BenMAR)
Environmental Benefits Mapping
and Analysis Program
User's Manual
December 2022
Updated for BenMAP Version 0.01
-------
TOC
Table of Contents
Chapter 1 1-i
1.1 How to Acesss BenMAP 1-1
1.1.1 New User Registration 1-1
1.2 Home Screen 1-5
1.3 Conducting an Analysis 1-6
1.4 Step 1: Where? 1-7
1.5 Step 2: What pollutant? 1-8
1.6 Step 3: What air quality? 1-8
1.7 Step 4: Who will be exposed? 1-13
1.8 Step 5: What health effects? 1-16
1.9 Step 6: Value of effects? 1-17
1.10 Step 7: Review & Submit 1-19
1.11 View and Export Results 1-20
Chapter 2 2-i
2.1 Introduction to Health Benefits Assessment 2-1
2.1.1 Overview of BenMAP & Benefits Assessment 2-1
2.2 Browser and Computer Requirements 2-7
2.3 Contacts for Comments, Questions, and Bug Reporting 2-8
2.4 Frequently Asked Questions 2-8
Chapter 3 3-i
3.1 Introduction to Estimating Health Incidence Changes 3-1
3.2 Pollutant Change 3-3
BenMAP User Manual v 0.01 December 2022
-------
TOC
3.3 Population 3-4
3.4 Baseline Incidence 3-5
3.5 Health Impact Functions 3-6
3.5.1 Evaluating Sets of Health Effects 3-8
3.5.2 Evaluating Individual Health Effects 3-9
3.5.3 Details on Health Impact Functions 3-9
Chapter 4 4-i
4.1 Overview of Economic Valuation 4-1
4.2 Monetizing Health Benefits 4-2
4.3 Valuing Reductions in Premature Mortality 4-4
4.4 Overview of Discounting 4-5
4.5 Valuing Incidence Results 4-12
4.5.1 How to Specify Valuation Functions in BenMAP 4-13
4.5.2 Details on Valuation Functions 4-14
4.6 Currency Year and Income Growth 4-16
Chapter 5 5-i
5.1 Introduction to Customizing a Benefits Analysis 5-1
5.2 Custom Air Quality Surfaces 5-1
5.2.1 Model Data File Structure 5-2
5.2.2 Loading Custom Air Quality Surfaces 5-4
5.2.3 Validating Custom Air Quality Surfaces 5-6
Chapter 6 6-i
Appendix A. Monitor Rollback Algorithms A-l
BenMAP User Manual v 0.01 December 2022
ii
-------
TOC
Appendix B. Algorithms for Estimating Air Pollution Exposure B-l
Appendix C. Deriving Health Impact Functions C-l
Appendix D. U.S. Health Incidence & Prevalence Data in BenMAP D-l
Appendix E. Core Particulate Matter Health Impact Functions in BenMAP E-l
Appendix F. Core Ozone Health Impact Functions in BenMAP F-l
Appendix G. Additional Health Impact Functions in BenMAP G-l
Appendix H. Core Health Valuation Functions in BenMAP H-l
Appendix I. Additional Health Valuation Functions in BenMAP 1-1
Appendix J. Population & Other Data in BenMAP J-l
Appendix K. Uncertainty & Pooling K-l
Appendix L. Batch Run Approach L-l
Appendix M. Function Editor M-l
References R-l
BenMAP User Manual v 0.01
December 2022
-------
Chapter l
BenMAP
Quick Start Guide
BenMAP User Manual v 0.01
1-i
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
Chapter l Table of Contents
1.1 How to Acesss BenMAP 1-1
1.1.1 New User Registration 1-1
1.2 Home Screen 1-5
1.3 Conducting an Analysis 1-6
1.4 Step 1: Where? 1-7
1.5 Step 2: What pollutant? 1-8
1.6 Step 3: What air quality? 1-8
1.7 Step 4: Who will be exposed? 1-13
1.8 Step 5: What health effects? 1-16
1.9 Step 6: Value of effects? 1-17
1.10 Step 7: Review & Submit 1-19
1.11 View and Export Results 1-20
BenMAP User Manual v 0.01
1-ii
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
This chapter of the User Manual will introduce you to U.S. EPA's web-based BenMAP tool and
guide you through the process of setting up and running an analysis of the health benefits
associated with an improvement in air quality.
l.l How to Acesss BenMAP
Returning Users: If you have already registered, navigate to https://benmap.epa.gov. and
login using your username and password.
New Users: In order to use Benmap, you first need to complete a registration process,
described below.
1.1.1 New User Registration
The registration steps differ for EPA and non-EPA users.
EPA Users
1. Navigate to https://benmap.epa.gov
a. If you are on the EPA network, you will be automatically logged in.
b. If you are accessing BenMAP via the Internet, you will be greeted with a login
page and should use your PIV card to authenticate. If you are an EPA account
user without a PIV card, you should access the tool via VDI desktop or follow
the login.gov route described below.
2. If you are not already a member of the BenMAP Users group, you will see a page
directing you to https://waa.epa.gov to request access.
a. Click Community Access and then Request Community Access.
b. Select the BenMAP Users group.
BenMAP User Manual v 0.01
1-1
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
c. Click Submit at the bottom of the page.
3. Once you receive an email indicating that you have been added to the BenMAP Users
group, return to https://benmap.epa.gov and proceed with Section 1.2 of this Quick Start
Guide.
Non-EPA Users
1. Navigate to https://waa.epa.gov
2. Click the Login button on the login.gov tab.
Select a Login Method
Login.gov
PIV Card
0 LOGIN.GOV
EPA Gateway is using a credential provider to allow
you to sign in to your account safely and securely.
If you do not have an existing Login.gov account, you
will be able to create one before you log in.
LOGIN
BenMAP User Manual v 0.01 December 2022
1-2
-------
Chapter 1 - Ben MAP Quick Start Guide
EPA Production is using Login.gov
to allow you to sign in to your
account safely and securely.
Email address
Password
~ Show password
Sign in
First time using Login.gov?
Create an account
• If you already have a Login.gov account, enter your usernarne and password
and click Login, On the following web page, select the Community Access
menu and select Request Web Community Access.
• If you do not already have a Login.gov account, click Create an Account and
complete the process to configure and activate your account. Once the
account is active, you will be asked to complete the Request Web Community
Access form. If you are not automatically directed to the form, select the
Community Access menu and select Request Web Community Access.
3. Complete all the fields on the Request Web Community Access form (see screenshot
on next page). Your EPA contact is Neal Fann, fann.neal@epa.gov, 919-541-0209. At
the bottom of the page, select the BenMAP Users community from the dropdown list.
BenMAP User Manual v 0.01
1-3
December 2022
-------
Chapter 1 - Ben MAP Quick Start Guide
r\ rnA u™,°d s6""
Ehvirnnm»nlal Protection
1—1 n Agency
Tihank vdu for registering for EPA Web Applicant on Access with your fogin.gov credentials. Please complete this form
to gain access to EFA Web Community or Application.
ALL FIELDS ARE REQUIRED
EPA Contact Name:
EPA Contact's Email Adkiress: e.g. emailld^epa.gov
EPA Contact's Phone Number:
Your Information:
First IVame:
Last Narre:
Email Aa-iress:
Street Address:
City:
Country: v
St3tei'Provinca'P.sg:br:
Postal Code:
Phone Number:
Select (he Community or Application Far which you are requesting access:
Select One v
I accept tne EPA Privacy & Security No-lice Click here to read.
Submit Registration I Reset
BenMAP User Manual v 0.01
1-4
December 2022
-------
Chapter 1 - Ben MAP Quick Start Guide
4, After submitting the form, you will receive one email acknowledging receipt of your
request and then, after a period of time, another notifying you that your request has
been approved.
5. Once approved, proceed to https://benmap.epa.gov, log in using login.gov, and
proceed with Section 1.2 of this Quick Start Guide,
1.2 Home Screen
Across the top of the screen is a black banner with four icons: Home, Data Center, Help, and
Feedback. These features are described below.
SEPA ; -
Envi
iionincntal Topics
Laws & Regulations
About EPA j";?
BenMAP - Benefits Mapping and Analysis Program
Home ; Data Center ? Help Feedback
You can start a new analysis, or select a template to start your analysis
Start a new analysis
g SeJecs a template
NEW ANALYSIS A ANALYSIS FROM TEMPLATE
Home.
Analysis
NEW ANALYSIS
This is the BenMAP welcome screen. To begin a new analysis, click New
. If you have previously conducted analyses in BenMAP, you
option, which allows you
can also choose the Analysis from Template
to re-run or modify a previous analysis.
ANALYSIS FROM TEMPLATE ^
BenMAP User Manual v 0.01
1-5
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
Data Center. HUjiKUiHfl The data center contains records of your past analyses and the
custom file uploads that you created or that have been shared with you by others.
To view your active, pending, or completed tasks, click Manage Tasks
If you have any active or pending tasks, the panels on this screen
will identify your task, its status, and progress. Completed tasks are shown beneath
the list of active/pending tasks.
To review the air quality layers saved in the tool's database and add or delete any
custom datasets, click Review Air Quality Layers BiiiaaaiiMiiiittata&aB. Select a
pollutant from the drop-down menu, and the default air quality surfaces will be
displayed. Clicking on a specific airquality layerwill display the air quality values in
the table in the bottom portion of this screen.
Help. KfljiilThe Help screen links to https://www.epa.gov/benmap/benmap-cloud. a
webpage that includes this user documentation for BenMAP and other more detailed
references supporting the health benefits quantification and valuation approaches employed
in the tool.
Feedback. ¦IHiliiSl Clicking this icon directs you to a form where you can provide
feedback on the BenMAP tool: https://www.epa.gov/benmap/forms/contact-us-about-
benmap.
1.3 Conducting an Analysis
The following steps guide you through a basic BenMAP health impacts analysis. From the
Home screen, click New Analysis to start estimating health benefits. You will be directed to a
new screen with a progress bar at the top that shows all the steps of a BenMAP analysis: steps
BenMAP User Manual v 0.01
1-6
December 2022
-------
Chapter 1 - Ben MAP Quick Start Guide
shown in gray have yet to completed; steps shown in blue with a check mark have been
completed; and a step shown in blue with a pencil icon indicates the current step.
© © o o © o ©
where? whK pollutant' Whilst qwRy? who win be exposed? What neaMh effects? value of effects* Review 4 Submit
1.4 Step i: Where?
The first step is to select where you want to perform your analysis. Ben MAP analyses can be
conducted at a variety of scales, ranging from global to national to city or neighborhood
scale. The selection you make here will establish the geographical boundaries of your
analysis; later you will select the scale at which benefits will be estimated within those
boundaries. Currently, U.S. National is the only option available in this step.
Where do you want to perform your analysis?
SELECT FROM A LIST OF DEFAULTS
® U S National
Q Eastern U.S.
(3 Western u s
|0 Global (everywhere)
CONTINUE
Select U.S. National and click Continue to move on to the next step.
BenMAP User Manual v 0.01
1-7
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
l. 5 Step 2: What pollutant?
In this step, you select the pollutant of concern; this is the pollutant for which BenMAP will
estimate health effects. The tool currently supports health impact calculations for ozone
and PM2.s. Only one pollutant may be analyzed per run. Choosing a pollutant will determine
which air quality surfaces and health impact functions you will be able to select later. The
USEPA website provides more information on particulate matter pollution and ground-level
ozone pollution and associated health effects.
What pollutant do you want to assess?
(3 Ground-level O2one (•) Fine particles (<2.5 ^m)
cfb
Select a pollutant and click Continue to move on to the next step, or Back to edit previous
selections.
1.6 Step 3: What air quality?
To estimate population exposure to air pollution, BenMAP combines air quality data assigned
to a spatial grid (i.e., an "air quality surface") with spatially gridded population data. In this
step, you will select a pair of air quality surfaces for BenMAP to compare. The first of these
surfaces represents pollutant concentrations that represent the state of the world before an
air quality-related policy or action has been implemented (Pre-policy); the second represents
concentrations after implementing the policy or action (Post-policy). BenMAP calculates the
BenMAP User Manual v 0.01
1-8
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
difference in pollutant concentrations between these surfaces. These differences are used to
calculate health impacts. You may select an existing air quality surface from the BenMAP
database or create new surfaces from your own air quality data. BenMAP can generate new
surfaces from .csv files of air quality data (instructions for preparing your own .csvfile are
here). BenMAP currently accepts files of air quality concentrations generated from computer
models that simulate the transport and transformation of pollutants. (Future versions will
also accept air quality data from air quality monitor measurements).
BenMAP assigns air quality data to a spatial structure, where each cell contains an air quality
concentration for your selected pollutant. These grids are characterized either by regularly
shaped and sized cells (like those typically used by air quality models), or by irregular
polygons representing political designations such as counties, provinces, cities, or nations.
BenMAP assumes that the ambient pollutant concentrations assigned to a cell represent the
exposure experienced by people living in that cell. The tool uses these values to estimate
average pollutant concentrations that can be fed into formulas for estimating health impacts.
Currently, the tool accepts airquality inputs assigned to a 12km grid, however future versions
of the model will accept data at the U.S. National, State, or County level as well.
If you want to use a pre-loaded air quality surface:
Select your pre-policy air quality data. BenMAP currently contains 4 pre-loaded ozone
surfaces and 12 pre-loaded PM2.s surfaces. BenMAP will show you the surfaces that
correspond to the pollutant you selected in Step 2. First, select the scenario you would like to
represent the pre-policy (i.e., baseline) conditions of your analysis. Metadata for the selected
scenario are displayed below your selection and show information such as the count of grid
cells and mean pollutant concentration.
BenMAP User Manual v 0.01
1-9
December 2022
-------
Chapter 1 - Ben MAP Quick Start Guide
Step 1. Select your pre-policy air quality data
(•) RCILPM2.5,Annual,Baseiiie
O RCUJPM3 S.AnnuaLFinaJ
M+N4C. S**SOfit! MMr>C
DJdHowMean, QuartertyMean
input lite cftaiaclwiilic Value
Nijnbtf o' yrf c^tll 97416
Mewl value 4 26
2 3th »d 97 5th value C-3SS
Select your post-policy air quality data. Next, select the scenario you would like to
represent the air quality conditions following policy implementation. Metadata for the
selected post-policy scenario are displayed below your selection.
Step 2 Select your post-policy air quality data
0 RCU,PM2 S_AnnuaLB.»setine
(§) RCU.PM2 5_AnnuaLFinai
Metl ¦; S*«on»l
D24HourMean, OuarteftyMean
hivut file characteristic
Number of grid ceils
Mean value
2 Sth and 9? Slovak*
Value
97416
4 28
0-8 5*
BenMAP User Manual v 0.01
1-10
December 2022
-------
Chapter 1 - Ben MAP Quick Start Guide
If you want to use your own data:
To add a custom air quality layer, click the button at the top of the screen that says Add Air
ADD GROUND-LEVEL OZONE AIR QUALITY LAYER
Quality Layer . A window will pop up. In this window, you wil
need to provide a name for your custom air quality layer and select the grid definition it
uses. Then, click the + icon at the top right to upload the air quality surface from your
computer. For details on how to format custom air quality data files so that BenMAP
recognizes them, see Section 5.2 of the user manual.
After you've uploaded your custom air quality layer, BenMAP will validate the surface to
ensure it is formatted correctly. If your air quality layer passes the validation step and is
successfully uploaded, the following pop-up message will appear. Click OK to dismiss.
BenMAP User Manual v 0.01
1-11
December 2022
-------
Chapter 1 - Ben MAP Quick Start Guide
_ Upload your CSV
~~ 221.08/0.00%
AQ#7.CSV
221.08/0.00%
Add Fine partic
Name
Example PM
Your upload of AQ#7.csv was successful
CMAQ 12km Nation - Clipped
UPLOAD I CANCEL
If there is an error with your custom air quality surface, it will not pass the validation step and
an error message will appear. A pop-up with possible errors is shown below.
Your upload of 03_errors.csv has failed
Your file has the following errors
Error
2 records have Column values that are not valid
integers.
Error 2 records have Row values that are not valid integers.
1 record has air quality values that is not a valid
number.
Error 1 record has air quality values below zero.
PRINT ¦ OK
Other potential errors not shown here include using an unavailable pollutant or using an air
quality surface name that's already taken. Click OK and address the listed error(s) in the input
file you created before reuploading.
BenMAP User Manual v 0.01
1-12
December 2022
-------
Chapter 1 - Ben MAP Quick Start Guide
Once you've selected or uploaded yourpre-policy and post-policy air quality scenarios, click
Continue to move on to the next step, or Back to edit previous selections.
1.7 Step 4: Who will be exposed?
BenMAP uses population data to understand how many people are exposed at each modeled
or observed concentration of air pollution and can stratify exposures spatially and
demographically. This information is critical for estimating numbers of avoided premature
deaths or cases of illness associated with a change in air pollution. Population data are
associated with a grid definition, which specifies the geographic areas for which the data is
available (e.g., County or 12-km grid cells). The pre-loaded population data for each grid cell
are stratified by year and by age, race, ethnicity, and gender. The latter information helps you
understand how pollution exposures may be distributed across subpopulations.
You must currently select from pre-loaded U.S. population files in BenMAP. (Future versions
will allow you to import their own population data.)
Step 1 - What population dataset do you want to use'
US CMAQ 12Km Nation
Step 2 m what year will people be exposes)'
2015
Step 3 what baseline health dataset do you want to use'
County, race-stratified *
BenMAP User Manual v 0.01
1-13
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
Population details
Step 1. What population dataset do you want to use? Select your population dataset from
the drop-down menu. Currently, the only option in BenMAP is US CMAQ 12km Nation
population.
Step 2. In what year will people be exposed? Select a year for your population from the
drop-down menu. Currently, the tool supports population years from 2000 through 2055.
Baseline Health Incidence
In the third step on this screen, you will need to select data describing the baseline health
status of your selected population. BenMAP's baseline health datasets describe rates of
disease and death in the U.S. population and include data on both incidence (the rate of new
cases in the population per person pertime) and in some cases prevalence (the fraction of
the population with a specified disease or diagnosis at a given time). Because these data are
meant to capture baseline conditions, these values reflect cases resulting from all stressors or
risk factors, including air pollution. These data are specific to the set of health effects in the
BenMAP database.
BenMAP User Manual v 0.01
1-14
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
Step 3. What baseline health dataset do you want to use? Select a baseline health dataset
from the drop-down menu. Currently, the dataset options are:
County
This dataset contains baseline incidence and prevalence data for fatal and
nonfatal health effects at U.S. county scale. More information on this dataset can
be found here.
County, race-
stratified
This dataset contains baseline mortality incidence data stratified by race at the
U.S. county scale. More information on this dataset can be found here.
County,
ethnicity-
stratified
This dataset contains baseline mortality incidence data stratified by both race
and ethnicity at the U.S. county scale. More information on this dataset can be
found here.
National
This dataset contains baseline incidence and prevalence data for various nonfatal
health effects that are currently only available at the U.S. national level.
Your answer to the question above indicates your preferred data. If baseline health data are
available at the spatial scale you selected for the health effects you choose, BenMAP will use
those data. If baseline health data are not available at the indicated spatial scale, BenMAP
will select the next best available data. For example, baseline data for prevalence of asthma
symptoms is currently available at the national level only. If you select County in this step
and choose to evaluate the change in asthma symptoms associated with a change in air
quality, BenMAP will automatically select National incidence for this health effect. You can
check BenMAP's incidence data selections when you get to Step 6.
Click Continue to move on to the next step, or Back to edit previous selections.
BenMAP User Manual v 0.01
1-15
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
l. 8 Step 5: What health effects?
Health impact functions (HIFs) estimate the change in the number of adverse health
effects associated with a change in exposure to air pollution. The inputs to a health impact
function include: the change in air quality concentration for a pollutant, the size of the
exposed population, the baseline incidence rate of the adverse health effect, and an effect
coefficient derived from epidemiological studies. In the steps above you selected the
pollutant, population and baseline incidence rates. Below you will select the health effects to
estimate using these data.
BenMAP currently allows you to select broad sets of health effects to evaluate, specific to
yourchosen pollutant. (Future versions of the tool will allowyou either to specify individual
health effects or to specify a custom set of both health effects and the functions used to
quantify those effects.)
Estimate a standard set of health effects o
] Premature Death - Primary
~ Premature Death-All
3 Chronic Effects - All
rP
\ | Acute Effects • Primary
I | Acute Effects-All
BACK V CONTINUE
Estimate a standard set of health effects. Check the box next to each set of health effects
you would like to analyze. Available health impact functions are grouped into standard sets
based on the type of health impact and duration of disease.
BenMAP User Manual v 0.01
1-16
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
Premature Death
Primary
Premature Death - All
Chronic Effects - All
Acute Effects - Primary
Acute Effects - All
Contains functions that estimate changes in premature death
attributable to the air pollutant. These are the studies judged by EPA to
be most appropriate to use for its primary benefits results when
conducting major national regulatory impact analyses.
Contains an expanded set of functions for estimating premature
deaths, including functions used in EPA's sensitivity analyses in RIAs,
and functions specific to at-risk groups.
Contains functions that estimate longer-term health conditions, e.g.,
new cases of asthma. These are the studies judged by EPA to be most
appropriate to use for its primary benefits results when conducting
major national regulatory impact analyses.
Contains functions that estimate short-term health effects, e.g.,
respiratory emergency department (ED) visits. These are the studies
judged by EPA to be most appropriate to use for its primary benefits
results when conducting major national regulatory impact analyses.
Contains expanded set of functions that estimate short term health
effects, including functions used in EPA's sensitivity analyses in RIAs,
and functions specific to at-risk groups.
BenMAP will show you the health effects and studies corresponding to your selections above
in the next step. Click Continue to move on to the next step, or Back to edit previous
selections.
1.9 Step 6: Value of effects?
In addition to quantifying population changes in health, BenMAP lets you assign monetary
values to these changes. You can match each health effect with a valuation function
derived from peer-reviewed economic literature and/or extensive health-care databases of
BenMAP User Manual v 0.01
1-17
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
medical costs. Monetizing human health benefits helps express the economic value of
improved health to society and facilitates easier comparisons of a policy's health benefits to
its implementation costs.
After you select a set of health effects, you will be directed to a screen that lists all the
individual health functions you have chosen. Here, you can review the health effect sets,
individual health effects, and the key meta-data for each health impact function.
The valuation step is recommended, but optional. If you would like to value your selected
health effects, click on the pencil icon f " at the right of each row. A pop-up window will
appear with the same metadata for the health impact function of interest, as well as a drop-
down menu you can use to select a valuation function. Open the drop-down menu and scroll
to the valuation function orfunctions that match the health effect and age range of the health
impact function. Click on a valuation function to select it. You may click multiple rows to
select multiple valuation functions. To delete any valuation functions you do not wish to
include, click the next to the function in the pop-up window.
BenMAP User Manual v 0.01
1-18
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
Group
Author i vear
Endpoint
Age Range
Race / Ethnicity t Gendef
incidence (l) v Prevalence (P)
Chrome Effects - AH
Cbanbvand el al /2C16
incidence, Lung Cancer
30*99
ALL I ALL t ALL
National (2021)
incidence, Lung Cancer 55 - 64 ; SEER Age-Distribution Adjust G
153
Select all the valuation functions you wish to run and click Save to exit the pop-up window
and return to the main screen with the full list of health effects/functions. Repeat this process
for all health effects that you wish to value.
Once you have selected all valuation functions, click Continue to review and submit your
analysis, or Back to edit previous selections.
If you choose to skip valuation, leave the Valuation column blank, scroll to the bottom of the
screen, and click Continue to review and submit your analysis.
On the Review and Submit screen, you will be able to review all of your selections for this
analysis: the pollutant, pre- and post-policy air quality surfaces, population dataset, year of
analysis, baseline health dataset, health effect set(s), and number of health impact
functions. Confirm that these are correct or click Back to edit your previous selections.
l.io Step 7: Review & Submit
BenMAP User Manual v 0.01
1-19
December 2022
-------
Review And Submtl
Chapter 1 - Ben MAP Quick Start Guide
Pollutant
Pre-Potify
Post-Policy
PopuJitton Daiaser
Year
incidence
nea)ih Effects
Heaith impart Functions
Fine panicles (<2,5 pm)
RCU_PM2 SLAmHLB&settne
ROJ.PMi.S>nnu»LFtt»t
US CMA012km Nation
2015
County
Cfifon»c effects • Al
37
Ent
-------
Chapter 1 - BenMAP Quick Start Guide
View your active or completed tasks by clicking View your Task via the Review and Submit
window orthrough the Manage Tasks window in the Data Center. Ifyourtaskis still pending,
you can see information on its status in the top portion of Manage Tasks window. Also note
that each analysis that involves both quantification and valuation will be represented by two
tasks: one for quantification and a second for the valuation step. The valuation step will only
begin once the corresponding quantification task is completed.
fcctivr/Pfvidvui T«M
The bottom panel of the Manage Tasks window shows all completed tasks; each completed
task should have a blue drop-down button B under the Action column at the far right. Click
on this drop-down to access a menu that lets you either or View/Export Results to access the
results of the analysis orto Delete a task.
If you click View/Export Results, a screen will load displaying the Task Name at the top left,
along with the date and time of the run. The main panel of this screen displays the results of
your analysis by health effect, age, and study, aggregated to the U.S. National level. This
panel also provides the national, population-weighted average change in air quality, the
number of avoided health effects, the number of exposed people, and the baseline incidence
among the population for each health effect. You can select which columns are displayed by
clicking the Columns drop-down menu and selecting any additional characteristics, such as
BenMAP User Manual v 0.01
1-21
December 2022
-------
Chapter 1 - BenMAP Quick Start Guide
race, ethnicity, and gender, that you would like to display. If all columns cannot be shown on
screen scroll bars will appear to help you navigate.
Although BenMAP displays your results at the national level, your "raw," unaggregated results
have been saved to the BenMAP database, and you can access these data, aggregate your
DOWNLOAD
results to a variety of spatial scales, and export the results as a .csvfile. Click the
button at the top right. A window will pop up with the task name, as well as a list of available
aggregation levels to select. Check the box next to any grid levels you wish to aggregate to -
the options currently supported are CMAQ 12km Nation, CMAQ 12km Nation - Clipped,
County, State, and Nation.
CMAQ 12km Nation
This shapefile grid definition contains grid cells that are roughly 12
kilometers on each side, for use with air quality modeling data.
CMAQ 12km Nation -
Clipped
This shapefile grid definition contains grid cells that are roughly 12
kilometers on each side, clipped to the national boundary of the
contiguous U.S.
County
This shapefile grid definition contains county borders, for use with
county-based population and baseline health data.
State
This shapefile grid definition contains state borders, for use in generating
results aggregated to the state level.
BenMAP User Manual v 0.01
December 2022
1-22
-------
Chapter 1 - BenMAP Quick Start Guide
Nation
This shapefile grid definition contains an outline of the contiguous United
States, defining an overall area of interest.
Export Results
Select Grid Levels
Q CMAQ 12km Nation
[~| CMAQ Hkm Nation - Clipped
Q County
Q Nation
Q Stale
Click Export to download your results. The results file(s), in .csv format, and the task log (a
text file containing all the selections you made to generate your results) will be available in
your browser's download page. The results .csv files can be opened in programs such as
Microsoft Excel and Google Sheets.
BenMAP User Manual v 0.01
1-23
December 2022
-------
Chapter 2
Introduction to Health
Benefits Analysis
In this chapter, find...
• An overview of the tool.
• An overview of health benefits analysis.
• Frequently asked questions.
BenMAP User Manual v 0.01
2-i
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
Chapter 2 Table of Contents
2.1 Introduction to Health Benefits Assessment 2-1
2.1.1 Overview of BenMAP & Benefits Assessment 2-1
2.2 Browser and Computer Requirements 2-7
2.3 Contacts for Comments, Questions, and Bug Reporting 2-8
2.4 Frequently Asked Questions 2-8
BenMAP User Manual v 0.01
2-ii
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
2.1 Introduction to Health Benefits Assessment
The environmental Benefits Mapping and Analysis Program (BenMAP) is a powerful yet easy-
to-use program that estimates the number of cases and economic value of health impacts
resulting from changes in air pollution concentrations. The open-source BenMAP web tool
replaces the desktop version of the program (BenMAP-CE) that the U.S. Environmental
Protection Agency (U.S. EPA) first developed in 2003 to analyze national-scale air quality
policies. Previous analyses include health benefits assessments for the National Ambient Air
Quality Standards (NAAQS) for Particulate Matter (2006,2012) and Ozone (2008,2010) as well
as the Revised Cross-State Air Pollution Rule (2021).
