 1047297cient x x
Sources of variability
PM25 exposure concentration
bull Vehicle emissions variability on each roadway link arising
from the following sources temperature road grade
cruising speed and percent time spent decelerating idling
accelerating and cruising
x x x
Dosendashresponse function
bull Seasonal variability x x x x
Demographic characteristics of exposed population
bull Age race and gender (by census block) x x x x x
416 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 913
population identifying as black suggests the possible presence of racial
disparities in exposure risks The census block having the highest num-
berof total deathsattributable to traf 1047297c on the study corridor under cur-
rent conditions (block number 371350118002002 with a population of
201) also has a very high percentage of black residents at 47 com-
pared to 9 in the study area as a whole This census block is the
home of a public housing community Airport Gardens intended for
low-income families The block has the second-highest PM25 exposure
concentration among all blocks in the study area Of the 10 census
blocks with the highest number of attributable deaths seven have
higher percentage black populations (17ndash47) than the average for
the study area Nonetheless even in the highest-risk of these census
blocks the annual per-person risk of premature mortality due to
traf 1047297c-related PM25 exposure is vanishingly small 58 times 10minus8
(obtain-ed by dividing the annual attributable deaths by the total population
of the census block) Over a 70-year lifetime this equates to a risk of
41 times 10minus6 Along other busier roadways however the health signi1047297-
cance of such disparities could be much greater
Overall we predict that future risks of primary PM25 from increased
traf 1047297c associatedwith theCarolina North campus will be extremelylow
If the new campus is built then 9 times 10minus6 excess CVD deaths and 2 times
10minus6 excess respiratory deathsare expected compared to if thecampus
isnot built (Table 7) Summingthese two estimates and dividing by the
future study corridor population of 19140 yields a per-person annual
risk of about 6 times 10minus10 These risks are low even if one assumes a resi-
dent is exposed to such a risk level for a lifetime For a 70-year lifetime
the per-person lifetime risk is 4 times 10minus8 Even in the most-exposed cen-
sus block lifetime risks attributable exclusively to the new campus arerelatively low (about 1 times 10minus8 per year or less than one-in-one-
million over a lifetime)
43 Sensitivity and uncertainty analysis
The 95 con1047297dence interval values of the risk estimates in Table 7
range over a factor of about 6ndash7 For example theupper 95 con1047297dence
interval estimate of annual CVD deaths attributable to roadway traf 1047297c
10times 10minus4 isabout 7 times largerthan the lower 95 con1047297dence inter-
val estimate 15 times 10minus5 While from a policy standpoint the risks at
both ends of this con1047297dence interval are relatively low at other sites
the optimal policy decision might change if the actual risk were close
to the upper or lower 95 con1047297dence interval value rather than the cen-
tral estimate Hence in future applications of the HIA analysis approach
demonstrated in this article identifying the variables with the biggest
in1047298uence on the mean value of and uncertainty in the risk estimates
may be important in order to guide additional data collection prior tomaking a risk-informed decision
In a future application a decision-maker may wish to know the ef-
fects of changing each random variable in an HIA model to plausible
high and low values Three key random variables underlie this analysis
the PM25 concentration in each census block as predicted by the com-
bined MOVESCAL3QHCR model the model uncertainty factor
(representing the departure of this combined model from actual PM25
concentrations) and the dosendashresponse coef 1047297cient Fig 5 shows the ef-
fects on thepredicted number of CVD deaths of 1047297xing each of these var-
iables at its lower and upper 95 con1047297dence interval value while
keeping all other variables the same The effects vary by census block
and hence are presented as cumulative distribution functions (CDFs)
For example the dosendashresponse coef 1047297cient relating PM25 exposure