U.S. EPA and its partners designed BenMAP to serve the analytical needs of a range of users,
including scientists, policy analysts, and decision makers. Most users apply the BenMAP tool
to answer one of two types of questions:
1. What are the human health and economic benefits associated with a policy changing air
quality?
2. What is the human health burden attributable to total air pollution levels?
2.1.1 Overview of BenMAP & Benefits Assessment
The BenMAP program estimates the human health impacts and economic value of air quality
changes. That is — BenMAP calculates the human health benefits associated with air quality
changes. Such analyses are a critical component of air quality policy assessments. As such, a
variety of federal, state and local airpollution officials have used BenMAP to inform airquality
management decisions.1
i
For a list of peer-reviewed articles that used the BenMAP tool, see: www.epa.gov/benmap
BenMAP User Manual v 0.01 December 2022
2-1
-------
Chapter 2 - Introduction to Health Benefits Analysis
BenMAP estimates benefits from improvements in human health such as reductions in the
risk of premature death, heart attacks, and other adverse health effects. Non-health benefits
of reducing air pollution (i.e., visibility and ecosystem effects) are not quantified in the
current version of BenMAP. After estimating the changes in the incidence of adverse health
effects, BenMAP calculates the monetary benefits associated with those reductions. We
provide a high-level overview of this process below. Additional details on the health effect
quantification and valuation steps can be found in Chapters 3 and 4, respectively.
How does BenMAP estimate human health effects?
First, BenMAP determines the change in ambient air pollution using user-specified air quality
data. Because BenMAP does not model air quality changes, you must input these data into
BenMAP. Next, BenMAP applies a health impact function or a concentration-response (C-R)
function to pollution concentration changes in order to calculate the corresponding health
effect changes. HIFs are derived from epidemiology studies that calculate effect estimates
which relate a change in pollutant concentration to a health impact, alternatively called a
change in the incidence of health outcomes. Effect estimates are combined with the change
in pollutant concentration, population, and baseline incidence to create a HIF. Equation 2-1
shows a typical functional form in which AY is the change in incidence or health effect, Y0 is
the baseline incidence, (3 is the effect estimate, Aconc is the change in pollutant
concentration, and pop is the exposed population. You can specify their preferred HIFs in
BenMAP. This calculation is done for each location in the area of study. Figure 2-1 illustrates
a potential calculation, showing data over a specific area of study which are combined in an
HIF to estimate a health impact.
Equation 2-1
AY = Yq(1 — e_PAconc) * pop
BenMAP User Manual v 0.01
2-2
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
Figure 2-1. Calculating a change in health outcomes
• Population. The exposed population is the number of people affected by the air
pollution change. The Census Bureau is a good source for this information. In addition,
private companies may collect this information and offer it for sale.
• Pollutant change. The air quality change is the difference between the starting air
pollution level (i.e., the pre-policy scenario) and the air pollution level after some
change, such as a new regulation (i.e., the post-policy scenario).
• Baseline Incidence. The baseline Incidence is the probability that a person will suffer
a specified adverse health effect in a given population over a given period of time.
This value should represent the incidence of health outcomes in the population before
changes in air quality are considered. Baseline incidences and other health data are
typically collected by the government. The World Health Organization is another good
BenMAP User Manual v 0.01
2-3
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
source for baseline incidence data.2 Additionally, the Global Burden of Disease (BGD)
query site can be used to find baseline incidence data.3
• Effect estimate. The effect estimate approximates the percentage change in the risk
of an adverse health effect due to a one unit change in ambient air pollution (typically
1 |ig/m3 or 1 ppm). Epidemiological studies are a good source for effect estimates.
How does BenMAP estimate the economic value of human health effects?
BenMAP calculates the economic value of avoided human health effects by multiplying the
quantity of these effects by an estimate of the economic value per case (see Figure 2-2 as well
as Chapter4 for details):
Economic Value = Health Impact * Statistical Unit Value of Health Impact
Figure 2-2. Estimating the Economic Value of Human Health Effects
• •
• •
• • •
• • V
• •
• •
• •
• • • •
• • • • * •
, • • • • • •
• • • •
• • • • • \
* • • • •
$5,000/admission
100 • $5,000 =
$500,000
An air quality policy
reduces the number of
hospital admissions by
100
The economic value of
each avoided admission
is $5,000 in the year
2010
The economic value is
the number of cases
multiplied by the value
of each admission
There are several different economic valuation metrics that can be used when calculating the
value of the health effect. For example, the value of an avoided premature mortality is
2 The World Health Organization is a good source for international health data, see: http://www.who.int.
3 The GBD query site is available at: https://vizhub.healthdata.org/gbd-results/.
BenMAP User Manual v 0.01
2-4
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
generally calculated using the Value of Statistical Life (VSL) - the average monetary value
that a group of people are willing to pay to slightly reduce the risk of premature death in the
population. For other health effects, the costs of the illness may be the only valuation data
available. Cost of illness (COI) measures may include some or all of the direct medical costs,
indirect medical costs, lost wages, caregiving costs, and other incurred costs due to the
illness. COI measures do not measure economic utility. The BenMAP database includes
several different functions for VSL and valuation functions for other health effects for you to
choose, oryou can use the U.S. EPA's approach for quantifying and valuing air pollution
effects.4
Please note that BenMAP does not have air quality modeling capabilities; you must provide
externally created air quality data.
What types of data can you use within BenMAP?
Table 2-1 lists the 10 types of possible input datasets that you can use within BenMAP and
indicates which types of data are needed to perform certain analyses. These components are
built into the web tool; the current version of the tool also supports user-uploaded air quality
model data. Future versions will support additional custom user data inputs.
4
See https://www.epa.Rov/benmap/benmap-community-edition.
BenMAP User Manual v 0.01 December 2022
2-5
-------
Chapter 2 - Introduction to Health Benefits Analysis
Table 2-1. BenMAP Data Elements
Dataset Type
Required to Estimate Health
Impacts
Required to Quantify
Economic Impacts
Grid Definitions
~
~
Pollutants
~
~
Modeled Air Quality Data
~
~
Incidence/Prevalence Rates
~
~
Population Datasets
~
~
Health Impact Functions
~
~
Variable Datasets*
~
Inflation Datasets
~
Valuation Functions
~
Income Growth Adjustments
~
*This includes data tables of socio-economic and demographic data used to support valuation of health impacts as well as
the assessment of exposure and impacts stratified by indicators of poverty, education, and other factors.
Analysts can use BenMAP to:
• Compare benefits associated with various regulatory programs;
• Characterize the distribution of health impacts among population subgroups;
• Estimate health impacts and economic values of existing air pollution concentrations;
• Estimate the health benefits of alternative ambient air quality standards; and
• Perform sensitivity analyses of health or valuation functions, or of other inputs.
BenMAP User Manual v 0.01
2-6
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
Example Applications
Analysts have used BenMAP to investigate a variety of policy questions such as:
• What is the current health burden from PM2.5 levels in Addis Ababa?
• How large are the economic benefits of reduced maternal exposure to fine particulate
matter?
• What are the future health impacts of wildfire smoke under alternative climate scenarios?
• What are the Environmental Justice implications of alternative air quality strategies in
Detroit, Ml?
• How large are tree and forest effects on air quality and human health?
• What are the health benefits from vehicular pollution control strategies?
2.2 Browser and Computer Requirements
• We recommend using BenMAP with the latest release of any of the following desktop
browsers, with cookies and JavaScript enabled (mobile browsers not supported at
this time):
o Google Chrome
o Firefox
o Safari
o Microsoft Edge
• Microsoft Excel or other spreadsheet program (to read exported .csv files and prepare
airquality input files)
BenMAP User Manual v 0.01
2-7
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
2.3 Contacts for Comments, Questions, and Bug
Reporting
If you have comments and questions, or to report a bug, please send EPA a message at the
BenMAP website: https://www.epa.gov/benmap/forms/contact-us-about-benmap, or email
benmap@epa.gov.
2.4 Frequently Asked Questions
Is BenMAP free? Is there a Terms of Use agreement? Are there any restrictions on using
BenMAP?
BenMAP is free. There is no Terms of Use agreement, though new users must register with
EPA to obtain access. The only restriction is that individuals are limited to storing the output
of 10 BenMAP runs at anytime on the system. You may download results to store locally and
delete them from the web database to free up space to conduct additional runs.
How do I know which version of BenMAP I am using? How do I know if I have the most
current version of BenMAP?
You are automatically using the most up-to-date version of BenMAP at
https://benmap.epa.gov. The current version number is shown in the upper right portion of
the screen, beneath your username. The version number for any given run is also indicated at
the beginning of the Task Log exported with every result download.
Why do I get different results than someone else?
There are many possible reasons why your results might differ from someone else's results.
One good place to start is reviewing the Task Log. With the Task Log you can examine the
assumptions and selections that you have made to generate your results and compare your
BenMAP User Manual v 0.01
2-8
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
selections with those made in another analysis. You can also check and compare BenMAP
version numbers and database numbers, which are reported at the beginning of each Task
Log.
Does BenMAP estimate effects of air pollution that are not related to human health (i.e.,
ecological effects)?
No. BenMAP does not currently have impact functions to estimate otherthan human health
effects. In principle, it would be possible to estimate ecological effects, as BenMAP is
designed to combine different types of geographically variable data. To do so, you would
need to develop and load data and impact functions appropriate to estimating ecological
effects of interest.
Where can I find the source code for BenMAP?
BenMAP is an open-source program and the development team welcomes contributions and
scrutiny from the user community. If you are interested in receiving a current copy of the
source code, see https://github.com/BenMAPCE/BenCloudServer and
https://github.com/BenMAPCE/BenCloudApp.
How does the web tool differ from the desktop BenMAP-CE tool?
The user interface of the web tool is fundamentally different from the desktop BenMAP-CE
program, though both versions are used for the same purpose. Additionally, the desktop
BenMAP-CE program has advanced functionality that has not yet been incorporated into the
web tool but is expected in future versions. Future functionality includes pooling, mapping,
and additional support for custom inputs.
BenMAP User Manual v 0.01
2-9
December 2022
-------
Chapter 2 - Introduction to Health Benefits Analysis
Can I still use the desktop version of BenMAP?
Yes, the desktop tool is still fully functional and available for download at
https://www.epa.gov/benmap. However, it is expected that the desktop tool will eventually
no longer be supported by EPA. Updates will be provided to the BenMAP user community as
this date approaches.
BenMAP User Manual v 0.01
2-10
December 2022
-------
Chapter 3
Estimating Health
Incidence
In this chapter, find...
• A description of how BenMAP estimates the incidence of
health outcomes.
• The source of functions relating air pollutant exposures
to health effects.
• The types of data needed to estimate air quality-related
changes in health incidence.
BenMAP User Manual v 0.01
3-i
December 2022
-------
Chapter3 - Estimating Health Incidence
Chapter 3 Table of Contents
3.1 Introduction to Estimating Health Incidence Changes 3-1
3.2 Pollutant Change 3-3
3.3 Population 3-4
3.4 Baseline Incidence 3-5
3.5 Health Impact Functions 3-6
3.5.1 Evaluating Sets of Health Effects 3-8
3.5.2 Evaluating Individual Health Effects 3-9
3.5.3 Details on Health Impact Functions 3-9
BenMAP User Manual v 0.01
3-ii
December 2022
-------
Chapter3 - Estimating Health Incidence
3.1 Introduction to Estimating Health Incidence Changes
Health incidence refers to the counts of new cases of an adverse health effect in a population
over a specified period of time (e.g., a year). One of BenMAP's primary functions is to estimate
changes in health incidence that result from changes in air pollutant concentrations. In
order to generate these estimates, you will need to create and execute a BenMAP
configuration that specifies all the details needed for an incidence analysis. These include:
• The pollutant you want to evaluate;
• The pollutant concentrations in two scenarios you wish to compare, typically referred
to as the pre-policy and post-policy scenarios;
• The year for the analysis;
• The population dataset for the analysis;
• The health effects and health impact functions to be used; and
• The baseline rate of incidence for those health impacts in the specified population.
After entering this information, you may opt to save your configuration choices as a template
to reuse in the future (see box).
BenMAP gives you flexibility in creating, editing,
and saving configuration data. From the home
screen, you can specify the details of a new run
manually, open an existing template and run it as
is, or open a template that you then modify to
create a new run and/ortemplate. If this is your
first time running BenMAP or you have not saved a
template from previous runs, you will need to
click the New Analysis
button.
Fundamental Concept: Templates ^
A Template includes all the user-specified
data and choices for a BenMAP run. This
includes the specified pollutant, scale of
air quality grids, air quality surfaces,
population data, health impact
functions, and baseline incidence
datasets. Templates may be saved to your
user account priorto runninga
configuration to allow you to easily re-run
or adjust a past analysis without needing
to enter all your previous choices. You can
access your templates from the Home
screen.
BenMAP User Manual v 0.01
3-1
December 2022
-------
Chapter3 - Estimating Health Incidence
To load a configuration you saved previously, click the drop-down menu "Select a template",
which is beside New Analysis on the home screen of BenMAP, select your desired template,
and then click the Analysis From Template button.
SEPA *=&¦
Environmental Topics
BenMAP - Benefits Mapping and Analysis Program
^ Home (£ : Data center Help
[p Feedback
You can start a new analysis, or select a template to start your analysis
Start a new analysis
(?) Select a template
1 NEW ANALYSIS ft jl
[ — *
Health
impact
The steps to calculating health impacts are displayed in Figure 3-1 and described in detail
throughout the rest of this chapter.
Figure 3-1. Steps to Calculating Health Impacts
A Y =Yo (1-e BA PM) *Pop
Pollutant change Population Baseline incidence
Effect ^
^ estimate
BenMAP User Manual v 0.01
3-2
December 2022
-------
Chapter3 - Estimating Health Incidence
3.2 Pollutant Change
The Pre-Policy file contains the air quality concentrations for the conditions that exist either
prior to or without any change in existing policy. The Post-Policy file specifies the air quality
concentrations assuming that some type of policy or change has been implemented. The air
quality files should contain concentration data for the same pollutant, using the same
metrics (e.g., daily 24-hour average, 8-hour max) and the same units, such as micrograms per
cubic meter (|ig/m3) or parts per billion (ppb).
The air quality files must also be mapped at the same spatial scale. If you choose a particular
Grid Type (e.g., County) forthe Pre-Policy file, then the same grid type must be used in the
Post-Policy file.
Fundamental Concept: Pre-Policy and Post-Policy Scenarios
BenMAP requires both a Pre-Policy Scenario air quality surface and Post-Policy Scenario air
quality surface to estimate the effects of a change in air quality (Delta).
• The Pre-Policy Scenario characterizes the air quality levels observed or expected in the
absence of the policy change you are evaluating. It is sometimes referred to as "Business as
Usual." The Pre-Policy scenario is usually considered to be the reference scenario against which
to compare a potential scenario characterized by the implementation of regulations.
• The Post-Policy Scenario in BenMAP is the scenario in which emissions from one or more
source sectors are changed (increased or decreased) from the Pre-Policy scenario. The Post-
Policy scenario usually represents expected air quality levels after a new regulation or set of
regulations has been implemented.
The air quality Delta is the change in air pollution between the Pre-Policy air quality grid and the
Post-Policy air quality grid (Pre-Policy minus Post-Policy). A positive Delta indicates that air
pollution has decreased (i.e., air quality has improved in the post-policy scenario compared to the
pre-policy scenario). A negative Delta indicates that air pollution has increased, and is worse for
human health . BenMAP uses the air quality Delta as the input to the health impact function.
BenMAP User Manual v 0.01
3-3
December 2022
-------
Chapter3 - Estimating Health Incidence
In future versions of the tool, the Pollutant specified in the air quality grids will determine the
suite of Health Impact Functions available for the configuration. Only functions associated
with the specified Pollutant will be available forthe configuration. Furthermore, if only
certain Metrics associated with the pollutant are present in the air quality grids, Health
Impact Functions associated with those Metrics will show a notification that the air quality
surface does not provide the metric specified in the health impact function. Currently, you
may select any suite of Health Impact Functions, regardless of the selected pollutant.
3.3 Population
The population data is used to estimate population exposure and in turn any adverse health
effects associated with a change in air pollution. BenMAP allows you to specify race,
ethnicity, gender, and age of the population, as well as the year of the population estimate.
The basis of U.S. population estimates in
BenMAP are 2010 Census population counts.
Future years are projected using a 2015 source
(Woods and Poole) that estimates population
through 2050. The rate of growth from 2045 to
2050 is then applied to the 2050 to 2055 period
to estimate 2055 population.
Population data loaded into BenMAP must be
associated with a Population Configuration,
which defines the races, ethnicities, genders,
and age ranges present in the data. Race,
ethnicity, and gender are unique text values
representing population subgroups (e.g.,
"Asian", or "Female"). Age ranges are defined
by integer values for starting age and ending
Fundamental Concept: Population
Configuration
BenMAP requires population data in order to
estimate the adverse health effects associated
with a change in air pollution. Population data
may be stratified by age, sex, race, and/or
ethnicity. The population configuration is a
template that specifies the categories into
which your population data are organized -
specifically, the race, ethnicity, gender, and
age group subdivisions present in the
population data. Detailed population data
allow you to more accurately estimate health
impacts by better estimating who is exposed
and better aligning your data with the study
populations of population health studies. It
also allows you to estimate and report
benefits by age group, sex, race and/or
ethnicity (e.g., asthma symptoms in African
American males aged 5-17) that may be useful
to support environmental justice analyses.
BenMAP User Manual v 0.01
3-4
December 2022
-------
Chapter3 - Estimating Health Incidence
age (inclusive), and a unique text value representing the name of the age range. For example,
'OTOl' might be used as a name for the age range defined by a start age of zero and an end
age of one, thus consisting of infants through the first twelve months of life and all one-year-
old infants. The population data should contain counts for all combinations of race, ethnicity,
gender, and age range specified in the associated population configuration.
Population data must also be associated with a grid definition which specifies the
geographic areas for which the data are available. If population data are available for
multiple grid definitions (cities and neighborhoods, for example), you can choose to use
different sets of population data for different analyses.
BenMAP can also estimate populations for grid definitions for which no population data are
available by developing area-weighted crosswalks with grid definitions for which data are
available.
3.4 Baseline Incidence
Fundamental Concepts: Incidence and Prevalence
Incidence is a measure of the total number of new occurrences of an adverse health impact in a
geographic area overtime. The incidence rate is the average number of health effects (e.g.,
respiratory hospital admissions) per person per unit of time, typically a day or a year. The incidence
rate must be expressed at the same time scale as the specified by the health impact function. For
example, a health impact function quantifying day-to-day changes in premature death requires a
daily mortality rate. The baseline incidence rate, also called the background incidence rate, is
the incidence of a given adverse effect due to all causes including air pollution. BenMAP typically
estimates and reports benefits as the change in incidence between the Baseline and Control
scenarios, (e.g., the number of avoided asthma Emergency Department visits).
The prevalence rate is the percentage of individuals in a given population at a given point in time
who are experiencing or have been diagnosed with a given health condition (e.g., the prevalence of
asthmatics among children 0 - 17). It may be required for certain health impact functions, such as
those that focus on asthmatics or other groups for which an existing health condition may make
them particularly vulnerable to the health effect being studied.
BenMAP User Manual v 0.01
3-5
December 2022
-------
Chapter3 - Estimating Health Incidence
Most C-R functions, such as those developed from log-linear or logistic risk models, estimate
the percent change in a health effect associated with a specified pollutant concentration
change. In other words, the absolute effect of air pollution on a specific health effect depends
in part on the rate of occurrence of that effect in the exposed population in the baseline.
Therefore, most of the HIFs in BenMAP require the baseline incidence rates (and in some
cases the prevalence rate) of the adverse health effect as inputs.
The incidence rate can be expressed as the number of health effects per person in the
population per unit of time, and the prevalence rate is the percentage of people that have
been diagnosed with a particular illness or condition at a given point in time. For example, in
2006-2008, the incidence rate for new asthma diagnoses among children was estimated by
researchers to be 1.2 cases per 100 children peryear. A recent estimate of the prevalence rate
of asthma (measuring the percentage of the population that is already asthmatic) is 7.5
percent of the total population.5-6
NOTE: For both incidence and prevalence rates, BenMAP allows you to apply rates that vary
by race, ethnicity, gender, and age group. BenMAP supports multiple sets of incidence and
prevalence rates if the rates differ by year or by grid definition.
3.5 Health Impact Functions
Health impact functions relate a change in the concentration of a pollutant to a change in
the incidence of a health effect (e.g., premature mortality or work-loss days). It is typically
derived from the estimated relationship between the concentration of a pollutant and the
5 Example incidence rate from Winer, RA, Qin, X, Harrington, T, Moorman, J, and Zahran, H. 2012. Asthma incidence
among children and adults: findings from the Behavioral Risk Factor Surveillance System and asthma call-back survey
- United States, 2006-2008. J Asthma 49(1): 16-22.
6 Prevalence rate from 2018 National Health Interview Survey (NHIS) data, available at
https://www.cdc.gov/asthma/nhis/2018/table4-l.htm
BenMAP User Manual v 0.01
3-6
December 2022
-------
Chapter3 - Estimating Health Incidence
Fundamental Concepts: Health
Impact Function and Concentration-
Response Function
adverse health effects suffered by a given
population in an epidemiology study. For
example, the pollutant concentration being
measured may be annual fine particulate matter
(PM2.s), and the population response may be
yearly deaths from all causes. For the purposes of
estimating health benefits, the health impact
function itself describes the relationship between
the change in concentration of the pollutant, and
the corresponding change in the population-
health response. We may want to know, for
example, if the concentration of PM2.5 is reduced
by 10 |ig/m3, how many premature deaths will be
avoided?
To estimate changes in health incidence, the first
step is to calculate the change in pollution
concentrations for a particular policy scenario,
such as an air quality improvement produced by a
set of emissions controls. The concentration
change in a pollutant is the increment between
the pre-policy scenario, which could represent
current conditions or a best estimate of future
conditions based on "business-as-usual," and the
post-policy scenario that reflects the impacts of the expected pollution control actions.7 A
A health impact function calculates the
change in the number of adverse health
effects AE associated with a change in air
quality AQ. The inputs to a health impact
function include the change in air quality
concentration for a pollutant (using a
specified metric such as annual
24HourMean); the size of the affected
population (of specified age, race and
ethnicity); the baseline incidence rate of
the adverse health effect; and an effect
coefficient derived from epidemiological
studies.
The coefficient for the health impact
function is known as Beta (li) and is
derived from epidemiological studies. The
value of li typically represents the percent
change in a given adverse health impact
per unit change in pollutant
concentration.
Health impact functions are constructed
using hazard ratios (HR), which estimate
the relationship between the likelihood of
adverse health effects as a function of
concentration of an air pollutant. The
terms C-R function and health impact
function are often used interchangeably.
7 You can also design the scenarios to look retrospectively at past pollution control efforts, where the post-policy
scenario represents current conditions and the pre-policy represents a hypothetical counterfactual scenario that
projects the likely air quality based on historical emissions rates and expected emissions growth in the absence of
those historical strategies and controls.
BenMAP User Manual v 0.01
3-7
December 2022
-------
Chapter3 - Estimating Health Incidence
spatial grid that maps each of these changes, a gridded population dataset, and local data on
current baseline health rates are then fed into health impact functions to generate a map
of the resulting changes in health effects (i.e., health incidence) that can be attributed to the
changes in air pollution. These functions are based on epidemiological studies and are
selected by you. Typically, the positive results from applications of these functions indicate a
decrease in health incidence (e.g., the decrease in asthma, mortality) resulting from a
decrease in air pollution.
3.5.1 Evaluating Sets of Health Effects
BenMAP currently includes a broad range of health impact functions to evaluate, organized
into groups by type of health effect called sets. These sets are described in Table 3-1.
Table 3-1. Description of Health Effect Sets
Premature Death -
Primary
Premature Death - All
Chronic Effects - All
Acute Effects - Primary
Contains functions that estimate changes in premature death
attributable to the air pollutant. These are the studies judged by EPA to
be most appropriate to use for its primary benefits results when
conducting major national regulatory impact analyses.
Contains an expanded set of functions for estimating premature
deaths, including functions used in EPA's sensitivity analyses in RIAs,
and functions specific to at-risk groups.
Contains functions that estimate longer-term health conditions, e.g.,
new cases of asthma. These are the studies judged by EPA to be most
appropriate to use for its primary benefits results when conducting
major national regulatory impact analyses.
Contains functions that estimate short-term health effects, e.g.,
respiratory emergency department (ED) visits. These are the studies
BenMAP User Manual v 0.01
3-8
December 2022
-------
Chapter3 - Estimating Health Incidence
judged by EPA to be most appropriate to use for its primary benefits
results when conducting major national regulatory impact analyses.
Acute Effects - All
Contains expanded set of functions that estimate short term health
effects, including functions used in EPA's sensitivity analyses in RIAs,
and functions specific to at-risk groups.
3.5.2 Evaluating Individual Health Effects
Future versions of the tool will allow you to select individual health effects and health
impact functions to evaluate, and this documentation will be updated.
3.5.3 Details on Health Impact Functions
The complete list of health effects currently available in BenMAP is presented in Table 3-2.
Health effects that are also included in the "Primary" health effect sets are marked with a S.
Table 3-2. Health Effects in BenMAP
Health Effect Set
Pollutant
Primary?
Health Effect
PM2.5
¦/
Mortality, All Cause
Premature Death - All
Ozone
Mortality, All Cause
Ozone
¦/
Mortality, Respiratory
PM2.5
Acute Myocardial Infarction, Nonfatal
PM2.5
Hospital Admissions, Alzheimer's Disease
PM2.5
Hospital Admissions, Parkinson's Disease
PM2.5
Incidence, Out of Hospital Cardiac Arrest
Chronic Effects-All
PM2.5
Incidence, Stroke
PM2.5
Incidence, Asthma
PM2.5
Incidence, Hay Fever/Rhinitis
PM2.5
Incidence, Lung Cancer
Ozone
Incidence, Asthma
Ozone
Incidence, Hay Fever/Rhinitis
BenMAP User Manual v 0.01
3-9
December 2022
-------
Chapter3 - Estimating Health Incidence
Acute Effects-All
Results for Regulatory
Analysis
PM2.5
¦/
Minor Restricted Activity Days
PM2.5
¦/
Asthma Symptoms, Albuterol Use
PM2.5
Emergency Hospital Admissions, All Respiratory
PM2.5
¦/
ER Visits, All Cardiac Outcomes
PM2.5
¦/
ER Visits, Respiratory
PM2.5
ER Visits, Asthma
PM2.5
Hospital Admissions, All Cardiac Outcomes
PM2.5
¦/
Hospital Admissions, Cardio-, Cerebro- and Peripheral Vascular
Disease
PM2.5
¦/
Hospital Admissions, All Respiratory*
PM2.5
Hospital Admissions, Respiratory -1*
PM2.5
¦/
Hospital Admissions, Respiratory -2*
PM2.5
¦/
Work Loss Days
Ozone
¦/
Minor Restricted Activity Days
Ozone
¦/
Asthma Symptoms, Chest Tightness
Ozone
¦/
Asthma Symptoms, Cough
Ozone
¦/
Asthma Symptoms, Shortness of Breath
Ozone
¦/
Asthma Symptoms, Wheeze
Ozone
¦/
ER Visits, Respiratory
Ozone
ER Visits, Asthma
Ozone
¦/
Hospital Admissions, All Respiratory
Ozone
Hospital Admissions, Lower Respiratory Infection
Ozone
¦/
School Loss Days, All Cause
Ozone
Minor Restricted Activity Days
Ozone
Asthma Symptoms, Chest Tightness
Ozone
Asthma Symptoms, Cough
Ozone
Asthma Symptoms, Shortness of Breath
Ozone
Asthma Symptoms, Wheeze
Ozone
ER Visits, Respiratory
Ozone
Hospital Admissions, All Respiratory
Ozone
Incidence, Asthma
Ozone
Incidence, Hay Fever/Rhinitis
Ozone
Mortality, Respiratory
Ozone
School Loss Days, All Cause
* Respiratory hospital admissions health impact functions evaluate different subsets of respiratory effects and are named
accordingly to correspond with associated baseline incidence values. For more information, see section D.2 of Appendix D.
BenMAP User Manual v 0.01
3-10
December 2022
-------
Chapter3 - Estimating Health Incidence
More information about individual health impact functions is displayed in the web tool
under Step 5: What health effects, shown below.
Ctvomc CltcM
Ovwn*
Ml Sulliy m M M I JK> 5
M tuHnotn *1« rtOOS
Ml OwUr» M «< J 3M1
Ml PMfiMil 13X01
Ml INX«r»M«l /J»l
Ml 'SMI
Mi Fop*«*l/20M
Myoo/il i. WvctCfl,
MyotHtJlW IrJtrctiMi NiaV
ACUM M|l9C*nM« M vcton. NcMtMal
A<-*» Myoruilufl im*:*on Wort-.ul (4
0.IMM1I U
Court!* W*s**nflW»
~Jig Cfclfltf. WMfMIQMM
*) (20(41
*»C»U|
«*<»•<(
*Q9Mj
«y fS0t«|
tyaou>
~
/
~
~
~
~
~
~
Additionally, details are reported in the task log when you export results. Further information
on the functions and underlying studies is available in Appendices E and F of this user
manual.
BenMAP User Manual v 0.01
3-11
December 2022
-------
Chapter 4
Valuation and
Discounting
In this chapter, find...
• Introduction to health benefit valuation.
• How to value changes in mortality risks.
• Overview of discounting and income growth.
BenMAP User Manual v 0.01
4-i
December 2022
-------
Chapter4 - Valuation and Discounting
Chapter 4 Table of Contents
4.1 Overview of Economic Valuation 4-1
4.2 Monetizing Health Benefits 4-2
4.3 Valuing Reductions in Premature Mortality 4-4
4.4 Overview of Discounting 4-5
4.5 Valuing Incidence Results 4-12
4.5.1 How to Specify Valuation Functions in BenMAP 4-13
4.5.2 Details on Valuation Functions 4-14
4.6 Currency Year and Income Growth 4-16
BenMAP User Manual v 0.01
4-ii
December 2022
-------
Chapter4 - Valuation and Discounting
4.1 Overview of Economic Valuation
In conducting valuation, the program estimates the economic value of the cases of premature
death and illness described in the previous chapter." In the example below, we discuss how
monetary values for health effects are estimated using U.S. based values. We also provide a
brief introduction to discounting, which weights the value of health benefits depending on
whether they would be realized today or in the future.
Valuation Functions are used by BenMAP to estimate the economic values of changes in the
incidence of health effects. In the context of human health benefits assessment, these functions
help express society's preferences for avoiding certain health effects as an economic value (e.g., in
U.S. dollars).
For morbidity effects, BenMAP estimates monetized benefits using either Willingness to Pay (WTP)
or Cost of Illness (COI)-based valuation functions. WTP is viewed by economists as the most
complete and appropriate measure of the value of a risk reduction and reflects the willingness of
individuals to exchange money for a reduction in his or her risk of illness or death. COI estimates the
value of a health effect based on the observed direct and indirect costs associated with that
condition. Direct costs would include medical costs such as hospital stays and pharmaceutical
costs, while indirect costs include impacts such as lost earnings from days unable to work. ACOI-
based estimate is expected to understate the true economic value of reductions in risk of a health
effect because it often captures only a subset of cost categories, and it does not include the value of
impact categories such as avoided pain and suffering.
For mortality effects, BenMAP generally estimates monetized benefits using the Value of Statistical
Life (VSL), a WTP-based estimate derived from an extensive literature of observed or elicited
estimates of the monetary value that an individual is willing to exchange for small reductions in his
or her risk of death. It does NOT represent the value of the life of any one specified individual.
Improvements in ambient air quality generally lower the risk of developing an adverse health
effect by a fairly small amount at the individual level. When aggregated across a large
population exposed to air pollution, these small risk changes can result in substantial fewer
numbers of premature deaths or of anticipated cases of illness, hospitalizations, emergency
department visits and so forth. For mortality, it is the small risk changes that are being
BenMAP User Manual v 0.01 December 2022
3
4-1
-------
Chapter4 - Valuation and Discounting
valued, based on individuals'WTP (i.e., their willingness to exchange money for lower
mortality risk). For other health effects, we estimate costs on a case-by-case basis, either by
using individuals' measured WTP to avoid a particular effect or using the measurable costs
incurred for a given case of illness as a proxy for the value of avoiding that effect (a COI
approach). These values are incorporated into valuation functions in BenMAP used to
estimate the monetized health benefits of reducing air pollution.
These benefits (reductions in risk) may vary across the population (and could be zero for
some individuals). Likewise, the WTP for a given benefit is likely to vary from one individual to
another. In theory, the total social value associated with the decrease in risk of a given health
problem resulting from a given reduction in pollution concentrations is generally taken to be
the sum of everyone's WTP for the benefits they receive. For the COI approach, costs may
vary from patient to patient, depending on severity of the case and from location to location.
Wages may also vary from location to location. In BenMAP, we apply an estimate of the mean
WTP or COI value per case and use county-level data on wages to estimate indirect impacts
such as lost workdays.
4.2 Monetizing Health Benefits
Epidemiological studies allow us to estimate the number of cases of an adverse health effect
that would be avoided by a given reduction in pollutant concentrations. If we have an
estimate of the average individual's WTP forthe risk reduction conferred upon him, we can
derive an estimate of the value of a statistical case avoided. Suppose, for example, that a
given reduction in pollutant concentrations results in a decrease in mortality risk of 1/10,000.
Then for every 10,000 individuals, one individual would be expected to die in the absence of
the reduction in pollutant concentrations (who would not be expected to die in the presence
of the reduction in pollutant concentrations). If the average individual's WTP forthis 1/10,000
decrease in mortality risk is $1,000, then the VSL is 10,000 x $1,000, or $10 million. The same
type of calculation can produce values for statistical incidences of other health effects.
BenMAP User Manual v 0.01
4-2
December 2022
-------
Chapter4 - Valuation and Discounting
Sometimes those economic values come from stated preference studies in which study
participants are asked about their WTP to avoid a specific adverse health effect. Many other
WTP estimates are derived using revealed preference techniques, which rely on observed
behavior in actual markets to infer individuals' preferences for relevant "non-market" goods
such as changes in mortality risks. For example, wage-risk studies are a form of revealed
preference techniques which rely on labor markets to understand how individuals trade off
risks of death with a monetary value (wages). When estimates of WTP are not available,
economic values can be approximated by other measures, most notably COI measures.
An individual's WTP to avoid an adverse health effect will include the amount of money he or
she would have to pay for medical expenses associated with the illness. Because medical
expenditures are to a significant extent shared by society via medical insurance, Medicare,
etc., the medical expenditures actually incurred by the individual are likely to be less than the
total medical cost to society.
The COI approach attempts to estimate the total value of the medical resources used, the
value of the individual's time lost resulting from the illness, and other costs such as caregiver
time. Because this method does not include the value of avoiding the pain and suffering
resulting from the illness (a potentially large component), it is generally believed to
underestimate the total economic value of avoiding the illness, perhaps substantially.
The contingent valuation method attempts to elicit from people what they would be willing
to pay to avoid the illness. Because of the distortion in the market for medical goods and
services, whereby individuals generally do not pay the full value of medical care, this method
too is likely to understate the total economic value of avoiding the illness.
Although the COI and WTP are the two most common methods, other methods have been
used in certain circumstances. The method with which the benefit analyst chooses to value a
particular health effect will depend in part on what data are available. The unit values
BenMAP User Manual v 0.01
4-3
December 2022
-------
Chapter4 - Valuation and Discounting
currently available for use in BenMAP are data or estimates that have been collected or
generated by researchers and can be readily obtained in publicly available databases or in
the open literature. When reviewing the economic literature to determine the appropriate
valuation functions to use, it is important to have an economist assist.
4.3 Valuing Reductions in Premature Mortality
The economics literature discussing the value of changes in fatality risks is extensive and
provides a basis for monetizing benefits when the number of deaths avoided as a result of an
air quality improvement can be calculated, but the literature on certain issues regarding the
appropriate method for valuing reductions in premature mortality risk is still developing.
Issues such as the appropriate discount rate and whether there are factors, such as age or the
quality of life, that should be taken into consideration when estimating the value of avoided
premature mortality are still under discussion. BenMAP currently offers a variety of options
reflecting the uncertainty surrounding the unit value for premature mortality. See Appendices
H and I for more detail on the valuation functions available in BenMAP.
Monetary estimates of changes in premature mortality risk are often expressed in terms of
the VSL. This term is easily misinterpreted and should be carefully described when used in
benefit analysis. VSL is the aggregate dollar amount that a large group of people would be
willing to payfora reduction in their individual risks ofdyingin a year, such that we would
expect one fewer death among the group during that year on average. The basic assumption
underlying the VSL approach is that equal increments in fatality risks are valued equally. For
similar reasons, the VSL approach is only appropriate for small changes in the risk of death
and should not be used to value more extensive changes. Because changes in individual
fatality risks resulting from environmental regulation are typically very small, the VSL
approach is usually acceptable for these types of benefit analyses.
BenMAP User Manual v 0.01
4-4
December 2022
-------
Chapter4 - Valuation and Discounting
The U.S. EPA National Centerfor Environmental Economics provides answers to frequently
asked questions regarding the economic value of mortality risk on its website:
https://www.epa.gov/environmental-economics/mortality-risk-valuation. You may wish to
consult this site as you have questions regarding how U.S. EPA derives VSL and applies it in an
environmental benefits analysis.
4.4 Overview of Discounting
What is discounting?
In general, people prefer current consumption to future consumption. In other words, one
dollartoday is worth more than one dollar tomorrow, and that dollar continues to decrease
in value as you go further out into the future. (This concept is also referred to as the social
rate of time preference or the time value of
money. This is a different concept than
inflation, which is a general increase in the
price level of goods and services.) Discounting
is the process of converting a future dollar into
a value that can be compared to the value of a
dollar today. The discount rate expresses this
process in quantitative terms. The higherthe
discount rate, the faster value decreases over
time. For example, $1 twenty years from now
is worth $0.55 today at a 3% annual discount
rate, but worth only $0.26 at a 7% annual
discount rate.
Fundamental Concept: Discounting/
Discount Rate
In a cost-benefit analysis, discounting
accounts for the fact that people value
benefits that occur in the future less than
benefits received today. The rate at which
individuals discount the value of those
benefits is the discount rate. Typically, if a
benefit is expected to be realized as a stream
of benefits over multiple years, as is often
assumed for mortality risk reductions, the
economic value of that benefit stream would
be discounted back to the starting year of
analysis and summed as a Net Present Value
(NPV) of benefits.
BenMAP User Manual v 0.01
4-5
December 2022
-------
Chapter4 - Valuation and Discounting
A basic discounting function is as follows (Equation 4-1):
Equation 4-1
Future Value
Present Value =
(1 + r)t
where r is the discount rate and t is the time period (usually in years).
Example: $1 twenty years from now at a 3% annual discount rate is worth $0.55 today
$1.00 11
Present Value = — ^ = = - - - = 0.553676 = $0.55
(1 + 0.03)20 (1.03)20 1.806111
Why do we discount benefits?
The benefits of reductions in air pollution may need to be discounted for three key reasons:
1. Today's society values benefits that occur today more highly than benefits that will
occur in the future. Therefore, we must discount to compare those future benefits
with current benefits.
2. For a cost-benefit analysis, benefit estimates in a future year need to be
comparable to the cost estimates for that same year (which are also discounted).
3. Discounting can be used to compare the future streams of benefits and costs. The
BenMAP program estimates changes in adverse health effects based on changes
in air quality for one specified analysis year, even though certain health benefits
may occur after the analysis year. Discounting can be used to compare the value of
future benefits with the value of benefits occurring during the analysis year.
Under which scenarios would I need to discount benefits?
Discounting of monetary values may be necessary in multiple scenarios, such as the
following:
BenMAP User Manual v 0.01
4-6
December 2022
-------
Chapter4 - Valuation and Discounting
1. The costs and benefits of the rule or policy you are evaluating do not occur in the same
year, and there may be a different temporal pattern for the anticipated streams of
costs and benefits. For example, costs of implementing a rule may be incurred before
emissions begin to be reduced. By discounting the stream of costs and the stream of
benefits back to a common reference year and summing each of them into a single
integrated value (Net Present Value), we can better compare the two values.
2. Certain health effects result in long-term, multi-year medical expenses, lost earnings,
and/or other impacts. The future medical costs and impacts avoided will need to be
discounted to a single integrated value (Net Present Value) to represent the overall
value of that avoided health effect in the year it would have occurred. For example, an
acute myocardial infarction in one year could result in medical costs and lost earnings
for several years.
3. Following a pollution reduction, the resulting benefits may require a number of years
to be fully realized as the overall population health improves; this effect is often
referred to as a cessation lag. The monetized benefits of these future health risk
changes need to be discounted to a single NPV in the year of the pollution change. For
example, exposure to particulate matter in 2025 could result in a diagnosed case of
lung cancer several years later.
When would we not discount benefits?
In other instances, it is not necessary to discount the benefits estimates generated by
BenMAP. This may be true if:
1. The health benefit occurs within the same year as the exposure change; and
2. The health effect and its associated costs are expected to occur in the same
year
3. The exposure change occurs in the same year as the analysis.
BenMAP User Manual v 0.01
4-7
December 2022
-------
Chapter4 - Valuation and Discounting
The first condition may be satisfied by individual incidence changes that occur in close
proximity to the exposure change, such as changes in emergency department visits or inhaler
use, where fluctuations in daily concentrations may result in fewer health effects. The
second condition requires that the health effect results in no costs that extend beyond the
year of impact. Impacts such as emergency department visits and school loss days would fall
in this category, but not effects such as myocardial infarctions (heart attacks), where multiple
years of ongoing treatment and therapy are likely required. It is important to understand the
assumptions within the health and valuation functions before you decide whether you need
to discount. For example, if your analysis year for your benefits estimates does not match the
analysis year for your costs estimates, you may need to discount in order to compare your
benefits with your costs even if you meet the criteria listed above. The third condition
requires that health impacts be realized in the same year as the analysis. Health impacts
costs and benefits due to an exposure in a future yearshould typically be discounted to a
specified year of analysis, even if they occur concurrently with the exposure change.
Which discount rate should I choose?
Selecting a discount rate is challenging and is one of the most contentious methodological
issues encountered in economic analyses of environmental policies. Because environmental
regulations frequently have differing streams of costs and benefits overtime, the selected
discount rate may determine whether the benefits of a regulatory action exceed the costs. In
addition, selecting a higher discount rate may result in a smaller benefits estimate because
the future benefits are worth much less than they would be if a lower discount rate was
selected. For benefits that occur well into the future, the issue of intergenerational equity
further complicates the selection of the discount rate. (In the context of environmental policy,
intergenerational equity refers to the fairness of the distribution of the costs and benefits of a
long-lived policy when those costs and benefits are borne by different generations. Most
criteria pollutants are not considered to have intergenerational equity issues, but the issue
BenMAP User Manual v 0.01
4-8
December 2022
-------
Chapter4 - Valuation and Discounting
frequently arises in analyses of climate-related air pollution impacts and analyses of mercury
exposures.)
There are various economic arguments for and against various discount rates. To comply
with OMB and EPA's recommendations, EPA currently uses discount rates of 3% and 7% for
benefit analyses.
Which health effects accrue medical expenses or lost earnings for multiple years, and
how do I discount them?
BenMAP includes health and valuation functions for several chronic health effects,
including PM2.5-related asthma, and non-fatal acute myocardial infarctions (AMIs, or heart
attacks).
• Asthma is assumed to last from the initial onset of the illness throughout the rest of
the individual's life. BenMAP currently includes two COI functions representing the
two discount rates for asthma.
• Technically, AMIs are discrete, acute events, not chronic conditions. However, heart
attacks cause chronic follow-up health effects that accrue medical expenses over
time, similar to chronic conditions. You can discount the economic value of these
chronic effects through the valuation function in BenMAP. AMIs are assumed to accrue
costs over five years. Although WTP functions for AMIs are not available, BenMAP
currently includes several COI functions that incorporate the direct medical costs and
the opportunity cost (lost earnings) for specific age groups at two discount rates.
• Other health effects with multi-year valuations include Alzheimer's Disease,
Parkinson's Disease, Out of Hospital Cardiac Arrest, and Non-fatal Lung Cancer.
See Appendix H for details on the categories and temporal patterns of costs and the
discounting assumptions, if any, applied within the valuation functions.
BenMAP User Manual v 0.01
4-9
December 2022
-------
Chapter4 - Valuation and Discounting
Should I discount the health incidence as well as the valuation?
You should not discount the health incidence (i.e., the counts of cases avoided) for any of the
scenarios mentioned above. The cessation lag does not change the total number of
premature deaths attributable to a pollution change, but rather the timing of those deaths. If
you discounted the health incidence along with valuation, you would be discounting twice.
Which health benefits may not occur in the same year as exposure?
In many cases, the health benefit from a decrease in exposure to air pollution occurs shortly
after exposure, in a matter of hours or days for example, but there can be more significant
lags between exposure change and the realization of the full health benefit. If exposure and
the health effect do not occur within the same year, it is necessary to discount those benefits
back to the analysis year. The only PM2.5 health functions currently in BenMAP that fall into
this category are PM2.5-related premature mortality and incidence of non-fatal lung cancer.
Discounting the value of these impacts is subject to considerable uncertainty because in
both cases, the structure of how the population risk changes with time is unknown. However,
scientific literature on similar adverse health effects and new intervention studies suggest
that premature mortality, including effects related to lung cancer disease progression,
probably would not occur in the same year as the exposure. (See: Roosli M, Kunzli N, Braun-
Fahrlander C, Egger M. 2005. "Years of life lost attributable to air pollution in Switzerland:
dynamic exposure- response model." International Journal of Epidemiology 34[5]:1029-35.)
VSL valuation functions in BenMAP incorporate cessation lags using both 3% and 7%
discount rates.
EPA's Science Advisory Board recommends future research to support the development of
defensible lag structures and provides a lag structure that could be assumed until additional
research has been completed. Some example lag structures from the 2012 PM RIA are shown
BenMAP User Manual v 0.01
4-10
December 2022
-------
Chapter4 - Valuation and Discounting
in Figures 4-1 and 4-2 below.8 Currently, BenMAP incorporates a 20-year cessation lag by
default for VSL valuation of mortality and a separate cessation lag model for valuing non-
fatal lung cancer that is based on historical rates of incidence by age (see Appendix H for
details). At present, the web tool does not support estimating other lag structures, though
this functionality is expected in future versions of the tool.
Note: Discounting is not necessary for short-term ozone-related premature mortality
because it occurs within the analysis year. However, discounting is necessary for premature
mortality associated with long-term ozone exposure.
Figure 4-1. Graphical representation of assumed lag structures
analyzed in EPA's PM RIA as sensitivity analyses
Assumed Lag Structures for PM2 5 Premature Mortality
g 30%
i
CD
C
3
O
o
o
= 20%
ro
-E
O
a>
3
To
E
a)
0%
~ SAB segmented lag
¦ Alternate Segmented Lag
Exponential decay model
• 5-year distributed lag
1st 2nd 3rd 4th 5th 6th 7th 8th 9th 10th 11th 12th 13th 14th 15th 16th 17th 18th 19th 20th
year year year year year year year year year year year year year year year year year year year year
Year Following Reduction in PM2.5
8 Source: U.S. EPA. 2012. Regulatory Impact Analysis for the Final Revisions to that National Ambient Air Quality
Standards for Particulate Matter. Office of Air Quality Planning and Standards. Research Triangle Park, NC.
https://www3.epa.gov/ttnecasl/regdata/RIAs/finalria.pdf
BenMAP User Manual v 0.01
4-11
December 2022
-------
Chapter4 - Valuation and Discounting
Figure 4-2. Graphical representation of cumulative assumed lag
structures analyzed in EPA's PM RIA as sensitivity analyses
Assumed Lag Structures for PM2 5 Premature Mortality (Cumulative)
Year Following Reduction in PM2.5
4.5 Valuing Incidence Results
When BenMAP runs, it will generate a valuation result for each Valuation Method you select
by running the method's function on the selected incidence results. You do not need to select
a Valuation Method for every incidence result—incidence results without any Valuation
Method will simply be ignored when valuation results are generated.
By default, BenMAP will use the uncertainty estimates built into the valuation functions,
along with the uncertainty estimated for health incidence changes, to generate lower- and
upper-bound estimates of the value of health impacts (specifically, the 2.5th and 97.5th
percentiles of the distribution, representing a 95% confidence interval around the central
result). The web tool does not currently support exporting other percentiles of the
distribution, but this functionality will be included in a future release. To obtain the value of
the health incidence, BenMAP draws samples from both the valuation and health incidence
BenMAP User Manual v 0.01
4-12
December 2022
-------
Chapter4 - Valuation and Discounting
distributions and multiplies each combination of
values together, then sorts these results to
estimate the combined uncertainty contributed
by both incidence and valuation. More
information on this procedure is found in
Appendix K.
4.5.1 How to Specify Valuation
Functions in BenMAP
After you have specified your health effects, you
can choose valuation functions to generate a
monetized estimate of health benefits for each health effect. After selecting your health
effect sets, you will be presented with a screen similar to the one below that lists all the
individual health effects that you have chosen to analyze.
Guidance/Best Practices \/
When selecting Valuation Methods for your
analysis, it is important to match the
valuation function to the health incidence
estimate as closely as possible. For
example, if the health effect is Asthma
Symptoms, Albuterol use, then the valuation
function should correspond specifically to
albuterol use rather than other asthma
symptoms such as cough or wheeze.
For long-term health impacts, the valuation
function may also account for your
preferred time span of analysis and
discount rate.
To value a health effect, click the pencil icon /V> at the far right of its row. A pop-up window
will appear that includes a drop-down menu you can use to select a valuation function. Scroll
to the valuation function or functions that match the health effect and age range of the
health impact function you are valuing. Click on a valuation function to select it. You may
click multiple rows to select multiple valuation functions, each of which will be run to
generate a separate valuation estimate.
BenMAP User Manual v 0.01 December 2022
4-13
-------
Chapter4 - Valuation and Discounting
Once you have selected all the valuation functions you wish to run for a given health effect,
click Save to exit the pop-up window and return to the main screen with the full list of health
effects. Repeat this process for each health effect you wish to value.
For more guidance on how to value health effects, see Section 1.9 of this user manual.
4.5.2 Details on Valuation Functions
The complete list of health effects with valuation functions currently available in BenMAP is
presented in Table 4-1.
BenMAP User Manual v 0.01
4-14
December 2022
-------
Chapter4 - Valuation and Discounting
Table 4-1. Details on Valuation Functions
Health Effect
Age Range
Acute Myocardial Infarction, Nonfatal
0-24; 25-44; 45-54; 55-65; 66-99
Minor Restricted Activity Days
18-99
Asthma Symptoms, Chest Tightness
0-17; 18-99
Asthma Symptoms, Cough
0-17; 18-99
Asthma Symptoms, Shortness of Breath
0-17; 18-99
Asthma Symptoms, Wheeze
0-17; 18-99
Asthma Symptoms, Albuterol Use
0-99
ER Visits, All Cardiac Outcomes
0-99
ER Visits, All Respiratory
0-99
ER Visits, Asthma
0-99
Emergency Hospital Admissions, All Respiratory
65-99
Hospital Admissions, All Cardiac Outcomes
0-99
Hospital Admissions, Cardio-, Cerebro- and Peripheral
Vascular Disease
65-99
Hospital Admissions, All Respiratory
0-18; 65-99
Hospital Admissions, Respiratory-1*
0-99
Hospital Admissions, Respiratory-2*
65-99
Hospital Admissions, Alzheimer's Disease
65-99
Hospital Admissions, Parkinson's Disease
18-99; 65-99
Incidence, Out of Hospital Cardiac Arrest
35-99
Incidence, Stroke
18-99
Incidence, Asthma
0-17; 4-21; 35-99
Incidence, Hay Fever/Rhinitis
0-17
Incidence, Lung Cancer
30-34; 35-44; 45-54; 55-64; 65-74; 75-84; 85-99
Mortality, All Cause
0-99
School Loss days
0-17
Work Loss Days
18-65
* Respiratory hospital admissions valuation functions evaluate different subsets of respiratory effects and are named
accordingly to correspond with associated health impact functions and baseline incidence values. For more information, see
section H.2.1 of Appendix H.
BenMAP User Manual v 0.01
4-15
December 2022
-------
Chapter4 - Valuation and Discounting
4.6 Currency Year and Income Growth
The web tool maintains the Variable Datasets from the desktop tool, which includes
economic variables such as income and poverty data, inflation indices, and income growth
factor tables. The relevant year for inflation and income growth matches the year chosen for
population when you prepare your BenMAP run configuration. You can confirm the years
used for these adjustments in the Task Log.
Inflation Adjustment
The Inflation Adjustment needs to be carefully considered in relation to the valuation
datasetthat you are using. The web tool uses the same inflation factortable as the desktop
version of BenMAP. These indices were derived using data from the Bureau of Labor Statistics
(all goods index, medical cost index, wage index). The default valuation database in the
United States setup has a currency year of 2015, so the inflation dataset has a value of 1 for
the year20i5. The web tool sets the inflation year is equal to the Population Year you
selected in Step 4, orto the latest yearfor which inflation data are available if less than the
population year.
Income Growth Adjustment
Economic studies generally provide evidence that WTP estimates are influenced by the
income of individuals. As income rises overtime, people are expected to allocate more
money towards safety and risk reduction, and thus WTP estimates are likely to increase as
well. The Income Growth Adjustment is designed to take this phenomenon into account,
allowing you to account for income growth between the time when WTP estimates were
calculated and the year of your analysis.
As with the Inflation Adjustment, the Income Growth Adjustment has a close connection to
the valuation estimates. For example, the valuation estimates in the United States setup are
BenMAP User Manual v 0.01 December 2022
4-16
-------
Chapter4 - Valuation and Discounting
assumed to be based on income levels from 1990, so the income growth adjustment database
has a value of 1 for the year 1990.
The web tool automatically incorporates an Income Growth Adjustment, where the income
year is set equal to the Population Year you selected in Step 4, or to the latest year for which
income growth adjustment data are available if less than the population year. Historical
income growth (1990-2016) is from the U.S. Bureau of Economic Analysis (BEA). Future
changes in annual income are based on data presented in the 2020 Annual Energy Outlook.
BenMAP User Manual v 0.01
4-17
December 2022
-------
Chapter 5
Customizing a Benefits
Analysis
In this chapter...
• Learn more about the file structure for data inputs.
• Learn how to import custom datasets.
BenMAP User Manual v 0.01
5-i
December 2022
-------
Chapter 5 - Customizing a Benefits Analysis
Chapter 5 Table of Contents
5.1 Introduction to Customizing a Benefits Analysis 5-1
5.2 Custom Air Quality Surfaces 5-1
5.2.1 Model Data File Structure 5-2
5.2.2 Loading Custom Air Quality Surfaces 5-4
5.2.3 Validating Custom Air Quality Surfaces 5-6
BenMAP User Manual v 0.01
5-ii
December 2022
-------
Chapter 5 - Customizing a Benefits Analysis
5.1 Introduction to Customizing a Benefits Analysis
BenMAP contains numerous datasets that can be used to conduct a complete health benefits
assessment. However, you can also import custom datasets to the tool. Custom datasets
allow you to tailor a given analysis to use parameters specific to his or her scenario and/or
geographic area.
Currently, the web tool supports import of custom air quality surfaces. Future versions of the
tool will support customization via additional custom inputs, such as population datasets,
incidence and prevalence data, health impact functions, and valuation functions.
5.2 Custom Air Quality Surfaces
BenMAP does not include an air quality modeling
component; thus, it relies on user-provided
estimates of air quality based on air quality
models. Future versions of the tool will support
surfaces derived from measured air quality data
as well, such as concentrations based on
networks of air quality models. Any custom air
quality file must be assigned to a grid structure,
where each cell contains data for one or more air
quality metrics (e.g., 24-hour average, 8-hour
maximum) in that location.
The grids are either regularly shaped areas like
those used by air quality models, or irregular shapes, like provinces, local government areas,
cities, or nations. Each grid cell is identified by a Row and Column identifier, as described
below. At present, you must assign your custom air quality files to the 12km grid in BenMAP.
Fundamental Concept: Air Quality
Surface
An air quality surface contains modeled
or monitored air pollution data arranged
spatially in a series of cells; these cells
may be a regular shape (like a 12km by
12km grid) or an irregularshape (like a
county or census tract). BenMAP uses one
air quality surface to represent the pre-
policy scenario and a second surface to
represent the post-policy scenario. These
pre- and post-policy surfaces must use the
same air quality grid. The program
calculates the difference between pre- and
post-policy scenarios as an input to the
health impact functions.