concentration to the risk of CVD mortality in winter is represented inthe base model as a truncated normal distribution with mean 135 times
10minus3 and standard deviation 17 times 10minus3 the lower 95 CI of this
Fig 3 Effect on health impact estimates of including the variability and uncertainty sources shown in Table 6 Error bars represent 95 con1047297dence intervals
Table 7
Comparison of HIA results by development scenario
Scenario Number of census
blocks affecteda
Range of mean exposure
concentrations in affected
blocks (μ gm3)b
Total cases times 106
CVD
mortality
CVD hospital
admissions
Respiratory
mortality
Respiratory hospital
admissions
2009 118ndash148 00002ndash016 48 (15ndash100) 140 (47ndash280) 15 (5ndash30) 73 (21ndash160)
2025 without Carolina North 75ndash122 00002ndash010 19 (56ndash42) 61 (19ndash120) 55 (17ndash12) 30 (8ndash66)
2025 with Carolina North 84ndash137 00002ndash013 28 (79ndash61) 87 (27ndash170) 79 (24ndash17) 42 (12ndash93)
a Number of census blocks with exposure concentrations greater than zero (varies by season)b
Lowest and highest mean seasonal exposure concentration in affected census blocks (also varies by season)
417C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1013
distribution is 12 times 10minus4 and the upper 95 CI is 47 times 10minus3 The
ldquoDosendashResponse Coef 1047297cient Highrdquo curve in Fig 5 shows a CDF of the
risk estimates for census blocks when this coef 1047297cient and those for the
other three seasons are 1047297xed at their upper 95 CI values (in the case
of winter 47 times 10minus3) rather than varying randomly while leaving
the other model variables unchanged Fig 5 shows that for all census
blocks the risk estimates are more sensitive to the concentrationndash
response coef 1047297cient than to the other random variables in the risk
model (air quality model uncertainty factor and predicted PM25 expo-
sure concentration) When the effects of 1047297xing each seasonal dosendash
response coef 1047297cient for CVD mortality at lower or upper 95 CI valuesare summed across all census blocks then the estimated number of
CVD deaths changes from the mean estimate of 47 times 10minus6 to 25 times
10minus6 and 120 times 10minus6 respectively (Fig 6) These results illustrate the
potential importance for futureHIAs of strengthening the epidemiologic
basis for predicting the health effects of PM25 exposures in order to de-
crease the potential for producing risk estimates that are either too high
or too low (Note that results for other health outcomes not shown
here as similar to those illustrated in Figs 5ndash6)
A second question that decision-makers might ask is why the 95
con1047297dence intervals in estimated risks are so wide One approach to an-
swering this question is to examine the rank-order correlation between
the estimated risks and each random variable in the model A high rank-
order correlation between an input variable and the risk estimate indi-
cates that high values of the input variable drive the risk estimate
toward comparably high values For this analysis the rank-order corre-
lations differ by census block season and health outcome Fig 7 shows
CDFs of the rank-order correlations between each random input vari-
able andCVD mortality risks among thecensus blocksby season In win-
ter the season in which PM25 exposure concentrations are highest
uncertainty in the dosendashresponse coef 1047297cient drives uncertainty in the
risk estimates in all census blocks In spring and summer the air quality
model uncertainty factor drives the uncertainty in the risk estimates In
fall the model uncertainty factor drives uncertainty except for in about
20 of census blocks where the dosendashresponse coef 1047297cient contributes
the most uncertainty Hence overall to decrease uncertainty in therisk predictions both the strength of the epidemiologic evidence and
the performance of near-roadway air pollutant dispersion models
must be improved
In summary Figs 5ndash7 illustrate the importance for future
transportation-related HIAs of decreasing uncertainty in epidemiologic
estimates of the concentrationndashresponse coef 1047297cient and improving the
ability to model near-roadway concentrations of PM25 from traf 1047297c
5 Limitations
Key limitations in this analysis arise from de1047297ciencies in the avail-
able epidemiologic evidence the capabilities of the air quality model
and future population data In addition the attributable fraction ap-