BenMAP User Manual v 0.01
5-1
December 2022
-------
Chapter 5 - Customizing a Benefits Analysis
Future versions will accept custom air quality files at the U.S. National, State, and County
level as well. Future versions will also enable you to first upload a GISshapefile in orderto
define a custom grid to which you may then assign air quality concentrations.
5.2.1 Model Data File Structure
To create air quality surfaces, BenMAP uses a number of inputs, including modeling data. You
may enter your own modeling data, provided that the data are in a format recognized by
BenMAP. BenMAP will accept air quality data saved in the comma separated values (CSV)
format. Each CSV value must represent a single time period (e.g., one-year or a multi-year
average); data for additional time periods must be uploaded in a separate file.
Table 5-1 presents the required elements for these files; variable names must be matched
exactly. Table 5-2 presents a sample data file. Each CSV file may contain data for multiple air
quality metrics; for example, an ozone file may contain measurements at each grid cell for
both 8-hour maximum and 1-hour maximum ozone metrics, subject to the constraints in
Table 5-1.
BenMAP User Manual v 0.01
5-2
December 2022
-------
Chapter 5 - Customizing a Benefits Analysis
Table 5-1. Air Modeling Data File Variables
Variable
Type
Required
Notes
Column
Row
Integer
Integer
Yes
Yes
The column and the row uniquely identify each set of
modeling values and link the modeling data with cells in
a grid definition.
Metric
Text
No
This variable is either blank (signifying that the Values are
Observations, rather than Metric values), or must
reference an already defined Metric (e.g., 24-hour daily
mean) for the appropriate Pollutant.
Seasonal
Metric
Text
No
This variable is either blank (signifying that the Values are
not Seasonal Metric values) or must reference an already
defined Seasonal Metric forthe Metric (e.g., mean of the
1-hour maximum values forthe months of June through
August). Files may contain data for more than one metric,
but the values for each metric/grid cell combination
must appear in its own row.
Annual Metric
Text
Yes
For use with annual values, this variable must be one
of: None, Mean, Median, Max, Min, Sum (e.g., mean of the
1-hour maximum values forthe year)
Values9
Number
Yes
The tool currently accepts annual values, i.e., a single
value forthe year for each grid cell. See Table 5-2 for a
sample air quality file.
9 Future functionality of the tool will support daily values, which must be supplied as a comma-delimited string of
values for the year [e.g., 365 or 366 (leap year) values for daily data]. To input daily values, Annual Metric must be
blank. Missing values must be signified with a period
BenMAP User Manual v 0.01
December 2022
5-3
-------
Chapter 5 - Customizing a Benefits Analysis
Table 5-2. Sample Air Modeling Data File
Column
Row
Metric
Seasonal Metric
Statistic
Values
24
101
D24HourMean
QuarterlyMean
Mean
17.53
24
102
D24HourMeari
QuarterlyMean
Mean
17.15
24
103
D24HourMean
QuarterlyMean
Mean
20.00
24
104
D24HourMean
QuarterlyMean
Mean
17.07
24
105
D24HourMean
QuarterlyMean
Mean
13.63
24
106
D24HourMean
QuarterlyMean
Mean
17.97
24
107
D24HourMean
QuarterlyMean
Mean
15.05
24
108
D24HourMean
QuarterlyMean
Mean
13.47
24
109
D24HourMean
QuarterlyMean
Mean
19.56
24
110
D24HourMean
QuarterlyMean
Mean
13.94
25
101
D24HourMean
QuarterlyMean
Mean
17.07
25
102
D24HourMean
QuarterlyMean
Mean
15.32
25
103
D24HourMean
QuarterlyMean
Mean
16.20
25
104
D24HourMean
QuarterlyMean
Mean
18.13
25
105
D24HourMean
QuarterlyMean
Mean
17.13
25
106
D24HourMeari
QuarterlyMean
Mean
15.64
25
107
D24HourMean
QuarterlyMean
Mean
19.99
25
108
D24HourMean
QuarterlyMean
Mean
15.16
25
109
D24HourMean
QuarterlyMean
Mean
12.06
25
110
D24HourMean
QuarterlyMean
Mean
14.08
5.2.2 Loading Custom Air Quality Surfaces
There are multiple ways to add a custom airquality layerto BenMAP: through the Data
Center or during Step 3: What Air Quality.
To review airquality layers saved in the tool's database and add or delete any custom
datasets, go to the Data Center. Click Review Air Quality Layers.
BenMAP User Manual v 0.01
5-4
December 2022
-------
Chapter 5 - Customizing a Benefits Analysis
BenMAP - Benefits Mapping and Analysis Program
- Home nr Data Center 0 Help Feedback
View your active and complete tasks.
Review Air Quality Layets
MANAGE TASKS
REVIEW AIR QUALITY LAYERS
€t
Select a pollutant from the drop-down menu.
Review Air Quality
Ground-level ojow
Fine partides (<2-5 pm)
All air quality surfaces that fit the pollutant you selected will be displayed, as well as the
option to Add an Air Quality Layer.
Review Air Quality
Ground-level ozors
ADO GROUND-LEVEL OZONE AIR QUALITY Li
Name ^ Grid
2023 Poficy Baseline CMAQ 12km Nation
2023 Pofecy implementation CMAQ 12km Nation
flCU_03_MDA8_Baseli«e CMAQ 12km Nation
RCU_03„MDA8_Final CMAQ 12km Nation
You can also add a custom air quality layer in Step 3: What Air Quality. When you reach this
step, you will see a button at the top of the screen to Add an Air Quality Layer.
BenMAP User Manual v 0.01 December 2022
5-5
-------
Chapter 5 - Customizing a Benefits Analysis
o
O
O
O
o
o
©
Where?
what pollutant?
What air quality?
Dorft see the desired an quality dataset below?
Step 1 Select your pre-pollcy air quality data
PARTICLES (<2.3 MM) AIR QUALTTY LAYER
(S) RCU_PM2 5_ArmuaL8aseline
O RCU.PM2.5_Annual_Final
Step 2 Select your post-policy air quality data
O RCU_PM2.5Jvnnual_Baseline
(S) RCU_PM2 5_A^BJiaLFInal
D24HourMean. Quarter lyMean
D24HourMeart. QuarterlyMean
Number of ge desired air quality dataset below'
Upload your CSV __
Step 1. Select your pre-pollcy air quality data
0 08 / 0 OCX »¦
P».
(•) RCU.PM2 5_AnnuaLBaseime
O RCU_PM2 5_AnnueLF'nal
Add Fine particles (<2.5 pm) Air Ouality Layer
Name
"Grid -
024HourMean, QuartsilyMean
BB
Input me characteritlic
Num&er ol grid wfis
input (Ue characters!!
For more information on uploading custom air quality surfaces, refer to Section 1.5 of this
user manual.
5.2.3 Validating Custom Air Quality Surfaces
After uploading your custom air quality layer, Ben MAP will validate the surface to ensure it is
formatted correctly. If your air quality layer passes validation and is successfully uploaded,
the following pop-up message will appear. Click OK to dismiss.
BenMAP User Manual v 0.01
5-6
December 2022
-------
Chapter 5 - Customizing a Benefits Analysis
Upload your CSV
~~ 221.08/ 0.00%
AQ#7.csv
221.08/0.00%
Add Fine parti<
Name
Example PM
Your upload of AQ#7.csv was successful
OK
X
Your custom air quality surface will appear in the Review Air Quality Layers window of the
data center. You can select the layer and review the statistics displayed below to check the
data that you just imported.
Air Quafity Cells
¦VnmSMton.MHouMu
WirenSiMKM .BSHOvtMb
»«7U
IM5746
»5S7W
W 56076
faaaas
XM3942
am
If there is an error with your custom air quality surface, it will not pass the validation step and
an error message will appear. A pop-up with possible errors is shown below.
BenMAP User Manual v 0.01
5-7
December 2022
-------
Chapter 5 - Customizing a Benefits Analysis
Your upload of 03_errors.csv has failed
Your file has the following errors
2 records have Column values that are not valid
Error
Integers.
Error 2 records have Row values that are not valid integers.
1 record has air quality values that is not a valid
Error
number.
Error 1 record has air quality values below zero.
PRINT ¦ OK
Other potential errors not shown here include using an unavailable pollutant or using an air
quality surface name that's already taken. Click OK and address the listed error(s) in the input
file you created before reuploading.
BenMAP User Manual v 0.01
5-8
December 2022
-------
Chapter 6
Terminology
In this chapter, find...
• Definitions for common terms used in the BenMAP tool
and in this manual.
BenMAP User Manual v 0.01
6-i
December 2022
-------
Chapter 6 - Terminology
Aggregation. The summing of grid cell level resulting in a larger spatial scale, such as county,
state, or national levels.
Air Quality Metric. A value that expresses both the time period over which air quality values
are modeled or observed and whether that modeled or observed air quality value is an
average, maximum, or minimum. For example, the metric DailyMean represents the average
concentration for the sampling day. This could be taken directly from a single 24-hour
observation or from an average of hourly (or more frequent) observations. In addition to the
time period, some metrics also specify the method used for averaging or aggregation. For
example, a typical ozone metric D8HourMax represents the highest of the 8-hour moving
averages during the day.
Air Quality Model. Air quality management tools that mathematically describe pollution
transport, dispersion, and related physical and chemical processes in the atmosphere. Air
quality models (like CMAQ10 and CAMx11) are used to estimate the air pollutant concentration
at specific locations, which are referred to as receptors, or over a spatial area that has been
divided into uniform grid squares. The number of receptors or grid cells in a model far
exceeds the number of monitors one could typically afford to deploy in a monitoring study.
Therefore, models provide a cost-effective way to analyze pollutant impacts over a wide
spatial area where factors such as meteorology, topography, and emissions from both local
and remote sources could be important. BenMAP does not contain an air quality model but
can use the output from these models to estimate health impacts.
Air Quality Surface. A file containing modeled or monitored air pollution data in a series of
cells; these cells may be a regular shape (such as a 12km by 12km grid) or an irregular shape
10 Community Multi-scale Air Quality (CMAQ) Model is available at: https://www.epa.Rov/cmaq or
https://www.cmascenter.orR/cmaq/.
11 Comprehensive Air Quality Model with Extensions (CAMx) is available at: http://www.camx.com/.
BenMAP User Manual v 0.01
6-1
December 2022
-------
Chapter 6 - Terminology
(such as a county or census tract). These surfaces are also referred to as air quality grids.
BenMAP uses one air quality grid to represent the pre-policv scenario and a second grid to
represent the post-policy scenario. These pre-policy and post-policy grids must share the
same geographic structure. The program calculates the difference between the pre- and post-
policy grids as an input to the health impact function.
Attainment. The state of meeting the National Ambient Air Quality Standard (NAAQS)
standard for a pollutant and not contributing to a nearby area exceeding the standard. A
geographical area that meets the NAAQS is called an "attainment area." BenMAP modeling
results are not designed to designate attainment for regulatory purposes.
Background Concentration. The concentration of a pollutant, generally in the absence of
human sources.
Baseline Incidence. The incidence of a given adverse effect due to all causes including air
pollution. Also called background incidence rates.
Beta. The coefficient for the health impact function that measures the strength of the impact
of air pollution exposure on a health effect. The value of beta (ft) typically represents the
percent change in a given adverse health impact per unit of pollutant exposure.
Cessation Lag. This term represents the time that is expected to elapse between a reduction
in pollutant exposure and the achievement of new steady state health risk level in the
exposed population. The cessation lag model also specifies the proportion of the overall risk
reduction occurring in the population during each time step (e.g., year) of the lag period. The
result of applying the cessation lag model is an expected time stream of health benefits from
the initial exposure reduction to the end of the lag period, which is then used in the benefits
valuation step.
BenMAP User Manual v 0.01
6-2
December 2022
-------
Chapter 6 - Terminology
Concentration-Response (C-R) Function. An equation that estimates the relationship
between adverse health effects and ambient air pollution and is used to derive health impact
functions (defined below). You will often see that the term C-R function and health impact
function are used interchangeably.
Cost of Illness (COI). An estimate of the value of a health effect that includes the direct
medical costs incurred and indirect impacts such as lost earnings associated with illness.
These estimates generally understate the true economic value of reductions in risk of a health
effect because they only include the direct expenditures related to treatment and lost
earnings but not the value of other impacts such as avoided pain and suffering.
Deltas. The difference between two data points. As used in BenMAP, mapping the air quality
deltas shows the change in air pollution between the pre-policy air quality grid and the post-
policy air quality grid.
Discount Rate. In a cost-benefit analysis, the discount rate is a measure of the degree to
which individuals value benefits or costs today more highly than future benefits and costs
Typically, if a benefit occurs over multiple years, the economic benefit would be discounted.
Epidemiology. The study of factors affecting the health and illness of populations.
FIPS Code. Federal Information Processing Standard codes. FIPS codes uniquely identify
geographic areas; the number of digits vary depending on the level of geography. Each state
in the United States is assigned a 2-digit code (e.g., "06" refers to California). Each county is
assigned a 5-digit code of which the first two digits identify the state to which the county
belongs (e.g., "06037" refers to Los Angeles County, California). Counties can be further
subdivided into census tracts, denoted by FIPS codes of 11 digits.
Grid Cell. One of the many geographic, or spatial, components within a grid definition. These
can be regularly or irregularly shaped.
BenMAP User Manual v 0.01 December 2022
6-3
-------
Chapter 6 - Terminology
Grid Definition. A BenMAP grid definition provides a means of breaking a geographic region
into areas of interest (Grid Cells) that allow for spatial variation in data relevant to a benefits
analysis. All grid definitions must have a unique (i.e., non-repeating) column and row index.
BenMAP currently includes one grid definition with uniformly shaped and sized cells (12 km x
12 km).
Health Effect. A subset of a health effect set which represents a specific type of adverse
health effect and may span multiple medical diagnostic codes. For example, the health effect
set Chronic Effects-All includes health effects Asthma Incidence, Lung Cancer Incidence, and
Stroke Incidence.
Health Effect Set. A group of health impact functions that quantify impacts within certain
categories, including mortality, acute (short-term) morbidity effects, and chronic (long-term)
morbidity effects.
Health Impact Function. An equation that calculates the change in adverse health effects
associated with a change in exposure to air pollution. Based on a C-R function, a typical
health impact function accepts inputs specifying the air quality metric and pollutant; the
change in air pollutant concentration, the size, age, race and ethnicity of the population
affected; and the baseline rate of occurrence of the adverse health effect being evaluated.
Incidence. The total number of adverse health effects in a geographic region in a given unit of
time. In BenMAP, this is the total number of adverse health effects avoided due to a change in
air pollution levels, often reported for a single year.
Incidence Rate. The background rate of new health effects per person over a particular
period of time in a given geographic region. The unit of time is typically a day or a year, but
could be other periods as well (e.g., warm season). The incidence rate must be expressed at
the same time scale as the remainder of the health impact function. For example, a health
BenMAP User Manual v 0.01
6-4
December 2022
-------
Chapter 6 - Terminology
impact function quantifying day-to-day changes in premature death must specify a daily
death rate.
Income Growth Adjustment. Adjusting certain valuation functions to reflect increases in real
income overtime. Generally, an increase in real income implies an increase in the WTP values
used to monetize health effects.
Inflation Adjustment. Adjusting monetized benefits occurring over different years to a
constant dollar year (e.g., 2020 dollars) to account for historical or projected future increases
in prices overtime.
Metadata. Data that serves to provide context or additional information about other data.
BenMAP stores a minimum set of standardized metadata fields for imported data files (e.g.,
file name, file date, reference, import date, and description). For certain data types,
additional metadata are recorded.
Micrograms per Cubic Meter (ng/m3). The unit of measure for particulate matter in the
NAAOS. This unit represents the mass of PM and other pollutants found in a cubic meter of
air.
Model Data. Pollutant concentration data that are generated by running an air quality model
such as CMAQ. This is different from "monitor data," which are based upon observed
measurements of pollutant concentrations.
Monetize. In the context of human health benefits assessment, this is the practice of
expressing society's preferences for avoiding certain health effects as an economic value
(e.g., in U.S. dollars). In BenMAP we estimate monetized benefits by using either Willingness
to Pav or Cost of Illness valuation functions.
BenMAP User Manual v 0.01
6-5
December 2022
-------
Chapter 6 - Terminology
Monitor Data. Pollutant concentration data that are collected from an air quality monitor.
"Raw" monitor data usually refers to data that are taken directly from measurement
networks, with no additional processing of the data. Monitor data are different from "model
data," which are based upon numerical predictions from an air quality model. Use of monitor
data is not available in the current version of the tool but is expected in future versions.
Monitoring. The systematic, long-term assessment of pollutant levels by measuring the
quantity and types of certain pollutants in the surrounding outdoor air. The U.S. EPA reports
monitoring data, as well as other information related to monitoring, available through its Air
Quality System (AOS): https://www.epa.gov/aqs.
Monte Carlo Simulation. A technique used in BenMAP to quantify the confidence intervals
around mean incidence and economic value estimates by randomly sampling uncertainty
distributions related to effect coefficients and/or willingness to pay estimates.
Morbidity. A measure of being diseased or afflicted by an illness (generally non-fatal).
Mortality. A measure of the number of deaths in a given population.
National Ambient Air Quality Standards (NAAQS). The U.S. EPA establishes levels for
pollutants that are considered harmful to public health and the environment. The Clean Air
Act established two types of national air quality standards. Primary standards set limits to
protect public health, including the health of "sensitive" populations such as asthmatics,
children, and the elderly. Secondary standards set limits to protect public welfare, including
protection against decreased visibility and against damage to animals, crops, vegetation, and
buildings. The U.S. EPA has set NAAQS for six principal pollutants, which are called "criteria"
pollutants: carbon monoxide, lead, nitrogen dioxide, ozone, particulate matter (PM2.5, PM10),
and sulfur dioxide.
BenMAP User Manual v 0.01
6-6
December 2022
-------
Chapter 6 - Terminology
Net Present Value (NPV). The economic value of a stream of monetized benefits discounted
back to the starting year of the analysis and summed.
Odds Ratio. A quantitative measure reported in epidemiology studies of the relationship
between exposure to air pollution and a health outcome. Odds Ratios must be converted to
beta coefficients to be used in BenMAP.
Ozone (03). BenMAP focuses on ground-level or "bad" ozone, which is not emitted directly
into the air, but is created by chemical reactions between oxides of nitrogen (NOx) and
volatile organic compounds (VOCs) in the presence of sunlight. Emissions from industrial
facilities and electric utilities, motor vehicle exhaust, gasoline vapors, and chemical solvents
are some of the major sources of NOx and VOC. Breathing ozone can trigger a variety of health
problems, particularly for children, the elderly, and people of all ages who have lung diseases
such as asthma. Ground level ozone can also have harmful effects on sensitive vegetation and
ecosystems.
Particulate Matter. Particulate matter, also known as particle pollution or PM, is a complex
mixture of extremely small particles and liquid droplets. Particle pollution is made up of
multiple components, including acids (such as nitrates and sulfates), organic chemicals,
metals, and soil or dust particles. Once inhaled, these particles can affect the heart and lungs
and cause serious health effects. Includes PM2.5 (particles less than 2.5 microns in
aerodynamic diameter), PM10 (particles less than 10 microns in aerodynamic diameter), and
PM 10-2.5 (particles between 2.5 and 10 microns in aerodynamic diameter).
Parts per Billion (ppb). This unit represents the concentration of the pollutant in a billion
parts of air. Ozone concentrations in BenMAP are reported in units of ppb.
Parts per Million (ppm). This unit represents the concentration of the pollutant in a million
parts of air. Carbon monoxide is often measured in units of ppm.
BenMAP User Manual v 0.01
6-7
December 2022
-------
Chapter 6 - Terminology
Population Exposure versus Personal Exposure. Population (or ambient) exposure refers to
the average air pollution level measured in a grid cell. In contrast, personal exposure keeps
track over the course of a day the exposure individuals encounter in different micro-
environments, such as the freeway, outdoors and indoors. BenMAP only represents
population exposure.
Population-weighted Air Quality. Modeled or monitored ambient concentrations that have
been weighted according to the number of people exposed.
Post-Policy Scenario. In a modeling study, this is a scenario in which emissions from one or
more source sectors are changed (increased or decreased) from a given "pre-policy scenario".
The post-policy scenario generally represents air quality levels after a new policy has been
implemented.
Pre-Policy Scenario. The air quality levels prior to the policy change you are evaluating. The
pre-policy scenario is sometimes referred to as "Business as Usual." This scenario is usually
considered to be the reference scenario against which to compare a potential "post-policy
scenario," in which air quality levels are changed from the baseline levels.
Prevalence Rate. The percentage of individuals in a given population who already have a
given adverse health condition. Used to calculate changes in health conditions among those
who already have a health condition, such as asthmatics.
Regulatory Impact Analysis (RIA). A policy tool used to assess the likely effects of a
proposed regulation or regulatory change. It usually includes detailed analyses to quantify
the costs and benefits of the regulation.
Relative Risk. A measure of the change in risk of an adverse health effect associated with an
increase in air pollution levels in an epidemiology study. More specifically, relative risk is the
ratio of the risk of illness with a higher pollution level to the risk of illness with a lower
BenMAP User Manual v 0.01 December 2022
6-8
-------
Chapter 6 - Terminology
pollution level, where the "risk" is defined as the probability that an individual will become ill.
Also sometimes referred to as a risk ratio or hazard ratio.
Shapefile. A shapefile is a particular type of GIS file that has a .shp extension. These files are
accompanied by companion files with .shx and .dbf extensions and can be used to create
Shapefile Grid Definitions. See http://www.esri.com/librarv/whitepapers/pdfs/shapefile.pdf
for more information.
Task Log. This is a report that contains a record of all the choices involved in creating a
particular file. The task log is exported when exporting results files from the BenMAP tool.
Template. A saved configuration from a previous BenMAP run, specifying all the necessary
details to run a benefits analysis. After loading a template, you may run it as is or make
changes to one or more elements priorto running.
Unit Value. The estimated mean economic value of avoiding a single case of a particular
health effect.
Valuation Function. Valuation functions are used by BenMAP to estimate the economic
values of changes in the incidence of health effects. They usually include unit value estimates
and may also include potential adjustments for inflation, income growth, and discounting.
Variable Datasets. Health Impact functions and valuation functions may sometimes refer to
variables other than those for which BenMAP automatically calculates values. For example,
some valuation functions reference the median income within each area of analysis. To
facilitate this, BenMAP stores variables for use in functions; these variables may be used
globally (e.g., inflation adjustments) or may vary geographically (meaning they are associated
with a particular Grid Definition).
BenMAP User Manual v 0.01
6-9
December 2022
-------
Chapter 6 - Terminology
WTP (Willingness to Pay). The willingness of individuals to pay for a good or service, such as
a reduction in the risk of illness. In general, economists tend to view an individual's WTP for
an improvement in environmental quality as the appropriate measure of the value of a risk
reduction. An individual's willingness to accept (WTA) compensation for not receiving an
improvement is also a valid measure. However, WTP is generally considered to be a more
readily available and conservative measure of benefits.
BenMAP User Manual v 0.01
6-10
December 2022
-------
BenMAR)
Environmental Benefits Mapping and
Analysis Program
User's Manual
Appendices
December 2022
-------
TOC
Table of Contents
Appendix A. Monitor Rollback Algorithms A-l
Appendix B. Algorithms for Estimating Air Pollution Exposure B-l
B.l Direct Modeling B-2
B.2 Monitor Data B-2
Appendix C. Deriving Health Impact Functions C-l
C.l Overview C-2
C.2 Review Relative Risk and Odds Ratio C-3
C.3 Linear Model C-5
C.4 Log-linear Model C-6
C.5 Logistic Model C-9
C.6 Cox Proportional Hazards Model C-17
Appendix D. U.S. Health Incidence & Prevalence Data in BenMAP D-l
D.l Mortality D-l
D.l.l Mortality Data for 2012-2014 D-l
D.1.2 Mortality Rate Projections 2015-2060 D-6
D.1.3 Race-Stratified Mortality Incidence D-7
D.1.4 Ethnicity-Stratified Mortality Incidence D-9
D.2 Hospitalizations D-9
D.3 Nonfatal Heart Attacks D-16
D.4 Emergency Department Visits D-17
D.5 School Loss Days D-19
D.6 Asthma-Related Health Effects D-21
BenMAP User Manual v 0.01 December 2022
-------
TOC
D.6.1 New Onset Asthma D-22
D.6.2 Shortness of Breath D-23
D.6.3 Wheeze D-23
D.6.4 Cough D-23
D.6.5 Albuterol Use D-24
D.6.6 Upper Respiratory Symptoms D-24
D.6.7 Asthma Population Estimates D-24
D.7 Other Acute and Chronic Effects D-25
D.7.1 Acute Bronchitis D-26
D.7.2 Chronic Bronchitis Incidence Rate D-27
D.7.3 Chronic Bronchitis Prevalence Rate D-27
D.7.4 Lower Respiratory Symptoms D-27
D.7.5 Minor Restricted Activity Days (MRAD) D-28
D.7.6 Work Loss Days D-28
D.8 Other Health Effect Incidence D-28
D.8.1 Allergic Rhinitis D-29
D.8.2 Lung Cancer D-29
D.8.3 Out of Hospital Cardiac Arrest D-30
D.8.4 Stroke D-31
Appendix E. Core Particulate Matter Health Impact Functions in BenMAP E-l
E.l Long-term Mortality E-l
E.l.l Wu et al. (2020) E-3
E.1.2 Pope et al. (2019) E-4
BenMAP User Manual v 0.01 December 2022
ii
-------
TOC
E.1.3 Di et al. (2017) E-5
E.1.4 Turner et al. (2016) E-6
E.1.5 Woodruff et al. (2008) E-7
E.2 Chronic/Severe Illness E-8
E.2.1 Ensoret al. (2013) E-9
E.2.2 Gharibvand et al. (2017) E-10
E.2.3 Kioumourtzoglou et al. (2016) E-ll
E.2.4 Kloogetal. (2012) E-13
E.2.5 Peters et al. (2001) E-14
E.2.6 Pope et al. (2006) E-16
E.2.7 Rosenthal et al. (2008) E-17
E.2.8 Silverman et al. (2010) E-18
E.2.9 Sullivan et al. (2005) E-19
E.2.10 Zanobetti and Schwartz (2006) E-20
E.2.12 Wei etal. (2019) E-23
E.3 Hospitalizations E-24
E.3.1 Bell etal. (2015) E-25
E.3.2 Ostro et al. (2009) E-26
E.4 Emergency Room Visits E-27
E.4.1 Ostro et al. (2016) E-28
E.4.2 Kralletal. (2016) E-29
E.5 Minor Effects E-30
E.5.1 Ostro (1987) E-31
BenMAP User Manual v 0.01
December 2022
-------
TOC
E.5.2 Ostro and Rothschild (1989) E-33
E.5.3 Parker et al. (2009) E-35
E.6 Asthma-Related Effects E-36
E.6.1 Rabinovitch et al. (2006) E-37
E.6.2 Tetreaultetal. (2016) E-38
E.7 Sensitivity Analysis - General E-39
E.7.1 Dietal. (2017) E-41
E.7.2 Jones et al. (2015) E-42
E.7.3 McConnell et al. (2010) E-43
E.7.4 Nishimura et al. (2013) E-44
E.7.5 Pope etal. (2015) E-46
E.7.6 Talbott et al. (2014) E-47
E.7.7 Turner et al. (2016) E-48
E.7.8 Zanobetti et al. (2009) E-49
E.8 Sensitivity Analysis - At-Risk Populations E-50
E.8.1 Alhanti et al. (2016) E-51
E.8.2 Dietal. (2017) E-52
Appendix F. Core Ozone Health Impact Functions in BenMAP F-l
F.l Short-term Mortality F-l
F.l.l Katsouyanni et al. (2009) F-3
F.1.2 Turner et al. (2016) F-5
F.1.3 Zanobetti and Schwartz (2008) F-6
F.2 Hospital Admissions F-7
BenMAP User Manual v 0.01 December 2022
iv
-------
TOC
F.2.1 Katsouyanni et al. (2009) F-8
F.3 Emergency Room Visits F-9
F.3.1 Barry et al. (2018) F-10
F.4 Minor Effects F-ll
F.4.1 Gillilandetal. (2001) F-12
F.4.2 Ostro and Rothschild (1989) F-15
F.4.3 Parker et al. (2009) F-17
F.5 Asthma-Related Effects F-18
F.5.1 Lewis et al. (2013) F-19
F.5.2 Tetreaultetal. (2016) F-20
F.6 Sensitivity Analysis - General F-21
F.6.1 Dietal. (2017) F-22
F.6.2 Garcia et al. (2019) F-23
F.6.3 Katsouyanni et al. (2009) F-24
F.6.4 Turner et al. (2016) F-25
F.7 Sensitivity Analysis - At-Risk Populations F-26
F.7.1 Cakmak et al. (2006) F-30
F.7.2 Jerrettet al. (2009) F-31
F.7.3 Katsouyanni et al. (2009) F-32
F.7.4 Lin etal. (2005) F-33
F.7.5 Mar and Koenig (2009) F-34
F.7.6 Medina-Ramon & Schwartz (2008) F-35
F.7.7 Paulu and Smith (2008) F-37
BenMAP User Manual v 0.01
v
December 2022
-------
TOC
F.7.8 Villeneuveetal. (2007) F-38
F.7.9 Zanobetti and Schwartz (2008b) F-39
Appendix G. Additional Health Impact Functions in BenMAP G-l
Appendix H. Core Health Valuation Functions in BenMAP H-l
H.l Mortality H-2
H.l.l Value of a Statistical Life Based on 26 Studies H-3
H.2 Hospital Admissions & Emergency Room Visits H-4
H.2.1 Hospital Admissions H-4
H.2.2 Emergency Room Visits H-9
H.2.3 Emergency Room Visits for Asthma H-ll
H.3 Other Health Effect Occurrence H-12
H.3.1 Lung Cancer H-12
H.3.2 Out of Hospital Cardiac Arrest H-15
H.3.3 Stroke H-16
H.4 Acute Symptoms and Illness Not Requiring Hospitalization H-17
H.4.1 Non-Fatal Myocardial Infarctions (Heart Attacks) H-18
H.4.2 Minor Restricted Activity Days (MRADs) H-20
H.4.3 New Onset Asthma H-21
H.4.4 Asthma Symptoms H-22
H.4.5 Allergic Rhinitis H-23
H.4.6 Work Loss Days (WLDs) H-23
H.4.7 School Loss Days H-24
H.5 Developing Income Growth Adjustment Factors H-26
BenMAP User Manual v 0.01 December 2022
vi
-------
TOC
H.6 Inflation Indices H-30
Appendix I. Additional Health Valuation Functions in BenMAP 1-1
Appendix J. Population & Other Data in BenMAP J-l
J.l Population Data for the U.S J-2
J.1.1 How BenMAP Forecasts Population J-3
J.1.2 Data Needed for Forecasting J-5
J.2 U.S. Demographic Datasets in BenMAP J-ll
J.2.1 Household Size J-ll
J.2.2 Educational Attainment J-12
J.2.3 Poverty Status J-12
J.2.4 Unemployment Rates J-13
J.2.5 Health Insurance J-14
J.2.6 Blue Collar Workers J-14
Appendix K. Uncertainty & Pooling K-l
K.l Uncertainty K-l
K.l.l Characterization of Uncertainty Surrounding Incidence Changes K-2
K.l.2 Characterization of Uncertainty Surrounding Dollar Benefits K-4
Appendix L. Batch Run Approach L-l
Appendix M. Function Editor M-l
References R-l
BenMAP User Manual v 0.01 December 2022
vii
-------
Appendix A: Monitor Rollback Algorithms
Appendix A. Monitor Rollback Algorithms
This Appendix explains the algorithms BenMAP applies when you perform a "monitor
rollback" to create an air quality surface. The monitor rollback procedure adjusts the air
quality monitoring data using simplified rules to reflect hypothetical broad-scale changes in
air pollution across a given study area. This functionality is not implemented in the current
version of the web-based tool, though it is planned for a future version.
BenMAP User Manual v 0.01
A-l
December 2022
-------
Appendix B: Algorithms for Estimating Air Pollution Exposure
Appendix B. Algorithms for Estimating Air
Pollution Exposure
BenMAP groups counts of individuals into what we refer to as "population grid cells," where
the grid cells typically correspond to a regular grid such as those used in an air quality
models, or to a grid defined by political boundaries such as the counties of the United States.
The default data in the BenMAP web tool allocates the United States population to the same
12km by 12km grid used in the Community Multiscale Air Quality modeling system (CMAQ).
BenMAP estimates the air pollution exposure forthe population in each grid-cell by assuming
that people living within a particular grid-cell experience the same air pollution levels.
The goal of estimating exposure is to provide the necessary input for concentration-response
functions, so that BenMAP can estimate the impact of air pollution on adverse health effects.
BenMAP can recognize multiple air pollution metrics. Table B-l lists the types of metrics
commonly used in concentration-response functions.
Table B-l. Example Metrics Used in Concentration-Response Functions for Criteria Air
Pollutants
Measurement
Frequency
Metric Name
Metric Description
Daily (e.g., PM2.5)
Daily Average
Daily 24-hour average
Annual Average
Average of four quarterly averages. The four
quarters are defined as: Jan-Mar, April-June,
Jul-Sep, Oct-Dec.
Hourly (e.g., Ozone)
1-hour Daily Max
Highest hourly value from 12:00 A.M. through
11:59 P.M.
8-hour Daily Average
Average of hourly values from 9:00 A.M. through
4:59 P.M.
24-hour Daily Average
Average of hours from 12:00 A.M. through 11:59
P.M.
BenMAP User Manual v 0.01
B-l
December 2022
-------
Appendix B: Algorithms for Estimating Air Pollution Exposure
Multiple approaches exist to estimate the exposure to air pollution for the people living
within a given population grid-cell, including the use of air quality model-based
concentrations, and data from ground-level measurements at monitors.
B.i Direct Modeling
When using direct modeling data to estimate exposure, BenMAP assumes that all people
living within a particular air pollution grid-cell experience the air pollution concentration
estimated by the air quality model for the given air quality metric.
B.2 Monitor Data
An alternative approach is to use air pollution monitoring data, where you may choose the
closest monitor data to the center of a grid-cell or spatially interpolate values based on of
nearby monitors. Placeholder: this functionality is planned for a future version of the
tool.
BenMAP User Manual v 0.01
B-2
December 2022
-------
Appendix C: Deriving Health Impact Functions
Appendix C. Deriving Health Impact
Functions
This Appendix provides of an overview regarding the health impact functions that BenMAP
uses to estimate the impact of a change in air pollution on adverse health effects. It provides
a description of the particular types of health impact functions that BenMAP uses.
The functional form of the relationship between the change in pollutant concentration, Ax,
and the change in population health response (usually an incidence rate), Ay depends on the
functional form of the C-R function from which it is derived, and this depends on the
underlying relationship assumed in the epidemiological study chosen to estimate a given
effect. For expository simplicity, the following subsections refer simply to a generic adverse
health effect, "y" and uses particulate matter (PM) as the pollutant - that is, Ax = APM - to
illustrate how the relationship between Ax and Ay is derived from each of several different C-
R functions.
Estimating the relationship between APM and Ay can bethought of as consisting of three
steps:
1. choosing a functional form of the relationship between PM and y (the C-R function),
2. estimating the values of the parameters in the C-R function assumed, and
3. deriving the relationship between APM and Ay (the health impact function) from the
relationship between PM and y (the C-R function).
Epidemiological studies have used a variety of functional forms for C-R functions. Some
studies have assumed that the relationship between adverse health and pollution is best
described by a linearform, where the relationship between y and PM is estimated by a linear
BenMAP User Manual v 0.01
C-l
December 2022
-------
Appendix C: Deriving Health Impact Functions
regression in which y is the dependent variable and PM is one of several independent
variables. Log-linear regression and logistic regression are other common forms.
Note that the log-linear form used in the epidemiological literature is often referred to as
"Poisson regression" because the underlying dependent variable is a count (e.g., number of
deaths), believed to be Poisson distributed. The model parameters may be estimated by
regression techniques but are often estimated by maximum likelihood techniques. The form
of the model, however, is still log-linear.
C.i Overview
The relationship between the concentration of a pollutant, x, and the population response, y,
is called the concentration-response (C-R) function. For example, the concentration of fine
particulate matter (PM2.5) may be in |ig/m3 per day, and the population response may be the
number of premature deaths per 100,000 population per day. C-R functions are estimated in
epidemiological studies. A functional form is chosen by the researcher, and the parameters of
the function are estimated using data on the pollutant (e.g., daily levels of PM2.5) and the
health response (e.g., daily mortality counts). There are several different functional forms,
discussed below, that have been used to estimate C-R functions. The one most commonly
used is the log-linear form, in which the natural logarithm of the health response is a linear
function of the pollutant concentration.
Forthe purposes of estimating benefits, we are not interested in the C-R function itself,
however, but the relationship between the change in concentration of the pollutant, Ax, and
the corresponding change in the population health response, Ay. We want to know, for
example, if the concentration of PM2.5 is reduced by 10 |ig/m3, how many premature deaths
will be avoided? The relationship between Ax and Ay can be derived from the C-R function, as
described below, and we refer to this relationship as a health impact function.
BenMAP User Manual v 0.01
C-2
December 2022
-------
Appendix C: Deriving Health Impact Functions
Many epidemiological studies, however, do not report the C-R function, but instead report
some measure of the change in the population health response associated with a specific
change in the pollutant concentration. The most common measure reported is the relative
risk associated with a given change in the pollutant concentration. A general relationship
between Ax and Ay can, however, be derived from the relative risk. The relative risk and
similar measures reported in epidemiological studies are discussed in the sections below. The
derivation of the relationship of interest for BenMAP - the relationship between Ax and Ay - is
discussed in the subsequent sections.
C.2 Review Relative Risk and Odds Ratio
The terms relative risk and odds ratio are related but distinct. Table C-l provides the basis for
demonstrating their relationship.
Table C-2. Relative Risk and Odds Ratio Notation
Exposure
Fraction of Population
Adverse Effect Measure
Affected
Not Affected
Relative Risk
Odds
Baseline Pollutant Exposure
yo
l-yo
Yo/Yc
Yo/(1-Yo)
Control Pollutant Exposure
Yc
1-Yc
Yc/(1-Yc)
The "risk" that people with baseline pollutant exposure will be adversely affected (e.g.,
develop chronic bronchitis) is equal to y0, while people with control pollutant exposure face a
risk, y0, of being adversely affected. The relative risk (RR) is simply:
Equation C-l
RR= —
yc
BenMAP User Manual v 0.01
C-3
December 2022
-------
Appendix C: Deriving Health Impact Functions
The odds that an individual facing high exposure will be adversely affected is:
Equation C-2
yo
Odds =
i-yo
The odds ratio is then:
Equation C-3
( yo )
u-yJ
Odds Ratio =
(t^)
This can be rearranged as follows:
Equation C-4
Odds Ratio = — x (] ~yc) = RR x —)
vr VI - Vn/ VI - yJ
As the risk associated with the specified change in pollutant exposure gets small (i.e., both y0
and yc approach zero), the ratio of (l-yc) to (l-y0) approaches one, and the odds ratio
approaches the relative risk. This relationship can be used to calculate the pollutant
coefficient in the C-R function from which the reported odds ratio or relative risk is derived, as
described below.
BenMAP User Manual v 0.01
C-4
December 2022
-------
Appendix C: Deriving Health Impact Functions
C.3 Linear Model
A linear relationship between the rate of adverse health effects (incidence rate) and various
explanatory variables is of the form:
Equation C-5
y = a + (3 x PM
where a incorporates all the other independent variables in the regression (evaluated at their
mean values, for example) times their respective coefficients. The relationship between the
change in the rate of the adverse health effect from the baseline rate (y0) to the rate after
control (yc) associated with a change from PM0 to PMC is then:
Equation C-6
Ay = y0 - yc = P * (.pm0 - pmc) = p * apm
For example, Ostro et al. (1991, Table 5) reported a PM2.5 coefficient of 0.0006 (with a standard
error of 0.0003) for a linear relationship between asthma and PM2.5 exposure.
The lower and upper bound estimates for the PM2.5 coefficient are calculated as follows:
Equation C-7
Plowerbound = P~ (l-96 X Op) = 0.0006 - (1.96 X 0.0003) = 1.2 X 10-5
Pupperbound = P + (l-96 X Ojg) = 0.0006 + (1.96 X 0.0003) = 0.00119
It is then straightforward to calculate lower and upper bound estimates of the change in
asthma.
BenMAP User Manual v 0.01
C-5
December 2022
-------
Appendix C: Deriving Health Impact Functions
C.4 Log-linear Model
The log-linear relationship defines the incidence rate (y) as:
Equation C-8
y = B x eP*PM
Or, equivalently,
Equation C-9
ln(y) = a + (3 * PM,
where the parameter B is the incidence rate of y when the concentration of PM is zero, the
parameter 3 is the coefficient of PM, ln(y) is the natural logarithm of y, and a = ln(B). Other
covariates besides pollution clearly affect mortality. The parameter B might be thought of as
containing these other covariates, for example, evaluated at their means. That is,
Equation C-10
B = B0 x e^lXl+"'+^nXn
where Bo is the incidence of y when all covariates in the model are zero, and xl,..., xn are the
other covariates evaluated at their mean values. The parameter B drops out of the model,
however, when changes in y are calculated, and is therefore not important.
The relationship between APM and Ay is:
Equation C-ll
A y = y0-yc=B(ePpM°-ePpMc)
BenMAP User Manual v 0.01
C-6
December 2022
-------
Appendix C: Deriving Health Impact Functions
This may be rewritten as:
Equation C-12
Ay = Bx e"r"'(l - l)= ^
! A
ex p(fi x APM)
where y0 is the baseline incidence rate of the health effect (i.e., the incidence rate before the
change in PM).
The change in the incidence of adverse health effects can then be calculated by multiplying
the change in the incidence rate, Ay, by the relevant population (e.g., if the rate is number per
100,000 population, then the relevant population is the number of 100,000s in the
population).
When the PM coefficient (3) and its standard error (a3) are published (e.g., Ostro et al., 1989),
then the coefficient estimates associated with the lower and upper bound may be calculated
easily as follows:
Equation C-13
Plowerbound ~ P ~ (l.96 X Ojg)
Pupperbound ~ P (l.96 X Ojg),
where the adjustment on the mean of ±1.96 times the standard error produces the 2.5th and
97.5th percentiles of the normal distribution, which are used to approximate a 95% confidence
interval. These values can be changed to capture different lower and upper bounds.
Epidemiological studies often report a relative risk for a given APM, rather than the
coefficient, 3 (e.g., Schwartz et al., 1995, Table 4). Recall that the relative risk (RR) is simply
the ratio of two risks:
BenMAP User Manual v 0.01
C-7
December 2022
-------
Appendix C: Deriving Health Impact Functions
Equation C-14
RR = — = ep'^M
yc
Taking the natural log of both sides, the coefficient in the C-R function underlying the relative
risk can be derived as:
Equation C-15
Mm)
APM
The coefficients associated with the lower and upper bounds (e.g., the 2.5th and 97.5th
percentiles) can be calculated by using a published confidence interval for relative risk, and
then calculating the associated coefficients.
Because of rounding of the published RR and its confidence interval, the standard error for
the coefficient implied by the lower bound of the RR will not exactly equal that implied by the
upper bound, so an average of the two estimates is used. The underlying standard error for
the coefficient (ap) can be approximated by:
P, 2.5 percentile
Equation C-16
_P~P2 .5 percentile
1.96
_ P'>i .5percentile P
^p,97'.5 percentile ^
^ ® j3,2.5 percentile ® j3,97.5 percentile
P ~ o
BenMAP User Manual v 0.01
C-8
December 2022
-------
Appendix C: Deriving Health Impact Functions
C.5 Logistic Model
In some epidemiological studies, a logistic model is used to estimate the probability of an
occurrence of an adverse health effect. Given a vector of explanatory variables, X, the logistic
model assumes the probability of an occurrence is:
f ex'p \
where 3 is a vector of coefficients. Greene (1997, p. 874) presents models with discrete
dependent variables, such as the logit model. See also Judge et al. (1985, p. 763). This may be
rewritten as:
Equation C-18
ex-» ex-f 1
Equation C-17
y = probioccurrence X / fJ>) =
BenMAP User Manual v 0.01
C-9
December 2022
-------
Appendix C: Deriving Health Impact Functions
The odds of an occurrence is:
Equation C-19
1
odds =
y
i-y
l + e
-x*p
1-
1 + e
-x*p
( i W i ^
odds = ¦
l + e
-x»p
l + e
-x»p
1-
f e-z-P \
l + e
-x*p
l + e
-x*p
-x*p
= ex'p
=> In(odds) = X x p
The odds ratio for the control scenario (oddsc) versus the baseline (odds0) is then:
Equation C-20
odds ratio =
oddsc
oddsn
yc
i-v,
c J
-XC*P
f \ f
^0
1 - V0
1
-x0'P
The change in the probability of an occurrence from the baseline to the control (Ay),
assuming that all of the other covariates remain constant, may be derived from this odds
ratio:
BenMAP User Manual v 0.01
C-10
December 2022
-------
Appendix C: Deriving Health Impact Functions
Equation C-21
y0
1 _ v pX»'P
odds ratio = = ep^x = ep'*™
yc ex''p
1-yc
yc _ yo w
— Xc-
l~yc l~y0
y +y xe"^"
~
e^+y0,
: pop
BenMAP User Manual v 0.01
C-ll
December 2022
-------
Appendix C: Deriving Health Impact Functions
When the coefficient (3) and its standard error (ap) are published (e.g., Pope et al., 1991, Table
5), then the coefficient estimates associated with the lower and upper bound may be
calculated easily as follows:
Equation C-22
Plowerbound ~ P ~ (l.96 X Ojg)
Pupperbound ~ P (l.96 X Ojg),
where the adjustments to the mean of plus or minus 1.96 times the standard error represent
the 2.5th and 97.5th percentiles of the normal distribution and are used to approximate a 95%
confidence interval. These values can be changed to capture different lower and upper
bounds.
Often the logistic regression coefficients are not published, and only the odds ratio
corresponding to a specified change in PM is presented (e.g., Schwartz et al., 1994). It is easy
to calculate the underlying coefficient as follows:
Equation C-23
ln(odds ratio) =j?x A PM
ln(odds ratio)
^ ^ ~ APM
BenMAP User Manual v 0.01
C-12
December 2022
-------
Appendix C: Deriving Health Impact Functions
The coefficients associated with the lower and upper bound estimates of the odds ratios are
calculated analogously. The underlying standard errorforthe coefficient (a a) can be
approximated by:
Equation C-24
_P~P2 .5 percentile
^(3,2.5 percentile
1.96
_ /^97.5 percentile P
^f!,97.5 percentile j
^ ^ P,2.5 percentile *^"/?,97.5 percentile
Sometimes, however, the relative risk is presented. The relative risk does not equal the odds
ratio, and a different procedure should be used to estimate the underlying coefficient. Note
that ESEERCO (1994, p. V-21) calculated (incorrectly) the underlying regression coefficient for
Abbey etal. (1993,Table 5) bytakingthe logarithm ofthe relative risk and dividing by the
change in TSP.
The relative risk (RR) is simply:
Equation C-25
RR = ^
yc
where y0 is the risk (i.e., probability of an occurrence) at the baseline PM exposure and yc is
the risk at the control PM exposure. When the baseline incidence rate (y0) is given, then it is
easy to solve for the control incidence rate (yc):
BenMAP User Manual v 0.01
C-13
December 2022
-------
Appendix C: Deriving Health Impact Functions
Equation C-26
y = *L,
c RR
The odds ratio, may then be calculated:
Equation C-27
Vq
odds ratio = -——
yc
1 -vc
Given the odds ratio, the underlying coefficient (3) may be calculated as before:
Equation C-28
_ In(odds ratio)
APM '
The odds ratio and the coefficient calculated from it are dependent on the baseline and
control incidence rates. Unfortunately, it is not always clear what the baseline and control
incidence rates should be. Abbey et al. (1995b, Table 2) reported that there are 117 new cases
of chronic bronchitis out of a sample of 1,631, or a 7.17 percent rate. In addition, they
reported the relative risk (RR = 1.81) for a new case of chronic bronchitis associated with an
annual mean concentration "increment" of 45 |ig/m3 of PM2.5 exposure.
Assuming that the baseline rate for chronic bronchitis (yO) should be 7.17 percent, the
question becomes whether the "increment" of 45 |ig/m3 should be added to or subtracted
from the existing PM2.5 concentration. If added, the control incidence rate (yc) would be
BenMAP User Manual v 0.01 December 2022
C-14
-------
Appendix C: Deriving Health Impact Functions
greater than the baseline rate (y0), while subtraction would give a control rate less than the
incidence rate. In effect, one might reasonably derive two estimates of the odds ratio:
odds ratio, =
Equation C-29
1.81x0.0717
f
f
I1
O
1
ll
f
I1
-yj
1 — (l.81 x 0.0717)y
0.0717 ^
1-0.0717
= 1.931
odds ratio-, =
r v0 ^ ( 0.0717
,1-VoJ U- 0.0717
2 A vc ^ f 0.0717 ^
_ 181
0.0717
= 1.873
1-v,
v -ycJ
1-
1.81 )
^ = Mill) = 0.01462
45
ML8Z1), 0.01394
45
An alternative is to simply assume that the relative risk (1.81) is reasonably close to the odds
ratio and calculate the underlying coefficient. It is easy to show that the relative risk equals:
Equation C-30
RR=— =
yc
BenMAP User Manual v 0.01
C-15
December 2022
-------
Appendix C: Deriving Health Impact Functions
Assuming that:
Equation C-31
-APM»P
RR = e
It is then possible to calculate the underlying coefficient:
Equation C-32
In(RR)
-APM
= P
ML8l)= 9
3 45
Since this coefficient estimate is based on the assumption that
Equation C-33
= (1 -y0)xC-a»'-" + v0
it should be used in a C-R function that maintains this assumption. In effect, it should be
applied to a log-linear C-R function:
Equation C-34
A.v = [v0x (/•«-!)]
BenMAP User Manual v 0.01
C-16
December 2022
-------
Appendix C: Deriving Health Impact Functions
Usingthe formula forthe change in the incidence rate and assuming a 10 |ig/m3 decline in
PM2.5, it is shown that this results in changes within the bounds suggested by the two
estimates based on usingthe estimated odds ratios:
Equation C-35
0717
Avi = 7 , ' ,0Qm46O 0.0717 = -0.00914
(l-0.0717)xe10 0 01462 + 0 . 07 1 7
0717
Av0 = 7 , ' 10 oon94 0.0717 = -0.00874
" * (l-0.0717)xe10 0 01394 + 0 . 07 1 7
Av3 = 0.0717 x (e-10 0 01319 _ i)= -0.00886
In this instance, it seems that simply usingthe relative risk to estimate the underlying
coefficient results in a good approximation of the change in incidence. Since it is unclear
which of the two other coefficients (3i or p2) should be used - as the published work was not
explicit - the coefficient based on the relative risk and the log-linear functional form is a
reasonable approach.
C.6 Cox Proportional Hazards Model
Use of a Cox proportional hazards model in an epidemiological study results in a C-R function
that is log-linear in form. It is often used to model survival times, and as a result, this
discussion focuses on mortality impacts.
The Cox proportional hazards model is based on a hazard function, defined as the probability
that an individual dies at time t, conditional on having survived up to time t (Collet, 1994, p.
BenMAP User Manual v 0.01
C-17
December 2022
-------
Appendix C: Deriving Health Impact Functions
10). More formally, the hazard function equals the probability density function for the risk of
dying divided by one minus the cumulative probability density function:
where X is a vector of explanatory variables, 3 is a vector of coefficients, and h0(t) is the so-
called "baseline hazard" rate. This terminology differs from that used in most of this
discussion: this "baseline hazard" is the risk when all of the covariates (X) are set to zero; this
is not the risk in the baseline scenario.
The Cox proportional hazards model is sometimes termed a "semi-parametric" model,
because the baseline hazard rate is calculated using a non-parametric method, while the
impact of explanatory variables is parameterized. Collet (1994) details the estimation of Cox
proportional hazards models; in particular, see Collet's discussion (pp. 95-97) of
nonparametric estimation of the baseline hazard.
Equation C-36
The proportional hazards model takes the form:
Equation C-37
h(x,t) = h0(t)ex'p
BenMAP User Manual v 0.01
C-18
December 2022
-------
Appendix C: Deriving Health Impact Functions
Takingthe ratio of the hazard functions forthe baseline and control scenarios gives the
relative risk:
Equation C-38
nn _ KXq ' 0 _ K (fy ° P _ APA/./J
m-h(xc,,)--e
}
where it is assumed that the only difference between the baseline and control is the level of
PM pollution.
The relative risk is often presented rather than the coefficient 3, so it is necessary to estimate
3 in orderto develop the functional relationship between APM and Ay, as described
previously for log-linear C-R functions.
BenMAP User Manual v 0.01
C-19
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Appendix D. U.S. Health Incidence &
Prevalence Data in BenMAP
Health impact functions developed from log-linear or logistic models estimate the percent
change in an adverse health effect associated with a given pollutant change. In order to
estimate the absolute change in incidence using these functions, we need the baseline
incidence rate of the adverse health effect. And for certain health effects, such as asthma
exacerbation, we need a prevalence rate, which estimates the percentage of the general
population with a given ailment like asthma. This appendix describes the data used to
estimate baseline incidence and prevalence rates forthe health effects considered in this
analysis.
D.i Mortality
This section describes how we developed county mortality rates forthe years 2015 through
2050 to use in BenMAP. First, we describe the source of 2012-2014 baseline mortality data and
how we calculated county-level mortality rates. We then describe how we used national-level
Census mortality rate projections to develop county-level mortality rate projections for years
2015-2060.
D.l.l Mortality Data for 2012-2014
We obtained county-level mortality and population data from 2012-2014 for 11 causes forthe
contiguous United States by downloading the data from the Centers for Disease Control
(CDC) WONDER database (http://wonder.cdc.gov).
Since the detailed mortality data obtained from CDC do not include population, we combined
them with U.S. Census Bureau population estimates exported from BenMAP. We then
BenMAP User Manual v 0.01
D-l
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
generated age-, cause-, and county-specific mortality rates using the following formula
(Equation D-l):
Equation D-l
R _ D i,j,k (2012) ¦+D y, fc (2013) ¦+D y-k (2014)
i,],k ~ Pi,fe(2012)+Pu(2013)+Pu(2014)
where Ru,k is the mortality rate for age group /', cause j, and county k\ D is the death count; and
Pis the population.
For county-age group cells with fewer than 10 deaths, CDC WONDER suppresses the exact
death count. For these observations, a mortality rate cannot be calculated. For each
combination of age group and mortality cause, we used the following procedure to deal with
suppressed counts.
For each combination of state, age group and mortality cause, we grouped counties with
unsuppressed mortality figures and summed their reported death counts. We then
subtracted these unsuppressed deaths from the state-level age- and cause-specific death
count, which includes suppressed deaths. We divided the resulting state-wide death count in
suppressed counties by the age-specific populations in those counties. This calculation
resulted in an age- and cause- specific average mortality rate for suppressed counties;
Equation D-2
_ Dr.i.j ~ Du,i,j
s,i,j ~ p
s,i,j
Where Rs,u is the state average suppressed mortality rate for age group /' and cause/; DTjj, is the
total state death count for age group /' and cause/; Dujj is the aggregated state-level
unsuppressed death count for age group /' and cause/; and Ps,u is the aggregated population
for age group /' and cause j in suppressed counties.
BenMAP User Manual v 0.01
D-2
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
In some instances, age- and cause-specific death counts were suppressed at both the county
and state level. In these cases, we substituted national-level age- and cause-specific mortality
rates for the respective missing county mortality rates.
Following CDC WONDER (http://wonder.cdc.gov), we treated mortality rates as "unreliable"
when the death count is less than 20. For each combination of age group and mortality cause,
we used the following procedure to deal with the problem of "unreliable" rates:
For a given state, we grouped the counties where the death count was less than 20 and
summed those death counts across those counties. If the sum of deaths was greater than or
equal to 20, we then summed the populations in those counties, and calculated a single rate
for the "state collection of counties" by dividing the sum of deaths by the sum of populations
in those counties. This rate was then applied to each of those "unreliable" counties.
If the sum of deaths calculated in the above step was still less than 20, the counties in the
"state collection of counties" were not assigned the single rate from the above step. Instead,
we proceeded to the regional level, according to the regional definitions shown below in
Table D-l. In each region, we identified all counties whose death counts were less than 20
(excluding any such counties that were assigned a rate in the previous step). We summed the
death counts in those counties. If the sum of deaths was greater than or equal to 20, we then
summed the populations in those counties, and calculated a single rate forthe "regional
collection of counties" by dividing the sum of deaths by the sum of populations in those
counties. This rate was then applied to each of those counties in the "regional collection of
counties."