proach considers effects of PM25 exposure on the incidence of
Fig 4 Spatial distribution of cardiovascular deaths (times 106) attributable to PM25 before and after Carolina North development
418 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1113
cardiovascular and respiratory deaths but may overlook effects on the
population prevalence of CHD and respiratory diseases
One limitation arises from the assumption that all PM25mdashincluding
that generated by traf 1047297cmdashhas thesame health effects asPM25measured
at population-oriented central site monitors used as the basis for expo-
sure estimates in the epidemiologic studies from which the concentra-
tionndashresponse coef 1047297cients are drawn This assumption is common to
nearly all air quality risk assessments to date (eg Cohen et al 2005
Liet al2010Fann et al 2012) because the understanding of differen-
tial health effects of PM25 from different sources is still emerging Ac-
cording to a meta-analysis by Janssen et al traf 1047297c-associated PM25
may have greater health risks than PM25 from non-combustion sources
( Janssen et al 2011) Janssen et al found that theoretically risk esti-mates using black carbon particles which are associated with combus-
tion from motor vehicle engines and other sources as an indicator of
traf 1047297c-related pollution yielded risk estimates 4ndash9 times higherthan es-
timates using overall PM25 as an indicator However our analysis re-
quired use of PM25 since MOVES and CAL3QHCR do not provide the
capability to estimate black carbon particle concentrations Further-
more the available epidemiologic evidence on the association between
black carbon particlesand health risks is not nearlyas extensiveor thor-
oughly reviewed as that for PM25 ( Janssen et al 2011) Updating near-
roadway dispersion models to predict black carbon particle
concentrations and conducting further epidemiologic studies examin-
ing the effects of vehicle emissions on health are important areas of re-
search Nonetheless forthe case study site theestimated risks would be
very low even assuming the risks are under-estimated by a factor of 9
(the upper bound of Janssen et als predicted under-estimation when
using PM25 rather than black carbon particles as an air pollution indica-
tor) In the baseline scenario (year 2009) the annual average CVD or re-
spiratory mortality risk to an individual from traf 1047297c-related air pollution
predicted by our model is 36 times 10minus9 (=45 times 10minus6 CVD deaths plus
13 times 10minus6 respiratory deaths divided by a population of 16000) As-
suming a 70-year lifetime exposure period the resulting lifetime risk
is 25 times 10minus7 Increasing these risks by a factor of 9 results in an annual
risk of 33 times 10minus8 and a lifetime risks of 23 times 10minus6mdashrisks that are con-sidered very low accordingto US EPA guidelines which in general have
long designated as acceptable risks of less than 10minus4 to 10minus6 (EPA
1989)
A second limitation is that the concentrationndashresponse coef 1047297cients
assume that the exposure histories of current and future residents of
the case study area will be similar to those in the areas from which
the epidemiologic studies were drawn (Atlanta and the southeastern
United States) Once again this limitation is inherent in current airqual-
ity risk assessments due to the costs of conducting epidemiologic stud-
ies and theresulting lack of studies for each US metropolitan area This
limitation may bias the absolute results of the risk estimates but it does
not affect the estimates of risks of one scenario relative to another
Hence the conclusion that the development of the Carolina North cam-
pusis unlikely to lead to substantial traf 1047297c-related air quality health im-pacts is valid even if exposure histories of the Chapel Hill population
differ from those of the populations from which relative risk estimates
were derived
A third limitation is that Eqs (3a) (3b) (3c) and (3d) which have
been used as the basis for assessing health impacts of air pollution
exposure by nearly all researchers to date may neglect the effects of
airpollutionexposureon thedisease progression leading up to hospital-
izations for respiratory illnessesand CVD (Perez et al 2013) Perez et al
recently found that including such effects in analyzing health impacts of
traf 1047297c-related road pollution increased estimated health impacts on av-