BenMAP User Manual v 0.01
D-3
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-3. Regional Definitions from U.S. Census
Region
States Included
Northeast
Maine, New Hampshire, Vermont, Massachusetts, Rhode Island, Connecticut, New
York, New Jersey, Pennsylvania
Midwest
Ohio, Indiana, Illinois, Michigan, Wisconsin, Minnesota, Iowa, Missouri, North Dakota,
South Dakota, Nebraska, Kansas
South
Delaware, Maryland, District of Columbia, Virginia, West Virginia, North Carolina,
South Carolina, Georgia, Florida, Kentucky, Tennessee, Alabama, Mississippi,
Arkansas, Louisiana, Oklahoma, Texas
West
Montana, Idaho, Wyoming, Colorado, New Mexico, Arizona, Utah, Nevada,
Washington, Oregon, California, Alaska, Hawaii
If the sum of deaths calculated in the previous (regional) step was still less than 20, the
counties in the "regional collection of counties" were not assigned the single rate from the
above step. Instead, we proceeded to the national level, identifying all counties in the nation
whose death counts were less than 20 (excluding any such counties that were assigned a rate
in the previous steps). We summed the death counts in those counties and divided by the
sum of the populations in those counties to derive a single rate for the "national collection of
counties." This rate was then applied to each of those counties in the "national collection of
counties." In these cases where national adjustment still did not yield a death count greater
than 20, we simply calculated a single rate for the "national collection of counties, even
though it was "unreliable," and assigned it to those counties in the "national collection of
counties."
BenMAP User Manual v 0.01
D-4
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-4. National Mortality Rates (per 100 people per year) by Health Effect and Age
Group, 2012-2014
Mortality
Category
ICD-10
codes
Infant*
1-17
18-24
25-34
35-44
45-54
55-64
65-74
75-84
85+
Mortality,
All Cause
All
0.19242
0.01951
0.07804
0.10665
0.17264
0.40542
0.86162
1.79670
4.62837
13.5803
4
Mortality,
Non-
Accidental
A00-R99
0.15747
0.00949
0.01874
0.04112
0.10876
0.33084
0.79395
1.73208
4.49595
13.2086
7
Mortality,
Respiratory
J00-J98
0.01250
0.00102
0.00127
0.00253
0.00570
0.02013
0.06560
0.20585
0.57827
1.42362
Mortality,
Chronic
Lung
J40-J47,
J 67
0.00052
0.00032
0.00040
0.00074
0.00186
0.01033
0.04045
0.13873
0.36008
0.68593
Mortality,
Lung
Cancer
C34
0.00000
0.00001
0.00007
0.00033
0.00282
0.02378
0.07992
0.19701
0.32952
0.31820
Mortality,
Ischemic
Heart
Disease
120-125
0.00018
0.00004
0.00039
0.00234
0.01242
0.04854
0.12174
0.25698
0.68000
2.27271
Mortality,
Cardio-
pulmonary
100-178,
J10-J18,
J40-J47,
J 67
0.01365
0.00069
0.00099
0.00214
0.00502
0.01794
0.05877
0.18453
0.51055
1.26213
Mortality,
NCD + LRI
"kit
0.08961
0.00618
0.01168
0.02751
0.08129
0.26214
0.63767
1.37694
3.44731
9.47467
Mortality,
Lower
Respiratory
Infection
A48.1, A70,
B97.4-
B97.6, J 09-
J 15.8, J16,
J20-J21,
P23.0-
P23.4, U04
0.00249
0.00618
0.01168
0.00030
0.00062
0.00112
0.00196
0.00300
0.00758
0.02693
Mortality,
Cerebro-
vascular
G45-G46.8,
160-163.9,
165-166.9,
167.0-167.3,
167.5-167.6,
168.1-168.2,
169.0-169.3
0.00097
0.00012
0.00034
0.00096
0.00314
0.00809
0.01455
0.02892
0.08553
0.20863
Mortality,
COPD
J40-J44,
J 47
0.00048
0.00005
0.00004
0.00015
0.00102
0.00904
0.03888
0.13689
0.35661
0.67457
*We estimate post-neonatal mortality (deaths after the first month) for infants because the health impact function (see
Appendix E) estimates post-neonatal mortality. Neonatal deaths were removed from the infant mortality total, and total
infant population was used as the denominator in post-neonatal mortality incidence.
**For a full list of codes for non-communicable diseases (NCD) and lower respiratory infections (LRI), see the IHME GBD Code
mapping: http://ghdx.healthdata.org/record/ihme-data/gbd-2017-cause-icd-code-mappings.
BenMAP User Manual v 0.01
D-5
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
D.1.2 Mortality Rate Projections 2015-2060
To estimate age- and county-specific mortality rates in years 2015 through 2060, we
calculated annual adjustment factors, based on a series of Census Bureau projected national
mortality rates (for all- cause mortality), to adjust the age- and county-specific mortality rates
calculated using 2012-2014 data as described above. We used the following procedure:
For each age group, we obtained the series of projected national mortality rates from 2013 to
2050 (seethe 2013 rate in Table D-3) based on Census Bureau projected life tables.
We then calculated, separately for each age group, the ratio of Census Bureau national
mortality rate in yearY (Y = 2014,2015,..., 2060) to the 2013 rate. These ratios are shown for
selected years in Table D-4.
Finally, to estimate mortality rates in yearY (Y = 2015,2020,..., 2060) that are both age-group-
specific and county-specific, we multiplied the county- and age-group-specific mortality rates
for 2012-2014 by the appropriate ratio calculated in the previous step. For example, to
estimate the projected mortality rate in 2015 among ages 18-24 in Wayne County, Ml, we
multiplied the mortality rate for ages 18-24 in Wayne County in 2012-2014 by the ratio of
Census Bureau projected national mortality rate in 2015 for ages 18-24 to Census Bureau
national mortality rate in 2013 for ages 18-24.
BenMAP User Manual v 0.01
D-6
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-5. All-Cause Mortality Rate (per 100 people per year), by Source, Year, and
Age Group
1-17
18-24
25-34
35-44
45-54
55-64
65-74
75-84
85+
Calculated
CDC 2012-2014
0.192*
0.020
0.078
0.107
0.173
0.405
0.862
1.797
4.628
13.580
Census Bureau
2013**
0.654
0.029
0.088
0.102
0.183
0.387
0.930
2.292
5.409
13.091
*The Census Bureau estimate is for all deaths in the first year of life. BenMAP uses
post-neonatal mortality (deaths after the
first month, i.e., 0.23 per 100 people) because the health impact function (see Appendix E) estimates post- neonatal
mortality. For comparison purpose, we also calculated the rate for all deaths in the first year, which is 0.684 per 100 people).
**For a detailed description of the model, the assumptions, and the data used to create Census Bureau projections, see the
working paper, "Methodology and Assumptions for the 2012 National Projections," which is available on
http://www.census.gov/population/projections/files/methodology/methodstatementl2.pdf
Table D-6. Ratio of Future Year All-Cause Mortality Rate to 2013 Estimated All-Cause
Mortality Rate, by Age Group
Year
18-24
25-34
35-44
45-54
55-64
65-74
75-84
85+
2015
0.93
0.93
0.96
1.02
0.96
0.96
1.01
1.02
1.03
1.00
2020
0.94
0.94
0.98
1.04
0.97
0.98
1.02
1.03
1.03
1.00
2025
0.85
0.81
0.74
0.80
0.75
0.77
0.85
0.91
0.93
0.97
2030
0.81
0.75
0.66
0.70
0.67
0.69
0.78
0.86
0.89
0.92
2035
0.76
0.70
0.58
0.62
0.60
0.62
0.71
0.81
0.87
0.87
2040
0.73
0.65
0.51
0.53
0.53
0.56
0.64
0.76
0.84
0.86
2045
0.70
0.60
0.45
0.46
0.46
0.50
0.58
0.71
0.80
0.86
2050
0.67
0.56
0.39
0.40
0.40
0.44
0.53
0.66
0.77
0.87
2055
0.64
0.52
0.34
0.35
0.35
0.39
0.48
0.62
0.73
0.88
2060
0.61
0.48
0.30
0.30
0.31
0.34
0.43
0.58
0.70
0.87
D.1.3 Race-Stratified Mortality Incidence
To estimate race-stratified and age-stratified incidence rates at the county level, we
downloaded all-cause and respiratory mortality data from 2007 to 2016 from the CDC
BenMAP User Manual v 0.01
D-7
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
WONDER mortality database (https://wonder.cdc.gov/). Race-stratified incidence rates were
calculated for the following age groups: < 1 year, 1-4 years, 5-14 years, 15-24 years, 25-34
years, 35-44 years, 45-54 years, 55-64 years, 65-74 years, 75-84 years, and 85+ years.12 To
address the frequent county-level data suppression for race-specific death counts, we
stratified the county-level data into two broad race categories, White and Non-White. In a
later step, we stratified the non-White incidence rates by race (Black, Asian, Native American)
using the relative magnitudes of incidence values by race at the regional level, described in
more detail below.
We followed the methods outlined in section D.l.l with one notable difference in
methodology; we included an intermediate spatial scale between county and state for
imputation purposes. We designated urban and rural counties within each state using CDC
WONDER and, where possible, imputed missing data using the state-urban and state-rural
classifications before relying on broader statewide data. We followed methods for dealing
with suppressed and unreliable data at each spatial scale as described in section D.l.l.
A pooled non-White incidence rate inherently underestimates the mortality riskforsome race
groups and overestimates mortality risk for others. To estimate county-level mortality rates
by individual race (Black, Asian, Native American), we applied regional race-specific incidence
relationships to the county-level pooled non-White incidence rates. We calculated a weighted
average of race-specific incidence rates using regional incidence rates for each
region/age/race group normalized to one reference population (the Asian race group) and
county population proportions based on race-specific county populations from CDC WONDER
where available. In cases of population suppression across two or more races per county, we
replaced all three race-specific population proportions derived from CDC WONDER with
population proportions derived from 2010 Census data in BenMAP.
12 Infant mortality dates for race- and ethnicity-stratified datasets do not currently exclude neonatal deaths.
BenMAP User Manual v 0.01 December 2022
D-8
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
D.1.4 Ethnicity-Stratified Mortality Incidence
To estimate ethnicity-stratified and age-stratified incidence rates at the county level, we
downloaded all-cause and respiratory mortality data from 2007 to 2016 from the CDC
WONDER mortality database (https://wonder.cdc.gov/). Ethnicity-stratified incidence rates
were calculated for the following age groups: < 1 year, 1-4 years, 5-14 years, 15-24 years, 25-34
years, 35-44 years, 45-54 years, 55-64 years, 65-74 years, 75-84 years, and 85+ years. We
stratified county-level data by two groups, Hispanic and non-Hispanic, and did not stratify
further by race due to suppression constraints. We followed the methods outlined in section
D.l.l to deal with suppressed and unreliable data. We also included an intermediate spatial
scale between county and state designating urban and rural counties for imputation
purposes, described in detail in section D.1.3.
D.2 Hospitalizations
Hospitalization rates were calculated using data from the Healthcare Cost and Utilization
Project (HCUP). HCUP is a family of health care databases developed through a Federal-State-
Industry partnership and sponsored by the Agency for Healthcare Research and Quality
(AHRQ). HCUP products include the State Inpatient Databases (SID), the State Emergency
Department Databases (SEDD), the Nationwide Inpatient Sample (NIS), and the Nationwide
Emergency Department Sample (NEDS). HCUP databases can be obtained from the following
data services:
HCUP Central Distributor: Many of the HCUP databases are available for purchase through
the HCUP Central Distributor. The databases include detailed information for individual
discharges, such as primary diagnosis (in ICD-9 codes), patient's age and residence county.
HCUP categorizes hospital admissions in various ways. Hospitalization admissions are
reported as emergency (admitted from the emergency department), urgent (admitted from
BenMAP User Manual v 0.01
D-9
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
another hospital), elective (admitted from another health facility, including long-term care),
newborn (admitted for delivery), trauma (not used by all states), or other/missing/invalid.
While a substantial subset of the ISA-identified literature evaluating respiratory
hospitalizations restricted analyses to emergency hospital admissions (EHAs), all hospital
admission baseline incidence data within BenMAP reflects total hospital admissions due to
time constraints limitingthe ability to stratify incidence by admission type. In general, the
vast majority of respiratory and cardiovascular hospitaizations appear to be emergency or
urgent admissions. As such, the total hospital admissions rates in BenMAP should largely
align with analogous EHA rates (albeit biased upward due to the small share of
hospitalizations that are elective).
HCUP State Partners: Some HCUP participating states do not release their data to the
Central Distributor; however, the data may be obtained through contacting the State
Partners. South Carolina provided county-level data.
HCUPnet: This is a free, on-line query system based on data from HCUP. It provides access to
summary statistics at the state, regional and national levels.
Figure D-l shows the level of hospitalization data (e.g, discharge-level or state-level) for each
state. Note that for some states neither discharge-level, county-level nor state-level data
were available. In such cases we used regional statistics from HCUPnet to estimate
hospitalization rates for those states. The data year for states using HCUPnet data is 2014. For
discharge-level data, the data year for most states is 2014; however, some states provided
data for 2011 (CA, MS); 2012 (ME); and 2013 (AR, MA, MD, NV, SD, UT). We assume
hospitalization rates are reasonably constant from 2011-2014 and consider all as 2014 rates.
BenMAP User Manual v 0.01
D-10
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Figure D-l. Hospitalization Data from HCUP
More information about HCUP can be found at http://www.hcup-us.abrq.gov/
The procedures for calculating hospitalization rates are summarized as follows:
For states with discharge-level data:
• We calculated age-, health effect-, and county-specific hospitalization counts. South
Carolina was the only state that, while not providing discharge-level data, did provide
county-level data for each age group-effect combination.
• The above calculation excluded hospitalizations with missing patient age or county
FIPS, which may lead to underestimation of rates. Therefore, we scaled up the
previously calculated age-, effect-, and county-specific counts using an adjustment
factor obtained as follows:
BenMAP User Manual v 0.01
D-ll
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
o We first counted the number of discharges for a specific health effect in the
state including those discharges with missing age or county FIPS.
o We then counted the number of discharges for the health effect in the state
excluding those records with missing age or county FIPS.
o The adjustment factor is the ratio of the two counts.
• For California and West Virginia, patient county was unavailable for all observations.
For these two states, we used hospital county in place of patient county.
• For health outcomes deemed acute (acute myocardial infarction; cerebrovascular
events; stroke; pneumonia; lower respiratory infection; acute cases of asthma), we
distributed patients within the hospital state in cases where the patient resided out of
state. We assume that everyone admitted to the hospital in a given state developed
that acute condition while in that state.
• We calculated hospitalization rates for each county by dividing the adjusted county-
level hospitalization counts by the Census estimated county-level population for the
corresponding year (2011 - 2014). Following CDC Wonder, we treated rates as
"unreliable" when the hospitalization count was less than 20, using the same
procedure we used for mortality rates (see Section D.l.l).
For states with summarized state statistics (from HCUPnet) we calculated the state-, age-,
effect- specific hospitalization rates and applied them to each county in the state. We used
the previously described procedure to adjust the "unreliable" rates.
For states without discharge-level or state-level data:
• We obtained the effect-specific hospitalization counts in each region from
HCUPnet/NIS (we refer to this count for the ith effect in the jth region as "TOTALij")
• For those states in the jth region that do have discharge-level or state-level data, we
summed the hospital admissions by effect (we refer to this count for the ith effect in
the jth region as "SUB ij").
BenMAP User Manual v 0.01
D-12
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
• We then estimated the hospitalization count for states without discharge or state data
forthe ith effect in the jth region as TOTALij - SUB ij. Note that while this count is
effect- and region- specific, it is not age-specific. We obtained the distribution of
hospital admission counts across age groups based on the Central Distributor data
and assumed the same distribution forthe HCUPnet hospitalizations. We then applied
this distribution to the estimated hospital counts (i.e., TOTALij - SUB ij) to obtain
effect-, region-, and age-specific counts.
• Using the corresponding age- and region-specific populations in BenMAP from Woods
and Poole (2015), we calculated age-specific hospitalization rates for the ith effect in
the jth region and applied them to those counties in the region that didn't have
discharge-level or state-level data.
The health effects in hospitalization studies are defined using different combinations of ICD
codes. Ratherthan generating a unique baseline incidence rate for each ICD code
combination, for the purposes of this analysis, we identified a core group of hospitalization
rates from the studies and applied the appropriate combinations of these rates in the health
impact functions:
• congestive heart failure (ICD-9 428)
• dysrhythmia (ICD-9 427)
• heart rhythm disturbances (ICD-9 426-427)
• acute myocardial infarction (ICD-9 410)
• ischemic heart disease -1 (ICD-9 410-414)
• ischemic heart disease - 2 (ICD-9 410-414,429)
• ischemic heart disease (less myocardial infarction) (ICD-9 411-414)
• all cardiovascular (ICD-9 390-429)
• all cardiovascular (less myocardial infarctions) (ICD-9 390-409,411-429)
• cardiovascular, cerebrovascular and peripheral vascular diseases (ICD-9 410-414,429,
426-427,428,430-438,440-449)
• all cardiac outcomes (ICD-9 390-459)
BenMAP User Manual v 0.01
D-13
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
• cerebrovascular events (ICD-9 430-438)
• stroke (ICD-9 431-437)
• peripheral vascular disease -1 (ICD-9 440-448)
• peripheral vascular disease-2 (ICD-9 440-449)
• all respiratory (ICD-9 460-519)
• respiratory illness -1 (ICD-9 466,480-486,490-493)
• respiratory illness -2 (ICD-9 464-466,480-487,490-492)
• chronic lung disease (ICD-9 490-496)
• chronic lung disease (less asthma) (ICD-9 490-492,494-496)
• chronic lung disease (less asthma) -2 (ICD-9 490-492,494,496)
• chronic lung disease (less asthma) -3 (ICD-9 490-492)
• chronic lung disease (less asthma) -4 (ICD-9 491,492,494,496)
• pneumonia (ICD-9 480-486)
• asthma (ICD-9 493)
• lower respiratory infection (ICD-9 466.1,466.0,480-487,490,510-511)
• respiratory - 1 (ICD-9 491,492,493,496)
• respiratory - 2 (ICD-9 464-466, 480-487,490-492,493)
• alzheimer's disease (ICD-9 331.0)
• parkinson's disease (ICD-9 332)
In addition to the hospitalization effects above, we developed a set of county level baseline
incidence for one EHA effect, All Respiratory (see Section E.7.8 for epidemiological
description). We generated the EHA rates by applying the HCUPnet national ratio of All
Respiratory hospitalizations originating from the emergency department (77%) to the county
level incidence rates developed from the discharge and state-level data.
For each C-R function, we selected the baseline rate or combination of rates that most closely
matches to the study health effect definition. For studies that define chronic lung disease as
ICD 490- 492,494-496, we subtracted the incidence rate for asthma (ICD 493) from the chronic
BenMAP User Manual v 0.01 December 2022
D-14
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
lung disease rate (ICD 490-496). In some cases, the baseline rate will not match exactly to the
health effect definition in the study. For example, Burnett et al. (2001) studied the following
respiratory conditions in infants <2 years of age: ICD 464.4,466,480-486,493. For this C-R
function we apply an aggregate of the following rates: ICD 464,466,480-487,493. Although
they do not match exactly, we assume that relationship observed between the pollutant and
study-defined health effect is applicable for the additional codes. Table D-5 presents a
summary of the national hospitalization rates for 2014 from HCUP.
Table D-7. Hospitalization Rates (per 100 people per year), by Health Effect and Age
Hospitalization
Category
Age
0-1
2-17
18-24
25-34
35-44
45-54
55-64
65-74
75-84
85+
Respiratory
All Respiratory
460-
519
2.387
0.363
0.166
0.212
0.340
0.737
1.297
2.292
4.151
6.343
Pneumonia
480-
486
0.477
0.101
0.039
0.063
0.103
0.196
0.336
0.640
1.426
2.660
Chronic Lung
Disease
490-
496
0.226
0.151
0.041
0.056
0.105
0.281
0.496
0.837
1.276
1.306
Asthma
493
0.217
0.147
0.036
0.048
0.076
0.123
0.136
0.157
0.218
0.243
Cardiovascular
All Cardiovascular
390-
429
0.044
0.017
0.061
0.138
0.377
0.914
1.747
3.131
5.886
8.832
Acute Myocardial
Infarction,
Nonfatal
410
0.000
0.000
0.002
0.010
0.068
0.202
0.380
0.575
0.921
1.332
Ischemic Heart
Disease
410-
414
0.000
0.000
0.002
0.014
0.105
0.350
0.689
1.090
1.570
1.734
Dysrhythmia
427
0.016
0.005
0.014
0.025
0.057
0.145
0.319
0.684
1.357
1.917
Congestive Heart
Failure
428
0.010
0.001
0.005
0.021
0.061
0.165
0.344
0.700
1.727
3.513
Stroke
431-
437
0.009
0.003
0.007
0.021
0.070
0.199
0.417
0.816
1.639
2.488
Neurological
Alzheimer's
Disease
331.0
0.000
0.000
0.00
0.00
0.00
0.0004
0.0035
0.027
0.129
0.248
Parkinson's
Disease
332
0.000
0.000
0.00011
0.0037
0.020
0.025
BenMAP User Manual v 0.01
D-15
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
D.3 Nonfatal Heart Attacks
The relationship between short-term particulate matter exposure and heart attacks was
quantified in a case-crossover analysis by Peters et al. (2001). The study population was
selected from heart attack survivors in a medical clinic. Therefore, the applicable population
to apply to the C-R function is all individuals surviving a heart attack in a given year. Several
data sources are available to estimate the number of heart attacks per year. For example,
several cohort studies have reported estimates of heart attack incidence rates in the specific
populations under study. However, these rates depend on the specific characteristics of the
populations under study and may not be the best data to extrapolate nationally. The
American Heart Association reports approximately 785,000 new heart attacks peryear (Roger
et al., 2012). Exclusion of heart attack deaths reported by CDC Wonder yields approximately
575,000 nonfatal cases peryear.
An alternative approach to the estimation of heart attack rates is to use data from the
Healthcare Cost and Utilization Project (HCUP), assuming that all heart attacks that are not
instantly fatal will result in a hospitalization. Details about HCUP data are described in
Section D.2. According to the 2014 HCUP data there were approximately 608,795
hospitalizations due to heart attacks (acute myocardial infarction: ICD-9 410, primary
diagnosis). We used rates based on HCUP data over estimates extrapolated from cohort
studies because the former is a national database with a larger sample size, which is intended
to provide reliable national estimates. The incidence rate calculation is also described in
Section D.2 and the incidence rates for AMI hospitalization are presented in Table D-5.
Rosamond et al. (1999) reported that approximately six percent of male and eight percent of
female hospitalized heart attack patients die within 28 days (either in or outside of the
hospital). We, therefore, applied a factor of 0.93 to the estimated number of PM-related acute
myocardial infarctions to exclude the number of cases that result in death within the first
month. Note that we did not adjust for fatal AM Is in the incidence rate estimation, due to the
BenMAP User Manual v 0.01 December 2022
D-16
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
way that the epidemiological studies are designed. Those studies consider total admissions
for AMIs, which includes individuals living at the time the studies were conducted. Therefore,
we use the definition of AMI that matches the definition in the epidemiological studies.
D.4 Emergency Department Visits
The data source for emergency department/room (ED or ER) visits is also HCUP, i.e., SID,
SEDD, and NEDS. And the types of data providers are also the same as those described in
Section D.2. Figure D-2 shows the emergency department data in each state.
Figure D-2. Emergency Department Data from HCUP
The calculation of ER visit rates is also similar to the calculation of hospitalization rates,
except for the following differences:
BenMAP User Manual v 0.01
D-17
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
The SEDD databases include only those ER visits that ended with discharge. To identify the ER
visits that ended in hospitalization, we used a variable called "admission source" in the SID
databases. Admission source identified as "emergency room" indicates that the hospital
admission came from the ER - i.e., the ER visit ended in hospitalization. For each combination
of age group, health effect, and county, we summed the ER visits that ended with discharge
and those that resulted in hospitalization.
The data year varies across the states from 2011 to 2014; we assumed that ER visit rates are
reasonably constant across these three years and consider them as 2014 rates.
Instead of using HCUPnet/NIS in the last step as described in Section D.2., we used
HCUPnet/NEDS to calculate ER visit rates for states without discharge level or state level data.
Table D-6 presents the estimated asthma emergency room rates by health effect and age
group.
BenMAP User Manual v 0.01
D-18
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-8. Emergency Department Visit Rates (per 100 people per year) by Health
Effect and Age Group
Emergency
Department
Category
Age 0-
17
18-24
25-34
35-44
45-54
55-64
65-74
75-84
85+
Asthma
493
0.959
0.601
0.556
0.538
0.552
0.408
0.331
0.368
0.350
491-
493,
460-
466,
477.0-
Respiratory
477.9,
480-
486,
496,
786.07,
786.09
6.069
3.214
2.837
2.332
2.447
2.418
2.908
4.382
5.651
410-
414,
427-
428,
433-
Cardiovascular
437,
440.0-
440.9,
443-
445,
451-453
0.030
0.107
0.212
0.496
1.151
2.023
3.451
6.726
11.028
All Cardiac
Outcomes
390-459
0.067
0.314
0.568
1.105
2.021
3.086
4.921
9.345
14.596
D.5 School Loss Days
Epidemiological studies have examined the relationship between air pollution and a variety
of measures of school absence. These measures include: school loss days for all causes,
illness- related, and respiratory illness-related. We have two sources of information. The first
is the National Center for Education Statistics, which provided an estimate of all-cause school
loss days, and the other is the National Health Interview Survey (Adams et al., 1999, Table 47),
which has data on different categories of acute school loss days. Table D-7 presents the
estimated school loss day rates. Further detail is provided below on these rates.
BenMAP User Manual v 0.01
D-19
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-9. School Loss Day Rates (per student per year)
Type
Northeast
Midwest
South
West
Respiratory illness-related
absences
1.3
1.7
1.1
2.2
Illness-related absences
2.4
2.6
2.6
3.7
All-cause
9.9
9.9
9.9
9.9
* We based illness-related school loss day rates on data from the 1996 NHIS and an estimate of 180 school days per year. This
excludes school loss days due to injuries. We based the all-cause school loss day rate on data from the National Center for
Education Statistics.
All-Cause School Loss Day Rate
Based on data from the U.S. Department of Education (1996, Table 42-1), the National Center
for Education Statistics estimates that for the 1993-1994 school year, 5.5 percent of students
are absent from school on a given day. This estimate is comparable to study-specific
estimates from Chen et al. (2000) and Ransom and Pope (1992), which ranged from 4.5 to 5.1
percent.
Illness-Related School Loss Day Rate
The National Health Interview Survey (NHIS) has regional estimates of school loss days due to
a variety of acute conditions (Adams et al., 1999). NHIS is a nationwide sample-based survey
of the health of the noninstitutionalized, civilian population, conducted by NCHS. The survey
collects data on acute conditions, prevalence of chronic conditions, episodes of injury,
activity limitations, and self-reported health status. However, it does not provide an estimate
of all-cause school loss days.
In estimating illness-related school loss days, we started with school loss days due to acute
problems (Adams et al., 1999, Table 47) and subtracted lost days due to injuries, in orderto
match the definition of the study used in the C-R function to estimate illness-related school
absences (Gilliland et al., 2001). We then divided by 180 school days per to estimate illness-
related school absence rates perschool day. Similarly, when estimating respiratory illness-
BenMAP User Manual v 0.01
D-20
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
related school loss days, we use data from Adams et al. (1999, Table 47). Note that we
estimated 180 school days in a year to calculate respiratory illness-related school absence
rates per year.
D.6 Asthma-Related Health Effects
Several studies have examined the impact of air pollution on asthma development or
exacerbation. Many of the baseline incidence rates used in the health impact functions are
based on study-specific estimates. The baseline rates for the various health effects are
described below and summarized in Table D-9. The prevalence of asthma is summarized in
Table D-10.