erage by a factor of about 10 in a study of 10 major European cities
(Perez et al 2013) However implementing the approach of Perez
et al is not possible when attempting to predict changes in health effect
estimates in the distant future because Perezs calculation relies on
Fig 5 Effects of changingrisk model input variables to their upper andlower95 con1047297dence interval valuesThe cumulativedistribution functions illustrate thevariability in these effects
by census block in the case study roadway corridor
Fig 6 Overall effect (across all census blocks) of changing random variables in the risk
modelto theupperand lowerendsof their95con1047297dence intervals Thechart is centered
on the mean value of theriskestimate 48times 10minus6 Theendsof each barcorrespond tothe
new risk estimate if the variable is changed to its low (left side) or high (right side) 95
con1047297
dence interval value
419C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1213
epidemiologic studies that useproximityto a busy roadwayas theexpo-
sure metric For estimatingthe effect of roadway emissions on coronary
heart disease (CHD) prevalence for example Perez relies on an epide-
miologic study in Germany showing that living within 150 m of a
busy roadway (de1047297ned as an autobahn or federal highway) increased
therelative riskof CHD by 85 compared to not living near such a road-
way Because per-vehicle emissions are expected to decrease substan-
tially in the future such studies cannot be used as the basis for
predicting the effects of road traf 1047297c pollution on populations in the dis-
tant future We expect that future health impacts of living near busy
roads will decrease as vehicle emissions controls improve so including
the effects on disease prevalence also would not change the conclusion
that thefuture risks will be less than todays risks even if thenew cam-pus is built
6 Conclusions
This study developed an improved modeling approach for estimat-
ing the health impacts of traf 1047297c-related PM25 air pollution under alter-
native future urban development scenarios We then demonstrated the
approach by quantifying health impacts in a case study roadway corri-
dor that could be affected by a new UNC campus extension in Chapel
Hill The new approach accounts for the effects of variability in traf 1047297c
emissions factors and for seasonal variabilityin concentrationndashresponse
coef 1047297cients It also accounts for uncertainty in concentrationndashresponse
coef 1047297
cients and air quality model prediction error The approach could
serve as a model for future health impact assessments considering
traf 1047297c-related PM25
Comparisons to the conventional modeling approach used in other
quantitative HIAs revealed that those HIAs could under-estimate poten-
tial health burdens by failing to consider variability and uncertainty in
input variables used to generate the health impact estimates Our anal-
ysis showed that in the case study corridor the conventional approach
under-predicted health impacts by a factor of 4 to 9 depending on the
health endpoint As such the conventional HIA approach could in
some circumstances lead to decisions that are not cost effective andor
are not suf 1047297ciently protective of public health
This analysis also showed that a 1047297ne-scale approach that quanti1047297es
impacts over a small grid (in this case US census blocks) accountingfor demographic variability in each grid cell along with the previously
mentioned variability and uncertainty in model inputs can be useful
for identifying health disparities For example this analysis reveals
that the neighborhood in the study area with the highest health burden
also has a very high minority population compared to that in the rest of
the study area In addition this method of accounting for demographic
variability can be used to analyze differences in risks among age and
gender groups It could be extended to analyze impacts among popula-
tions with pre-existinghealth conditions andoramong groups with dif-
ferent income levels and educational attainment levels as suggested in
a recent analysis of distributional effects of air quality policies by Fann
et al (Fann et al 2011) An analysis of distributional effects would re-
quire data on health outcomes educational attainment and income at
the census block level such data were not available for this analysis
Fig 7 Cumulative distribution functions of rank-order correlations betweenmodelinputvariablesand thepredictedrisk of CVDmortalityby season forthe census blocks in thecase study
roadway corridor A highrank-ordercorrelation indicates that thevariable has a strong in1047298uence on theuncertainty in the estimatedrisk so reducinguncertainty in thevariable will sub-
stantially reduce uncertainty in the estimated risk The cumulative distribution functions show the variability in these effects by spatial location (ie by census block) and season
420 C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421
7232019 1-s20-S0048969714016027-mainpdf
httpslidepdfcomreaderfull1-s20-s0048969714016027-mainpdf 1313
but potentially could be obtained through special requests to thecounty
health department and US Census Bureau
Overall the annual mortality risks of traf 1047297c-related PM25 from the
proposed new campus are very low (less than 1 times 10minus8) even for the
most-exposed populations Nonetheless it is important to recognize
that we consider only onetypeof traf 1047297c-related pollutant and one road-
way corridor Risks would be higher if including all roadways affected
by traf 1047297c from the new campus and all traf 1047297c-related pollutants Fur-
thermore it is important to keep in mindthe manyother sources of am-bient air pollution exposure in the study area and the cumulative effects
of multiple exposures Taking steps to reduce traf 1047297c from the new cam-
pus (eg increasing the frequency of public transit service encouraging
carpooling charging for parking and other steps) will reduce air pollu-
tion exposures and produce bene1047297ts beyond those along the single
roadway considered in this case study
Overall this work highlights the sensitivity of traf 1047297c-related health
impact assessments to uncertainty and variability in concentrationndashre-
sponse coef 1047297cients air quality model prediction accuracy and traf 1047297c
emissions factors Future HIAs should account for these in1047298uential vari-
ability and uncertainty sources
References
Aunan K Exposurendashresponse functions for health effects of air pollutants based on epide-miological 1047297ndings Risk Anal Oct 199616(5)693ndash709
Bell ML Ebisu K Peng RDWalkerJ Samet JM ZegerSL Dominici F Seasonaland regionalshort-term effects of 1047297ne particles on hospital admissions in 202 US counties 1999ndash2005 Am J Epidemiol Dec 2008168(11)1301ndash10
Bhatia R Corburn J Lessons from San Francisco health impact assessments have ad-vanced political conditions for improving population health Health Aff (Millwood)Dec 201130(12)2410ndash8
Bhatia R Seto E Quantitative estimation in health impact assessment opportunities andchallenges Environ Impact Assess Rev Apr 201131(3)301ndash9
Centers for Disease Control and P revention CDC WONDER 2013Chart-asa C Sexton KG Macdonald Gibson J Traf 1047297c impacts on 1047297ne particulate matter air
pollution at the urban project scale a quantitative assessment J Environ Prot (IrvineCalif) December 20134
Cohen AJRoss AndersonH Ostro B Pandey KD Krzyzanowski M Kuumlnzli N GutschmidtKPope A Romieu I Samet JM Smith K The global burden of disease due to outdoor airpollution J Toxicol Environ Health A 200568(13ndash14)1301ndash7
Dannenberg AL Bhatia R Cole BL Heaton SK Feldman JD Rutt CD Use of health impactassessment in the US 27 case studies 1999ndash2007 Am J Prev Med Mar 200834
(3)241ndash56EPA Risk Assessment Guidance for Superfund (part A) 1989 [Washington DC]FannN Lamson AD Anenberg SC Wesson K Risley D HubbellBJ Estimating the national
public health burden associated with exposure to ambientPM25 and ozone RiskAnal Jan 201232(1)81ndash95
Fann N Roman HA Fulcher CM Gentile MA Hubbell BJ Wesson K Levy JI Maximizinghealth bene1047297ts and minimizing inequality incorporating local-scale data in the de-sign and evaluation of air quality policies Risk Anal Jun 201131(6)908ndash22
Federal Highway Administration Economic analysis primer bene1047297tndashcost analysis 2003[Washington DC]
Frey HC Burmaster DE Methods for characterizing variability and uncertainty compari-son of bootstrap simulation and likelihood-based approaches Risk Anal Feb 199919(1)109ndash30
Human Impact Partners Pittsburg Railroad Avenue speci1047297c plan health impact assess-ment 2008 [Oakland CA]
Human Impact Partners Pathways to community health evaluating the healthfulness of affordable housing opportunity sites along the San Pablo Avenue Corridor usinghealth impact assessment 2009 [Oakland CA]