BenMAP User Manual v 0.01
D-21
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-10. Asthma-Related Health Effects Incidence Rates
Health Effect
Age
Parameter
Rate
Source
New Onset Asthma
0-4
Incidence
0.0234
New Onset Asthma
5-11
Incidence
0.0111
New Onset Asthma
12-17
Incidence
0.0044
New Onset Asthma
18-34
Incidence
0.0040
Winer etal. (2012,
New Onset Asthma, Women
35-44
Incidence
0.0051
Table 1 &Table 2)
New Onset Asthma, Women
45-54
Incidence
0.0046
New Onset Asthma, Women
55-64
Incidence
0.0059
New Onset Asthma, Women
65+
Incidence
0.0039
Asthma Exacerbation, Shortness
of Breath, African American
8-13
Prevalence
7.40%
Asthma Exacerbation, Wheeze,
African American
8-13
Prevalence
17.30%
Ostroetal. (2001, p.
202)
Asthma Exacerbation, Cough,
African American
8-13
Prevalence
14.50%
Asthma Symptoms, Shortness of
Breath
5-12
Prevalence
18.50%
Asthma Symptoms, Wheeze
5-12
Prevalence
19.40%
Lewis et al. (2013, p.
Asthma Symptoms, Cough
5-12
Prevalence
30.10%
51)
Asthma Symptoms, Chest
Tightness
5-12
Prevalence
12.70%
Asthma Symptoms, Albuterol Use
6-13
Incidence
2.2
Rabinovitch et al.
(2006, Table 1)
Upper Respiratory Symptoms
(URS)
9-11
Incidence
124.79
Pope et al. (1991,
Table 2)
D.6.1 New Onset Asthma
The annual rate of new asthma onset is estimated from Winer et al. (2012, Table 1 and Table
2). Winer et al., 2012 identify newly diagnosed asthma from the 2006-2008 Asthma Call-Back
Survey (ACBS) and Behavioral Risk Factor Surveillance System (BRFSS) as individuals
BenMAP User Manual v 0.01
D-22
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
diagnosed by a doctor, or other health professional, within the 12 months prior to the
surveys.
D.6.2 Shortness of Breath
To estimate the annual rate of new shortness of breath episodes among African-American
asthmatics, ages 8-13, we used the rate reported by Ostro et al. (2001, p.202). To estimate the
annual rate of new shortness of breath episodes among asthmatic children ages 5-12, we
used the rate reported by Lewis et al. (2013, p.51).
D.6.3 Wheeze
The daily rate of new wheeze episodes among African-American asthmatics, ages 8-13, is
reported by Ostro et al. (2001, p.202) as 0.076. We multiplied this value by 100 and by 365 to
get the annual incidence rate per 100 people. To estimate the annual rate of new wheeze
episodes among asthmatic children ages 5-12, we used the rate reported by Lewis et al. (2013,
P-51).
D.6.4 Cough
The daily rate of new cough episodes among African-American asthmatics, ages 8-13, is
reported by Ostro et al. (2001, p.202) as 0.067. We multiplied this value by 100 and by 365 to
get the annual incidence rate per 100 people. To estimate the annual rate of new cough
episodes among asthmatic children ages 5-12, we used the rate reported by Lewis et al. (2013,
P-51).
BenMAP User Manual v 0.01
D-23
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
D.6.5 Albuterol Use
The average number of albuterol inhaler actuations ('puffs') per day for an asthmatic child,
age 6-13, is reported by Rabinovith et al. (2006, Table 1) as 2.2 'puffs' per child per day.
D.6.6 Upper Respiratory Symptoms
Upper Respiratory Symptoms are defined as one or more of the following: runny or stuffy
nose; wet cough; burning, aching, or red eyes. Using the incidence rates for upper respiratory
symptoms among asthmatics, published in Pope et al. (1991, Table 2), we calculated a
sample size-weighted average incidence rate.
D.6.7 Asthma Population Estimates
In studies examining the association between air pollution and the development or
exacerbation of asthma, oftentimes an estimate of the percent of the population with asthma
is required. Asthma percentages were obtained from an American Lung Association (2010b)
report summarizing data from NHIS. Table D-10 presents asthma prevalence rates used to
define asthmatic populations in the health impact functions.
BenMAP User Manual v 0.01
D-24
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-ll. Asthma Prevalence Rates Used to Estimate Asthmatic Populations
Population Group
Asthma Prevalence
Source
All Ages
7.80%
<5
6.14%
<18
9.41%
5-17
10.70%
American Lung Association (2010b, Table 7)
18-44
7.19%
45-64
7.45%
65+
7.16%
African-American,
<5
9.98%
American Lung Association (2010b, Table 9)
African-American,
5 to 17
17.76%
African-American,
<18
15.53%
American Lung Association*
^Calculated by ALA for U.S. EPA, based on NHIS data (CDC, 2008).
D.7 Other Acute and Chronic Effects
For many of the minor effect studies, baseline rates from a single study are often the only
source of information, and we assume that these rates hold for locations in the U.S. The use
of study- specific estimates are likely to increase the uncertainty around the estimate
because they are often estimated from a single location using a relatively small sample.
These health effects include: acute bronchitis, chronic bronchitis, upper respiratory
symptoms, lower respiratory symptoms. Table D-8 presents a summary of these baseline
rates.
BenMAP User Manual v 0.01
D-25
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-12. Selected Acute and Chronic Incidence (Cases / Person-Year) & Prevalence
(Percentage Population)
Health Effect
Age
Parameter
Rate
Source
Acute Bronchitis
8-12
Incidence
0.043
American Lung
Association
(2002b, Table 11)
Chronic
Bronchitis
27+
Incidence
0.00378
Abbey etal.
(1993, Table 3)
18+
4.37%
American Lung
Association
(2010a, Table 4).
The rate numbers
may be slightly
Chronic
Bronchitis
18-44
3.15%
45-64
Prevalence
5.49%
different from
those in Table 4
because we
received more
current estimates
form ALA.
65+
5.63%
Lower
Respiratory
Symptoms (LRS)
7-14
Incidence
0.483
Schwartz etal.
(1994, Table 2)
Minor Restricted
Activity Days
(MRAD)
18-64
Incidence
7.8
Ostro and
Rothschild (1989,
p. 243)
18-64
2.172
Adams et a I.
Work Loss Day
18-24
Incidence
1.971
(1999, Table) U.S.
Bureau of the
Census(1997,
No.22)
(WLD)
25-44
2.475
45-64
1.796
NOTE: The incidence rate is the number of cases per person peryear. Prevalence refers to the fraction of people that have a
particular illness during a particular time period.
D.7.1 Acute Bronchitis
The annual rate of acute bronchitis for children ages 5 to 17 was obtained from the American
Lung Association (2002b, Table 11). The authors reported an annual incidence rate per person
of 0.043, derived from the 1996 National Health Interview Survey.
BenMAP User Manual v 0.01
D-26
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
D.7.2 Chronic Bronchitis Incidence Rate
The annual incidence rate for chronic bronchitis13 is estimated from data reported by Abbey
et al. (1993, Table 3). The rate is calculated by taking the number of new cases (234), dividing
by the number of individuals in the sample (3,310), dividing by the ten years covered in the
sample, and then multiplying by one minus the reversal rate (estimated to be 46.6% based on
Abbey etal. (1995a, Table 1).
Age-specific incidence rates are not available. Abbey et al. (1995a, Table 1) did report the
incidences by three age groups (25-54,55-74, and 75+) for "cough type" and "sputum type"
bronchitis. However, they did not report an overall incidence rate for bronchitis by age-group.
Since, the cough and sputum types of bronchitis overlap to an unknown extent, we did not
attempt to generate age-specific incidence rates for the over-all rate of bronchitis.
D.7.3 Chronic Bronchitis Prevalence Rate
We obtained the annual prevalence rate for chronic bronchitis from the American Lung
Association (2010a, Table 4). Based on an analysis of 2008 National Health Interview Survey
data, they estimated a rate of 0.0437 for persons 18 and older; they also reported the
following prevalence rates for people in the age groups 18-44,45-64, and 65+: 0.0315,0.0549,
and 0.0563, respectively.
D.7.4 Lower Respiratory Symptoms
Lower respiratory symptoms (LRS) are defined as two or more of the following: cough, chest
pain, phlegm, wheeze. The proposed yearly incidence rate for 100 people, 43.8, is based on
the percentiles in Schwartz et al. (Schwartz et al., 1994, Table 2). The authors did not report
13 Please note that this health effect is not regularly considered in U.S. EPA analyses (July 2018).
BenMAP User Manual v 0.01 December 2022
D-27
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
the mean incidence rate, but rather reported various percentiles from the incidence rate
distribution. The percentiles and associated per person per day values are 10th = 0 percent,
25th = 0 percent, 50th = 0 percent, 75th = 0.29 percent, and 90th = 0.34 percent. The most
conservative estimate consistent with the data are to assume the incidence per person per
day is zero up to the 75th percentile, a constant 0.29 percent between the 75th and 90th
percentiles, and a constant 0.34 percent between the 90th and 100th percentiles.
Alternatively, assuming a linearslope between the 50th and 75th, 75th and 90th, and 90th to
100th percentiles, the estimated mean incidence rate per person per day is 0.12 percent. (For
example, the 62.5th percentile would have an estimated incidence rate per person per day of
0.145 percent.) We used the latter approach in this analysis.
D.7.5 Minor Restricted Activity Days (MRAD)
Ostro and Rothschild (1989, p. 243) provide an estimate of the annual incidence rate of MRADs
per person of 7.8.
D.7.6 Work Loss Days
The yearly work-loss-day incidence rate per 100 people is based on estimates from the 1996
National Health Interview Survey (Adams et al., 1999, Table 41). They reported a total annual
work loss days of 352 million for individuals ages 18 to 65. The total population of individuals
of this age group in 1996 (162 million) was obtained from (U.S. Bureau of the Census, 1997,
No. 22). The average annual rate of work loss days per individual is 2.17. Using a similar
approach, we calculated work-loss-day rates for ages 18-24,25-44, and 45-64, respectively.
D.8 Other Health Effect Incidence
Baseline incidence estimates for health effect occurrences other than a hospitalization or
emergency department visit are described below, listed in alphabetical order.
BenMAP User Manual v 0.01 December 2022
D-28
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
D.8.1 Allergic Rhinitis
Prevalence rates of hay fever/rhinitis are presented by Parker et al. (2009). Parker et al.
investigate the associations between long-term ozone exposure and respiratory allergies in
children ages 3 to 17 years old. The authors use prevalence data from the NHIS household
interview survey and define allergic rhinitis as children with reported hay fever, respiratory
allergy, or both within the 12 months prior to the survey. Of the eligible population (72,279),
19.2% of respondents experience allergic rhinitis symptoms within the year prior to the
survey, therefore, the national prevalence rate of hay allergic rhinitis is 0.192.
D.8.2 Lung Cancer
The baseline incidence rates for non-fatal lung cancer were calculated using the existent
baseline incidence rate for lung cancer mortality in combination with the five-year lung
cancer survival rate from NCI (2015). We first used the five-year lung cancer survival rate to
calculate the total incidence of lung cancer (both fatal and non-fatal) from the baseline
mortality rate using the following formula: baseline mortality rate / (1 - five-year survival
rate). We then calculated the incidence of non-fatal lung cancer as the difference between
total lung cancer incidence and fatal lung cancer incidence (NCI, 2015). Table D-ll presents
the baseline incidence of lung cancer mortality, the SEER five-year survival rate, the
estimated total lung cancer incidence, and the estimated non-fatal lung cancer incidence rate
by age group.
BenMAP User Manual v 0.01
D-29
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
Table D-13. Lung Cancer Incidence Rates
Age Group
Annual Lung
Cancer Mortality
Incidence [A]
Five-Yea r Survival
Rate [B]
Total Lung Cancer
Incidence
[C] = [A] / (1 - [B])
Non-fatal Lung
Cancer Incidence
[D] = [C] - [A]
25-34
0.0000033
34.6%
0.0000050
0.00000175
35-44
0.0000282
34.6%
0.0000431
0.00001492
45-54
0.0002378
22.1%
0.0003053
0.00006746
55-64
0.0007922
20.8%
0.0010003
0.00020805
65-74
0.00019701
21.0%
0.0002494
0.00005237
75-84
0.0032952
14.9%
0.0038722
0.00057695
85+
0.0031820
14.9%
0.0037391
0.00055713
D.8.3 Out of Hospital Cardiac Arrest
The baseline incidence of cardiac arrests occurring outside of the hospital (OHCA) is
estimated using the incidence and survival rates reported by Daya et al. (2015). Daya et al.
(2015) utilize Resuscitation Outcomes Consortium data to calculate the incidence per 100,000
of OHCA and the survival rate broken down into four age categories, 0 to 17,18 to 39,40 to 64,
and 65+. We combined the age-specific incidence and survival rates to calculate the baseline
incidence for non-fatal OHCA (Table D-12).
Table D-14. Out of Hospital Cardiac Arrest Incidence and Survival Rates
Age
Annual incidence per
100,000 people
Survival Rate
Annual non-fatal
incidence per 100
people
0-17
10.1
8.4%
0.008
18-39
33.5
9.8%
0.033
40-64
137.3
14.9%
0.205
65+
553.5
8.8%
0.487
BenMAP User Manual v 0.01
D-30
December 2022
-------
Appendix D: U.S. Health Incidence & Prevalence Data in BenMAP
D.8.4 Stroke
We developed non-fatal stroke baseline incidence rates using similar data as was utilized to
develop the non-fatal OHCA incidence rates. Yao et al. (2019) provide the annual incidence of
stroke in 2013 for individuals over 65 years old, as well as the survival rate, broken down by
race, gender, and stroke type (hemorrhagic or ischemic). We combined the incidence and
survival rates to calculate the rate of non-fatal stroke by gender and race. We then calculated
the overall annual baseline incidence rate of stroke in all individuals overthe age of 65 by
calculating a weighted averaged from the stratified. This resulted in a rate of 0.004 strokes
per person per year. Table D-13 presents the stratified incidence and survival rates.
Table D-15. Stroke Incidence and Survival Rates
Characteristic
Annual incidence per
100,000 people
Survival Rate
Non-fatal Incidence per 100 people
peryear
Weight
(Study
Ischemic
Hemorrhagic
Ischemic
Hemorrhagic
Ischemic
Hemorrhagic
Total
Population)
Black Men
551
93
92%
73%
0.00507
0.00068
0.00575
6,155
White Men
407
75
88%
60%
0.00358
0.00045
0.00403
54,079
Black
Women
641
94
91%
69%
0.00583
0.00065
0.00648
9,819
White
Women
466
77
85%
56%
0.00396
0.00043
0.00439
78,839
Weighted Average
0.00446
BenMAP User Manual v 0.01
D-31
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
Appendix E. Core Particulate Matter Health
Impact Functions in BenMAP
In this Appendix, we present the core PM-related health impact functions in BenMAP, i.e., the
functions that, as of the current release, U.S. EPA routinely uses in its regulatory analyses.
Each sub-section has a table with a brief description of the health impact function and the
underlying parameters. Following each table, we present a brief summary of each of the
studies and any items that are unique to the study.
Note that Appendix C mathematically derives the standard types of health impact functions
encountered in the epidemiological literature, such as, log-linear, logistic and linear, so we
simply note here the type of functional form. Appendix D presents a description of the
sources for the incidence and prevalence data used in each health impact function.
E.i Long-term Mortality
There are two types of exposure to PM that may result in premature mortality. Short-term
exposure may result in excess mortality on the same day or within a few days of exposure.
Long-term exposure over, say, a year or more, may result in annual mortality in excess of
what it would be if PM levels were generally lower, although the excess mortality that occurs
will not necessarily be associated with any particular episode of elevated air pollution levels.
In other words, long-term exposure may capture a facet of the association between PM and
mortality that is not captured by short-term exposure. Table E-l lists the long-term mortality
health impact functions.
BenMAP User Manual v 0.01
E-l
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
Table E-16. Core Health Impact Functions for Particulate Matter and Long-Term
Mortality
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Mortality,
All Cause
Wu etal.
2020
Nationwide
65-
99
Annual
0.006391
0.000383
Log-
linear
Mortality,
All Cause
Pope et
al.
2019
Nationwide
18-
99
Annual
0.011333
0.001602
Log-
linear
Mortality,
All Cause
Di etal.
2017
Nationwide
65-
99
o3
Annual
0.007046
0.000095
Log-
linear
Mortality,
All Cause
Turneret
al.
2016
Nationwide
30-
99
o3
Annual
0.005827
0.000963
Log-
linear
Mortality,
All Cause
Woodruff
etal.
2008
Nationwide
0-0
Annual
0.005603
0.004539
Logistic
BenMAP User Manual v 0.01
E-2
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.l.l Wuetal. (2020)
Wu et al. (2020) evaluated the relationship between long-term PM2.5 exposure and all-cause
mortality in more than 68.5 million Medicare enrollees (over the age of 64), using Medicare
claims data from 2000-2016 representing over 573 million person-years of follow-up and over
27 million deaths. This cohort included over 20% of the U.S. population and was, at the time
of publishing, the largest air pollution study cohort to date. The authors modeled PM2.5
exposure at a l-km2 grid resolution using a hybrid ensemble-based prediction model that
combined three machine learning models and relied on satellite data, land-use information,
weather variables, chemical transport model simulation outputs, and monitor data. Wu et al.
(2020) fit five different statistical models: a Cox proportional hazards model, a Poisson
regression model, and three causal inference approaches (GPS estimation, GPS matching,
and GPS weighting). All five statistical approaches provided consistent results; we report the
results of the Cox proportional hazards model here. The authors adjusted for numerous
individual-level and community-level confounders, and sensitivity analyses suggest that the
results are robust to unmeasured confounding bias.
All-Cause Mortality
In a single-pollutant model, the coefficient and standard errorfor PM2.5 are estimated from
the hazard ratio (1.066) and 95% confidence interval (1.058-1.074) associated with a change in
annual mean PM2.5 exposure of 10.0 ug/m3 (Wu et al. 2020, Table S3, Main analysis, 2000-
2016 Cohort, Cox PH).
BenMAP User Manual v 0.01
E-3
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.1.2 Popeetal. (2019)
Pope et al. (2019) examined the relationship between long-term PM2.5 exposure and all-
cause mortality in a cohort of 1,599,329 U.S. adults (aged 18-84 years) who were interviewed
in the National Health Interview Surveys (NHIS) between 1986 and 2014 and linked to the
National Death Index (NDI) through 2015. The authors also constructed a subcohort of
635,539 adults from the full cohort for whom body mass index (BMI) and smoking status data
were available. They employed a hybrid modeling technique to estimate annual-average
PM2.5 concentrations derived from regulatory monitoring data and constructed in a universal
kriging framework using geographic variables including land use, population, and satellite
estimates. Pope et al. (2019) assigned annual-average PM2.5 exposure from 1999-2015 to
each individual by census tract and utilized complex (accounting for NHIS's sample design)
and simple Cox proportional hazards models forthe full cohort and the subcohort. We report
the results of the complex model forthe subcohort, which controls for individual-level
covariates including age, sex, race-ethnicity, inflation-adjusted income, education level,
marital status, rural versus urban, region, survey year, BMI, and smoking status.
All-Cause Mortality
In a single-pollutant model, the coefficient and standard errorfor PM2.5 are estimated from
the hazard ratio (1.12) and 95% confidence interval (1.08-1.15) associated with a change in
annual mean PM2.5 exposure of 10.0 ug/m3 (Pope et al. 2019, Table 2, Subcohort).
BenMAP User Manual v 0.01
E-4
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.1.3 Di etal. (2017)
Di et al. (2017) evaluated the relationship between long-term PM2.5exposure and all-cause
mortality in nearly 61 million U.S. Medicare enrollees (over the age of 64) through 460 million
person-years of follow-up and roughly 22 million observed deaths. This cohort comprised
approximately 15% of the total U.S. population, included people living in rural areas, and is
one of the largest cohort studies published to date. The authors modeled PM2.5 exposure
across the contiguous U.S. using a hybrid methodology that included land use regression,
satellite data, and monitor data, and resolved estimations to 1 x 1-kilometer areas. Di et al.
(2017) used Cox proportional-hazards models with a generalized estimating equation.
Adjustment for potential confounding by the co-pollutant 03 was performed, which slightly
attenuated the relationship between PM2.5 and mortality. The authors also performed
statistical testing of the potential for non-linear effects and concluded that the data
supported a nearly-linear concentration-response relationship with no signal of a threshold
down to at least 5 |ig/m3.
All-Cause Mortality
In a two-pollutant model, the coefficient and standard error for PM2.5 are estimated from the
hazard ratio (1.073) and 95% confidence interval of (1.071-1.075) associated with a change in
annual mean PM2.5exposure of 10.0 |ig/m3 (Di et al., 2017, Table 2 Main Analysis, Cox PH with
GEE).
BenMAP User Manual v 0.01
E-5
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.1.4 Turner etal. (2016)
Turner et al. (2016) examined the relationship between long-term PM2.5 exposure (1982-2004)
and mortality (all-cause, cause-specific) in American Cancer Society Cancer Prevention Study-
II participants (aged 30-99 years). Estimated PM2.5concentrations were obtained using
monthly PM2.5 monitor data (1999-2008) and a national-level hybrid land use regression (LUR)
and Bayesian maximum entropy (BME) interpolation model. Turner et al. (2016) utilized
random-effects Cox proportional hazard models adjusted a priori for individual, socio-
demographic, and ecological variables. In addition to adjusting for individual-level and
ecological covariates, Turner et al. (2016) also controlled for occupational PM2.5 exposure and
adjusted for the potential co-pollutants 03 and nitrogen dioxide.
All-Cause Mortality
In a multi-pollutant model, the coefficient and standard error for PM2.5 are estimated from the
hazard ratio (1.06) and 95% confidence interval of (1.04-1.08) associated with a change of
10.0 |ig/m3 in the mean PM2.5 exposure level from 1999-2004 (Turner et al., 2016, Table E10
HBM PM2.5, MP model, controlling for HBM 031982-2004).
BenMAP User Manual v 0.01
E-6
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.1.5 Woodruff etal. (2008)
Woodruff et al. (2008) examined the relationship between long-term exposure to fine PM2.5 air
pollution and post-neonatal infant mortality in 3,583,495 births from 96 counties containing
>249,999 residents across the U.S. between 1999-2002 using data from the National Center for
Health Statistics (NCHS). They linked average PM2.5 monitoring data over the first two months
of life with 6,639 post neonatal deaths, using logistic regression that incorporated generalized
estimating equations (GEE) to estimate the odds ratios for all-cause and cause-specific post-
neonatal mortality by exposure to air pollution. The study population experienced a median
PM2.5 concentration of 14.8 |ig/m3, with 25% of the population experiencing concentrations
below 12 |ig/m3 and above 18.8 |ig/m3. The study included an evaluation of the
appropriateness of a linear form from analysis based on quartiles of exposure and identified
the linearform as a reasonable assumption.
All-Cause Mortality
In a single-pollutant model, the coefficient and standard error for PM2.5 are estimated from
the odds ratio (1.04) and 95% confidence interval of (0.98-1.11) associated with a change of 7
|ig/m3 in the mean PM2.5 exposure level during the first two months of life (Woodruff et al.,
2008, Table 4 PM2.5Single-pollutant model, all causes).
BenMAP User Manual v 0.01
E-7
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2 Chronic/Severe Illness
Table E-2 below summarizes the health impact functions used to estimate the relationship
between PM2.5 and chronic health effects. We present a brief summary of each of the studies
below.
Table E-17. Core Health Impact Functions for Particulate Matter and Chronic Illness
Effect
Author
Year
Location
Age
Metric
Beta
Std Err
Form
Notes
Cardiac Arrest
Ensoretal.
2013
Houston, TX
18-
99
D24HourMean
0.006376
0.002823
Logistic
Cardiac Arrest
Rosenthal et
al.
2008
Indianapolis,
Indiana
0-
99
D24HourMean
0.00198
0.005018
Logistic
Cardiac Arrest
Silverman et
al.
2010
New York City
0-
99
D24HourMean
0.003922
0.00222
Logistic
Lung Cancer
Gharibvand et
al.
2017
Nationwide
U.S. and 5
Canadian
provinces
30-
99
Mean
0.037844
0.013121
Log-
linear
Alzheimer's
Disease
Kioumourtzogl
ou et al
2016
50
Northeastern
U.S. cities
65-
99
Mean
0.139762
0.017753
Log-
linear
Parkinson's
Disease
Kioumourtzogl
ou et al
2016
50
Northeastern
U.S. cities
65-
99
Mean
0.076961
0.018905
Log-
linear
Stroke
Kloog et al.
2012
New England
65-
99
Mean
0.00343
0.001265
Log-
linear
Acute Myocardial
Infarction,
Nonfatal
Peters et al.
2001
Boston, MA
18-
99
D24HourMean
0.024121
0.009285
Logistic
Acute Myocardial
Infarction,
Nonfatal
Pope et al.
2006
Greater Salt
Lake City, UT
0-
99
D24HourMean
0.0048
0.0019
Logistic
Index Ml and
unstable
angina
Acute Myocardial
Infarction,
Nonfatal
Sullivan et al.
2005
King County,
WA
0-
99
D24HourMean
0.0019
0.0022
Logistic
Acute Myocardial
Infarction,
Nonfatal
Zanobetti and
Schwartz
2006
Greater
Boston, MA
0-
99
D24HourMean
0.0053
0.0022
Logistic
Age range
adjusted
Admissions
through ER
visits only.
Acute Myocardial
Infarction,
Nonfatal
Zanobetti et al.
2009
26 U.S. Comm
0-
99
D24HourMean
0.0022
0.0006
Log-
linear
Age range
adjusted. All
Seasons.
Acute Myocardial
Infarction,
Nonfatal
Wei et al.
2019
Nationwide
65-
99
D24HourMean
0.0011
0.0002
Logistic
BenMAP User Manual v 0.01
E-8
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.1 Ensoretal. (2013)
Ensor et al. (2013) studied the association between short-term ambient air pollution (PM2.5
and 03) exposure and out-of-hospital cardiac arrest (OHCA). Ensor et al. (2013) gathered
medical and demographic data for all ages from an Emergency Medical Services database in
Houston, Texas between 2004 and 2011. Authors assessed the medical data and defined out-
of-hospital cardiac arrest as emergency medical services performing chest compressions.
Authors collected ambient air pollution and weather data from Texas Commission of
Environmental Quality monitors and calculated hourly and daily averages for PM2.5 and 03.
The authors used a time-stratified case crossover analysis and conditional logistic regression.
Out-of-Hospital Cardiac Arrest
In a single-pollutant model, the coefficient and standard error are estimated from a reported
excess risk of OHCA of 3.9 percent (95% CI: 0.5 -7.4) for a 6 |ig/m3 increase in the averaged
daily mean PM2.5 concentration 0- and 1-days prior to onset (Ensor et al. 2013, Table 4).
BenMAP User Manual v 0.01
E-9
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.2 Gharibvand etal. (2017)
Gharibvand et al. (2017) evaluated whether positive associations exist between PM2.5
exposure and incidence of lung cancer in non-smokers among the Adventist Health and Smog
Study-2 (AHSMOG-2), a group of health-conscious individuals of which 81% are never
smokers. Authors collected ambient air pollution data (PM2.5 and 03) from the US EPA Air
Quality system over two years (January 2000-December 2001). Three a priori factors were
added to the models as covariates: time spent outdoors, residence length, and moving
distance during follow-up. Authors modeled the association between PM2.5 exposure and
incidence of lung cancer using a Cox proportional hazards regression, with attained age as
the time variable. The authors conducted both a single and a two-pollutant (PM2.5and 03)
analyses. The study concluded that each 10 |ig/m3 increase in ambient PM2.5concentrations
was positively associated with increased lung cancer risks within the single-pollutant and
two-pollutant multivariable models with 03.
Incidence, Lung Cancer
In a two-pollutant multivariable model with 03 (including a priori covariates), the coefficient
and standard error were estimated from a hazard ratio of 1.46 (95% CI: 1.13-1.89) for each 10
|ig/m3 increase in mean monthly ambient PM2.5Concentrations (Gharibvand et al. 2016, Table
3).