Janssen NAH Hoek G Simic-lawson M Fischer P Van Bree L Brink H Keuken M AtkinsonRW Anderson HR Cassee FR Van Bree L Black carbon as an additional indicator of
the adverse health effects of airborne particles Environ Health Perspect 201119(12)1691ndash9
Li Y Gibson JM Jat P Puggioni G Hasan M West JJ Vizuete W Sexton K Serre M Burdenof disease attributed to anthropogenic air pollution in the United Arab Emirates es-timates based on observed air quality data Sci Total Environ Nov 2010408(23)5784ndash93
Mathers C Vos T Lopez A Salomon JEzzati M National burdenof diseasestudies a prac-tical guide Edition 20 Global program on evidence for health policy 2001 [GenevaSwitzerland]
Minnesota Population Center National Historical Geographic Information System ver-sion 20 Minneapolis MN University of Minnesota 2011
Morgan MG Henrion M Small M Uncertainty a guide to dealing with uncertainty inquantitative risk and policy analysis Cambridge New York Cambridge UniversityPress 1990 p 332
Murray CJ Ezzati M Lopez AD Rodgers A Vander Hoorn S Comparative quanti1047297cation of health risks conceptual framework and methodological issues Popul Health Metr20031(1)1
National Research Council Improving health in the United States the role of health im-pact assessment Washington DC National Academy Press 2011
NCDC Quality Controlled Local Climatological Data (QCLCD) 2013NOAA NOAAESRL Radiosonde Database 2013North Carolina Capital Area Metropolitan Planning Organization Socio-economic demo-
graphic forecasts shape1047297le 2030 Long Range Transportation Plan 2005North Carolina State Center for Health Statistics Detailed mortality statistics 2010 2012Ostro B Outdoor air pollution assessing the environmental burden of disease at national
and local levels no 5 Geneva WHO 2004Ostro B ChestnutL Assessing the health bene1047297ts of reducingparticulate matter air pollu-
tion in the United States Environ Res 199876(2)94ndash106Perez L Declercq C Intildeiguez C Aguilera I Badaloni C Ballester F Bouland C Chanel O
Cirarda FB Forastiere F Forsberg B Haluza D Hedlund B Cambra K Lacasantildea M
Moshammer H Otorepec P Rodriacuteguez-Barranco M Medina S Kuumlnzli N Chronic bur-den of near-roadway traf 1047297c pollution in 10 Europeancities(APHEKOM network) EurRespir J Sep 201342(3)594ndash605
Pruumlss-uumlstuumln A Mathers CD Corvalan C Woodward A Introduction and methodsassessing the environmental burden of disease at national and local levels 2003[Geneva]
Ross CL Elliott ML Rushing MM Barringer J Cox S Frackelton A Kent J Rao AAerotropolis Atlanta Brown1047297eld redevelopment health impact assessment Vol IAtlanta Center for Quality Growth and Regional Development at the GeorgiaInstitute of Technology 2011 p 169
Singleton-Baldrey L The impacts of health impact assessment a review of 54 health im-pact assessments 2007ndash2012 University of North Carolina at Chapel Hill 2012
University of North Carolina at Chapel Hill The UNC Department of Emergency MedicineCarolina Center for Health Informatics report overview and analysis of NC DETECTemergency department data 2009 2011 [Chapel Hill NC]
UC Berkeley Health Impact Group Oak to Ninth Avenue health impact assessment 2006[Berkeley CA]
UC Berkeley Health Impact Group MacArthur BART health impact assessment 2007[Berkeley CA]
UC BerkeleyHealthImpactGroup Health impact assessment of the Portof Oakland 2010[Berkeley CA]
US Environmental Protection Agency Risk Assessment Guidance for Superfund (RAGS)volume III mdash part A process for conducting probabilistic risk assessment 2001[Washington DC]
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
US Environmental Protection Agency Quantitative health risk assessment for particulatematter EPA-452R-10-005 NC Research Triangle Park 2010
Vanasse Hangen BrustlinInc Transportation impact analysis forthe CarolinaNorth devel-opment 2009 [Watertown MA]
Wernham A Health impact assessments are needed in decision making about environ-mental and land-use policy Health Aff (Millwood) May 201130(5)947ndash56
YuraEA Kear T Niemeier D Using CALINE dispersion to assess vehicular PM25 emissionsAtmos Environ Dec 200741(38)8747ndash57
Zanobetti A Schwartz J Theeffectof 1047297ne andcoarseparticulate airpollution on mortalitya national analysis Environ Health Perspect 2009117(6)898ndash903
421C Chart-asa JM Gibson Science of the Total Environment 506 ndash507 (2015) 409ndash421