BenMAP User Manual v 0.01
E-10
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.3 Kioumourtzoglou et al. (2016)
Kioumourtzoglou et al. (2016) evaluated the potential impact of long-term PM2.5exposure on
first hospital admission for dementia, Alzheimer's, or Parkinson's diseases among Medicare
beneficiaries (>= 65 years old) in 50 cities in the northeastern U.S. (Connecticut, Delaware,
Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania,
Rhode Island, Vermont, and Washington, D.C.). Authors retrieved medical data from the
Center for Medicaid and Medicare from the years 1999-2010. The study followed enrollees as a
cohort, which included annual follow-up records identifying the first hospital admissions for
dementia (ICD-9 290), Alzheimer's (ICD-9 331.0), Parkinson's (ICD-9 332), and other
cardiovascular comorbidities. With respect to Alzheimer's disease, the study evaluated
9,817,806 Medicare enrollees and included 266,725 cause-specific hospital admissions
indicating disease onset. With respect to Parkinson's disease, the study evaluated 9,817,806
Medicare enrollees and included 119,425 cause-specific hospital admissions indicating
disease onset. Annual average PM2.5 concentrations were estimated for each city using data
from the U.S. EPA Air Quality System database. Kioumourtzoglou et al. (2016) fit a time-
varying Cox proportional hazards model for each city, using the city-wide annual PM2.5
concentrations as the time-varying exposure of interest and a linear term for the calendar
year. This eliminated the impact of PM2.5 variation by city and any PM2.5trends within cities.
The model adjusted for cardiovascular comorbidities, and incorporated a counting process
extension which created an observation for each year of follow-up per person. The results
were then pooled across individuals and cities.
Incidence, Alzheimer's Disease (ICD-9 331.0)
In a single-pollutant model, the coefficient and standard error were estimated from a hazard
ratio of 1.15 (95% CI: 1.11-1.19) for a 1 |ig/m3 increase in the average annual PM2.5
concentrations (Kioumourtzoglou et al. 2016, Table 1).
BenMAP User Manual v 0.01
E-ll
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
Incidence, Parkinson's Disease (ICD-9332)
In a single-pollutant model, the coefficient and standard error were estimated from a hazard
ratio of 1.08 (95% CI: 1.04-1.12) for a 1 |ig/m3 increase in the average annual PM2.s
concentrations (Kioumourtzoglou et al. 2016, Table 1).
BenMAP User Manual v 0.01
E-12
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.4 Kloogetal. (2012)
Kloog et al. (2012) analyzed the effects of long- and short-term PM2.5exposure on hospital
admissions due to strokes with a new PM2.5exposure model in New England (Connecticut,
Maine, Massachusetts, New Hampshire, Rhode Island, and Vermont) from 2000 to 2006. We
use this health effect as a surrogate for PM2.5-attributable stroke incidence. Authors collected
medical data from 67,678 adults aged 65 to 99 in the U.S. Medicare program database from
2000 to 2006. They defined all respiratory, cardiovascular disease, stroke, and diabetes based
on emergency department visits and primary discharge diagnosis records. Authors used a
hybrid exposure technique comprised of daily PM2.5concentration data from aerosol optical
depth (AOD) measurements and ambient air monitors from the U.S. EPA and Interagency
Monitoring of Protected Visual Improvements (IMPROVE). Authors also obtained land use
regressions, meteorological data (National Climatic Data Center), and socioeconomic data
(U.S. Census Bureau) matched to zip codes in orderto perform land use Poisson regressions.
Incidence, Stroke (ICD Codes 430-436)
In a single-pollutant model for patients over the age of 65, the coefficient and standard error
were estimated from the percent change (3.49%) and 95% confidence interval (0.09-5.18%)
for a 10 |ig/m3 increase in the 7-year mean PM2.5concentrations (Kloog et al., 2012, Table 3).
BenMAP User Manual v 0.01
E-13
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.5 Peters et al. (2001)
Peters et al. (2001) studied the relationship between increased particulate air pollution and
onset of heart attacks in the Boston area from 1995 to 1996. The authors used air quality data
for PMio, PMio-2.5, PM2.5, "black carbon", 03, CO, N02, and S02 in a case-crossover analysis. For
each subject, the case period was matched to three control periods, each 24 hours apart. In
univariate analyses, the authors observed a positive association between heart attack
occurrence and PM2.5 levels hours before and days before onset. The authors estimated
multivariate conditional logistic models including two-hour and twenty-four hour pollutant
concentrations for each pollutant. They found significant and independent associations
between heart attack occurrence and both two-hour and twenty-four hour PM2.5
concentrations before onset. Significant associations were observed for PM10 as well. None of
the other particle measures or gaseous pollutants was significantly associated with acute
myocardial infarction for the two hour or twenty-four hour period before onset.
The patient population for this study was selected from health centers across the United
States. The mean age of participants was 62 years old, with 21% of the study population
under the age of 50. In order to capture the full magnitude of heart attack occurrence
potentially associated with air pollution and because age was not listed as an inclusion
criteria for sample selection, we apply an age range of 18 and over in the C-R function.
According to the National Hospital Discharge Survey, there were no hospitalizations for heart
attacks among children <15 years of age in 1999 and only 5.5% of all hospitalizations
occurred in 15-44 year olds (Popovic, 2001, Table 10).
Acute Myocardial Infarction, Nonfatal
The coefficient and standard error are calculated from an odds ratio of 1.62 (95% CI 1.13-2.34)
for a 20 |ig/m3 increase in twenty-four hour average PM2.5 (Peters et al., 2001, Table 4, p. 2813).
BenMAP User Manual v 0.01
E-14
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
Incidence Rate: We use the county-specific daily AMI hospitalization rate (ICD-9 code 410) for
the population of individuals aged 18 years and older as the estimate for the incidence rate of
nonfatal heart attack, assuming all heart attacks that are not instantly fatal will result in a
hospitalization. We did not adjust for fatal AMIs in the incidence rate estimation, due to the
way that the epidemiological studies are designed. Those studies consider total admissions
for AMIs, which includes individuals living at the time the studies were conducted. Therefore,
we use the definition of AMI that matches the definition in the epidemiological studies.
Population: Population of ages 18 and older
Adjustment: As some fraction of the admitted individuals die in the hospital, we apply a
survival rate of 93% in calculating the avoided cases of AMI in orderto avoid double counting
(once in the calculation of AMI cases and once in the calculation of PM-related mortality).
BenMAP User Manual v 0.01
E-15
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.6 Popeetal. (2006)
Pope et al. (2006) evaluated the association between short-term exposure to PM2.5 and acute
ischemic heart disease events, including acute nonfatal myocardial infarction, all acute
coronary events, and subsequent myocardial infarctions in individuals living in greater Salt
Lake City, Utah. In a case-crossover study, these ischemic events were assessed in relation to
a 10 |ig/m3 increase in PM2.5. The researchers determined that a 10 |ig/m3 increase in PM2.5
resulted in a 4.5% increase (95% CI: 1.1-8.0) in unstable angina and myocardial infarction.
Acute Myocardial Infarction, Nonfatal
In a single-pollutant model the coefficient and standard error were estimated from the
percent increase (4.81%) and 95% confidence interval (95% CI: 0.98-8.79) for a 10 |ig/m3
increase in daily 24-hour mean PM2.5 (Pope et al., 2006, Table 3).
Incidence Rate: AMI hospital admission rate for all ages. See the incidence rate discussion
under Peters et al. (2001) in Section E.2.5.
Population: All ages
Adjustment: See the adjustment description in Section E.2.5.
BenMAP User Manual v 0.01
E-16
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.7 Rosenthal et al. (2008)
Rosenthal et al. (2008) examined the effects of short-term PM2.5 exposure on out-of-hospital
cardiac arrest incidence and whether these effects were connected to demographic data or
presence of heart rhythm. Additionally, Rosenthal et al. (2008) compared exposure time and
measurement method on the effects of short-term PM2.5 exposure and out-of-hospital cardiac
arrest incidence. Authors obtained medical data from the Wishard Ambulance Service, a local
emergency medical service in Indianapolis, Indiana, from July 2,2002 to July 7,2006. The
study defined out-of-hospital cardiac arrest using the same criteria as Ensor et al. (2013) and
Silverman et al. (2010). Authors collected daily and hourly PM2.5concentrations from two City
of Indianapolis monitoring sites and using two separate methods: the Federal Reference
Method (FRM) for 24-hour filter samples, and a Federal Equivalence Method (FEM). The
authors used a case crossover analysis with conditional logistic regressions in order to study
the effects of short-term PM2.5 exposure on out-of-hospital cardiac arrest incidence. Rosenthal
et al. (2008) found a positive but statistically insignificant association between non-dead on
arrival (DOA) out-of-hospital cardiac arrest cases and ambient PM2.5 concentrations. Although
they also noted a statistically significant positive association when restricted to witnessed,
non-DOA out-of-hospital cardiac arrest cases, that subgroup is less applicable to the available
baseline incidence rate of non-DOA out-of-hospital cardiac arrest cases.
Out-of-Hospital Cardiac Arrest
In a single-pollutant model of all non-DOA OHCA cases, the coefficient and standard error
were estimated from a hazard ratio of 1.02 (95% CI: 0.92-1.12) for each 10 |ig/m3 increase in
daily mean PM2.5Concentrations, lagged by 0-1 days (Rosenthal et al. 2008, Table 5).
BenMAP User Manual v 0.01
E-17
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.8 Silverman et al. (2010)
Silverman et al. (2010) investigated the link between short-term ambient air pollution
exposure (PM2-5, N02, S02,03, and CO) and out-of-hospital cardiac arrest in New York City
between 2002 and 2006. Authors obtained medical data from the Emergency Medical Services
of the New York City Fire Department for 8,216 subjects aged 0 to 99, average age 65.6 with
slightly more men than women. Authors collected air pollution and weather data from the US
EPA's Air Quality System monitors within a 20-mile radius of New York City and averaged over
24-hour periods. Authors conducted time series and case crossover analyses with 0- and 1-
day lagged air pollution levels and by season.
Out-of-Hospital Cardiac Arrest
In a single-pollutant case-crossover model, the coefficient and standard error were estimated
from a relative risk of 1.04 (95% CI: 0.99-1.08) for a 10 |ig/m3 increase in the averaged daily
mean PM2i5 concentration 0- and 1-day prior to onset (Silverman et al. 2010, Table 4).
BenMAP User Manual v 0.01
E-18
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.9 Sullivan etal. (2005)
Sullivan et al. (2005) studied the relationship between onset time of acute myocardial
infarction and the preceding hourly PM2.5 concentrations in 5,793 confirmed cased of
myocardial infarction through King County, Washington. In this case-crossover study from
1988-1994, air pollution exposure levels averaged 1 hour, 2 hours, 4 hours, and 24 hours
before onset of myocardial infarction were compared to a set of time-stratified referent
exposures from the same day of the week in the month of the case event. The authors
estimated that an associated risk of 1.01 (95% CI: 0.98-1.05) for myocardial infarction onset
could be attributed to a 10 |ig/m3 increase in PM 2.5 the hour before the Ml onset. No
increased risk was found in all cases with preexisting cardiac diseases with an odds ratio of
1.05 (95% CI: 0.95-1.16). Furthermore, stratification for hypertension, diabetes, and smoking
status did not modify the association between PM2.5 and onset of myocardial infarction.
Acute Myocardial Infarction, Nonfatal
In a single-pollutant model the coefficient and standard error were estimated from the odds
ratio (1.02) and 95% confidence interval (95% CI: 0.98-1.07) for a 10 |ig/m3 increase in daily 24-
hour mean PM2.5 lagged 1 day (Sullivan et al., 2005, Table 3).
Incidence Rate: AMI hospital admission rate for all ages. See the incidence rate discussion
under Peters et al. (2001) in Section E.2.5.
Population: All ages
Adjustment: See the adjustment description in Section E.2.5.
BenMAP User Manual v 0.01
E-19
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.10 Zanobetti and Schwartz (2006)
Zanobetti and Schwartz (2006) analyzed hospital admissions through emergency department
for myocardial infarction (ICD-9 code 410) and pneumonia (ICD-9 codes 480-487) for
associations with fine particulate air pollution, ozone, black carbon, nitrogen dioxide, PM not
from traffic, and CO in the greater Boston area from 1995-1999. The authors used a case-
crossover analysis with control days matched on temperature. Significant associations were
detected for N02 with a 12.7% increase 95% CI: 5.8-18.0), PM2.s with an 8.6% increase (95% CI:
1.2-15.4), and black carbon with an 8.3% increase (95% CI: 0.2-15.8) in emergency myocardial
infarction hospitalizations. Similarly, significant associations were identified for PM2.5 with a
6.5% increase (95% CI: 1.1-11.4) and CO with a 5.5% increase (95% CI: 1.1-9.5) in pneumonia
hospitalizations.
Acute Myocardial Infarction, Nonfatal
The study looked at hospital admissions of AMI through the ER. Under the assumption that all
heart attacks will end in hospitalization, we considerthe health effect as heart attack events
to be consistent with other studies. In a single-pollutant model, the coefficient and standard
error are estimated from the percent change in risk (8.65%) and 95% confidence interval (95%
CI: 1.22-15.38%) for a 16.32 |ig/m3 increase in daily 24-hour mean PM2.5 for an average of the 0-
and 1-day lag (Zanobetti A. and Schwartz, 2006, Table 4).
Incidence Rate: AMI hospital admission rate for all ages. See the incidence rate discussion
under Peters et al. (2001) in Section E.2.5.
Population: All ages. Note that although Zanobetti and Schwartz (2006) reports results for
the 65-99 year old age range, for comparability to other studies, we apply the results to all
ages. Since the vast majority of AMIs occur among population 65-99, over-counting may not
be an issue when applying the risk coefficient to all ages.
Adjustment: See the adjustment description in Section E.2.5.
BenMAP User Manual v 0.01
E-20
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.11 Zanobetti et al. (2009)
Zanobetti et al. (2009) examined the relationship between daily PM2.5 levels and emergency
hospital admissions for cardiovascular causes, myocardial infarction, congestive heart
failure, respiratory disease, and diabetes among 26 U.S. communities from 2000-2003. The
authors used meta-regression to examine how this association was modified by season- and
community-specific PM2.5 composition while controlling for seasonal temperature as a
substitute for ventilation. Overall, the authors found that PM2.s mass higher in Ni, As, and Cras
well as Br and organic carbon significantly increased its effects on hospital admissions. For a
10 |ig/m3 increase in 2-day averaged PM2.5, a 1.89% (95% CI: 1.34-2.45) increase in
cardiovascular disease admissions, a 2.25% (95% CI: 1.10-3.42) increase in myocardial
infarction admissions, a 1.85% (95% CI: 1.19-2.51) increase in congestive heart failure
admissions, a 2.74% (95% CI: 1.30-4.20) increase in diabetes admissions, and a 2.07% (95% CI:
1.20-2.95) increase in respiratory admissions were observed. The relationship between PM2.5
and cardiovascular admissions was significantly modified when the mass of PM2.5 was high in
Br, Cr, Ni, and sodium ions, while mass high in As, Cr, Mn, organic carbon, Ni and sodium ions
modified the myocardial infarction relationship and mass high in As, orgarnic carbon, and
sulfate ions modified the diabetes admission rates.
Acute Myocardial Infarction, Nonfatal
The study looked at hospital admissions of AMI through ER. Under the assumption that all
heart attacks will end in hospitalization, we considerthe health effect as heart attack events
to be consistent with other studies. In a single-pollutant model the coefficient and standard
error are estimated from the percent change in risk (2.25%) and 95% confidence interval (95%
CI: 1.10-3.42) for a 10 |ig/m3 increase in 2-day averaged PM2.5 (Zanobetti et al., 2009, Table 3).
Incidence Rate: AMI hospital admission rate for all ages. See the incidence rate discussion
under Peters et al. (2001) in Section E.2.5.
BenMAP User Manual v 0.01
E-21
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
Population: All ages. Note that although Zanobetti et al. (2009) reports results for the 65-99
year old age range, for comparability to other studies, we apply the results to all ages. Since
the vast majority of AMIs occurs among population 65-99, over-counting may not be an issue
when applying the risk coefficient to all ages.
Adjustment: See the adjustment description in Section E.2.5.
BenMAP User Manual v 0.01
E-22
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.2.12 Wei etal. (2019)
Wei et al. (2019) evaluated the relationship between short-term PM2.5 exposure and hospital
admissions for 214 mutually exclusive disease groups, including acute myocardial infarction,
in a time-stratified, case-crossover analysis of over 95 million Medicare inpatient hospital
claims from 2000-2012. The authors estimated daily PM2.5 levels at a l-km2 grid cell level
using a satellite based, neural network model that was calibrated using monitor data and
assigned 0-1 day lagged PM2.5 exposure to each participant by zip code of residence. For
each disease group, Wei et al. (2019) created a case crossover dataset that controlled for
individual level and zip code level variables, day of the week, seasonality, and long-term time
trends. They used conditional logistic regression models to estimate associations between
PM2.5 exposure and risk of hospital admission and found positive associations for numerous
rarely studied and numerous well-studied disease groups.
Acute Myocardial Infarction, Nonfatal
In a single-pollutant model, the coefficient and standard error are estimated from a reported
relative increase in risk (0.11%) and 95% confidence interval (0.07%-0.16%) associated with a
1 ug/m3 increase in 0-1 day lagged PM2.5 exposure (Wei et al. 2019, Figure 3, CCS 100 Acute
Myocardial Infarction).
BenMAP User Manual v 0.01
E-23
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.3 Hospitalizations
Table E-3 summarizes the health impacts functions used to estimate the relationship
between PM2.5 and hospital admissions. Below, we present a brief summary of each of the
studies and any items that are unique to the study.
Table E-18. Core Health Impact Functions for Particulate Matter and Hospital
Admissions
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
Cardiovascular
Bell et
al.
2015
213 U.S.
Counties
65-
99
D24HourMean
0.000648
0.000089
Log-
linear
Respiratory
Bell et
al.
2015
213 U.S.
Counties
65-
99
D24HourMean
0.00025
0.000120
Log-
linear
Respiratory
Ostro et
al.
2009
6
California
counties
0-
18
D24HourMean
0.002752
0.000772
Log-
linear
BenMAP User Manual v 0.01
E-24
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.3.1 Bell etal. (2015)
Bell et al. (2015) investigated the effects of short-term fine particulate matter (PM2.5) exposure
on respiratory health (ICD-9 464-466,480-487,490-492,493) and cardiovascular health (ICD-9
410, omitting 410.x2; 410-414; 426-427; 428; 429; 430-438; and 440-448) in older adults (>64
years). Authors acquired data for 213 U.S. counties (1999-2010) from the Medicare Claims
Inpatient Files for U.S. residents >65 years of age. Authors chose variables including sex, age,
county of residence, and cause of hospital admission, as determined by ICD-9 codes. Authors
collected PM2.5 exposure data from county population-based ambient monitors from the US
EPA Air Quality System and averaged for county and day. Data were present for 56.5% of
study days due to thesamplingschedule of the monitors. Bell etal. (2015) utilized Bayesian
hierarchal modeling to examine the links between PM2.5 and hospital admissions, running
separate models to generate risk models for time lags (0-2 days) and season for any
estimated variation in health effects.
Hospital Admissions, Cardio-, Cerebro- and Peripheral Vascular Disease (ICD Codes 410,
omitting 410.x2; 410-414; 426-427; 428; 429; 430-438; and440-448)
In a single-pollutant model, the coefficient and standard error are estimated from a percent
increase in risk of 0.65% (95% CI: 0.48-0.83%) for an increase of 10 |ig/m3 in same-day daily
mean PM2.5Concentrations (Bell et al. 2015, Table 1).
Hospital Admissions, Respiratory-2 (ICD Codes 490-492,464-466,480-487,493)
In a single-pollutant model, the coefficient and standard error are estimated from a percent
increase in risk of 0.25% (95% CI: 0.01-0.48%) for an increase of 10 |ig/m3 in same-day daily
mean PM2.5Concentrations (Bell et al. 2015, Table 1).
BenMAP User Manual v 0.01
E-25
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.3.2 Ostro et al. (2009)
Ostro et al. (2009) estimated the association between ambient PM2.5, EC, organic carbon (OC),
NOs, and S04 on hospital admissions for respiratory diseases in children ages 5 to 19. The
study used the California Office of Statewide Health Planning and Development, Healthcare
Quality and Analysis Division hospitalization data from six California counties for the 2000 to
2003 study period. Ostro et al. (2009) classified hospital admissions into: all respiratory
disease (ICD-9 codes 460-519), asthma (ICD-9 code 493), acute bronchitis (ICD-9 code 466),
and pneumonia (ICD-9 codes 480-486). They aggregated the hospital admission data to the
county level to create a daily time series of admissions for each county. Authors took air
quality measurements from the California Air Resources Board, which captured speciated 24-
hour average pollutant measurements using a filter-based Met One Speciation Air Sampling
System. Meteorological measurements for average daily temperature and relative humidity
came from the California Air Resources Board or the California Irrigation Management
Information System. Authors analyzed data using a Poisson regression with time, day of the
week, temperature, relative humidity, and pollutant as explanatory variables. Ostro et al.
(2009) controlled for seasonality and time dependent effects by including a natural spline
smoother for the daily time trend and meteorology.
Hospital Admissions, All Respiratory (ICD Codes 460-519)
In a single-pollutant model, the coefficient and standard error are estimated from an excess
risk of 4.1% (95% CI: 1.8-6.4%) for a 14.6 |ig/m3 increase in the daily mean PM2.5
concentrations, lagged by 3 days (Ostro et al. 2009, Table 2, pg. 477).
BenMAP User Manual v 0.01
E-26
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.4 Emergency Room Visits
Table E-4 summarizes the health impacts functions used to estimate the relationship
between PM2.5 and emergency room visits. Below, we present a brief summary of each ofthe
studies and any items that are unique to the study.
Table E-19. Core Health Impact Functions for Particulate Matter and Emergency Room
Visits
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
Cardiovascular
Ostro
et a I.
2016
8
California
cities
0-99
D24Hour
Mean
0.0006
12
0.0004
22
Logistic
Respiratory
Krall
et a I.
2016
Atlanta,
GA
0-99
D24Hour
Mean
0.0005
45
0.0002
67
Log-
linear
Respiratory
Krall
et a I.
2016
Birmingh
am, AL
0-99
D24Hour
Mean
0.0009
68
0.0003
52
Log-
linear
Respiratory
Krall
et a I.
2016
St. Louis,
MO
0-99
D24Hour
Mean
0.0008
32
0.0003
29
Log-
linear
Respiratory
Krall
et a I.
2016
Dallas, TX
0-99
D24Hour
Mean
0.0013
53
0.0005
88
Log-
linear
BenMAP User Manual v 0.01
E-27
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.4.1 Ostroetal. (2016)
Ostro et al. (2016) investigated the association between short-term, source-specific (vehicular
emissions, biomass burning, soil, and secondary NCT3 and S04 sources) PM2.5 concentrations
and emergency department visits for respiratory and cardiovascular diseases in eight cities in
California from 2005 to 2008. Authors obtained medical and demographic data from the
Office of Statewide Health Planning and Development in California, and diagnosis was
defined with ICD-9 codes: all cardiovascular (390-459), ischemic heart disease (410-414), AMI
(410), cardiac dysrhythmia (427), and heart failure (428). Ostro et al. (2016) conducted a case
cross-over analysis, stratified by year and month, controlling for weather and day of the week
covariates. Authors used a county-level logistic regression and random-effects meta-analysis
to examine the association between source-specific PM2.5 and emergency department visits
for respiratory and cardiovascular diseases. Results indicate a positive association between
vehicle PM2.5 emissions and emergency department visits for all cardiovascular diseases.
ER Visits, All Cardiac Outcomes (ICD Codes 390-459)
In a single-pollutant model, the coefficient and standard error were estimated from the
excess risk of 0.7% (95% CI: -0.2-1.7%) for a 11.4 |ig/m3 (interquartile range) increase in daily
mean PM2.5 concentration, lagged by 2 days (Ostro et al. 2016, Table 4).
BenMAP User Manual v 0.01
E-28
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.4.2 Kralletal. (2016)
Krall et al. (2016) investigated the associations between short-term, source-specific (traffic
and coal combustion) ambient PM2.5 exposure and emergency department visits for
respiratory diseases in U.S. cities (Atlanta, GA, Birmingham, AL, St. Louis, MO, and Dallas, TX).
Authors obtained medical data from hospital electronic billings for emergency department
visits due to respiratory disease, identified using ICD-9 codes (460-465,466,477,480-486,491,
492,493,496,786.07). Authors collected PM2.5 concentrations from one ambient air monitor in
each of the four cities and gathered meteorological data from the National Climactic Data
Center. Krall et al. (2016) estimated source-specific PM2.5 using apportionment models, which
separate PM2.5 sources based on chemical composition. This model also included data on
gaseous pollutant concentrations from the Community Multiscale Air Quality (CMAQ) with
Tracers model. Kralletal. (2016) used Poisson time series regression models to analyze
associations between short-term PM2.5 exposure and emergency department visits for
respiratory diseases. They then compared source-specific PM2.5 exposures across cities to
estimate associations with the emergency department visit data. To limit confounders, the
authors adjusted models for indicator variables, meteorological variables, and long-term
trends in emergency department visits.
ER Visits, Respiratory (ICD Codes 480-486,491,492,496,460-465,466,477,493, 786.07)
In a single-pollutant model, the coefficient and standard error were estimated from a relative
risk of 1.005 (95% CI: 1.000-1.010) for Atlanta, GA; 1.009 (95% CI: 1.003-1.015) for Birmingham,
AL; 1.008 (95% CI: 1.002-1.014) for St. Louis, MO; and 1.012 (95% CI: 1.002-1.023) for Dallas, TX.
All relative risks were calculated for a 9.16 |ig/m3 increase in daily mean PM2.5 concentrations,
lagged by 0 days (Krall et al. 2016, Figure 1).
BenMAP User Manual v 0.01
E-29
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.5 Minor Effects
Table E-5 summarizes the health impacts functions used to estimate the relationship
between PM2.sand minor effects. Below, wepresenta brief summary of each of the studies
and any items that are unique to the study.
Table E-20. Core Health Impact Functions for Particulate Matter and Minor Effects
Effect
Author
Year
Location
Age
Co-Poll
Metric
Beta
Std Err
Form
Work Loss
Days
Ostro
1987
Nationwide
18-64
D24Hour
Mean
0.004600
0.000360
Log-
linear
Minor
Restricted
Activity
Days
Ostro and
Rothschild
1989
Nationwide
18-64
Ozone
D24Hour
Mean
0.007410
0.000700
Log-
linear
Hay
Fever/
Rhinitis
Parker et
al.
2009
Nationwide
3-17
Summer
O3, PM2.5-10,
no2, so2
Annual
0.025464
0.009618
Logistic
BenMAP User Manual v 0.01
E-30
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.5.1 Ostro (1987)
Ostro (1987) estimated the impact of PM2.5 on the incidence of work-loss days (WLDs),
restricted activity days (RADs), and respiratory-related RADs (RRADs) in a national sample of
the adult working population, ages 18 to 65, living in metropolitan areas. The study
population is based on the Health Interview Survey (HIS), conducted by the National Center
for Health Statistics. The annual national survey results used in this analysis were conducted
in 1976-1981. Ostro reported that two-week average PM2.5 levels were significantly linked to
work-loss days, RADs, and RRADs, however there was some year-to-year variability in the
results. Separate coefficients were developed for each year in the analysis (1976-1981); these
coefficients were pooled. The coefficient used in the concentration-response function
presented here is a weighted average of the coefficients in Ostro (1987, Table III) using the
inverse of the variance as the weight.
Work Loss Days
The coefficient used in the C-R function is a weighted average of the coefficients in Ostro
(1987, Table III) using the inverse of the variance as the weight:
Equation E-l
1981
s
/=1976
BenMAP User Manual v 0.01
December 2022
E-31
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
The standard error of the coefficient is calculated as follows, assuming that the estimated
year-specific coefficients are independent:
Equation E-2
( 1981
n
\
( 1981 n
>
£
Pi
v A
2
Zu 2
1981
= Zvar
f
\
i-1976
CJ pi
i-1976 (J
I
var
= var
1981
1
7
2
I
z—1976
rj
x y
2
\yj Pi
y
z=l976
<7 n
V
^ Pi
V
y
This eventually reduces down to:
Equation E-3
= 0.00036
Incidence Rate: daily work-loss-day incidence rate per person ages 18 to 64 = 0.00595 (U.S.
Bureau of the Census, 1997, No. 22; Adams et al., 1999, Table 41)
Population: adult population ages 18 to 64
BenMAP User Manual v 0.01
E-32
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.5.2 Ostro and Rothschild (1989)
Ostro and Rothschild (1989) estimated the impact of PM2.5 and ozone on the incidence of
minor restricted activity days (MRADs) and respiratory-related restricted activity days (RRADs)
in a national sample of the adult working population, ages 18 to 65, living in metropolitan
areas. The study population is based on the Health Interview Survey (HIS), conducted by the
National Center for Health Statistics. In publications from this ongoing survey, non-elderly
adult populations are generally reported as ages 18-64. From the study, it is not clear if the
age range stops at 65 or includes 65 year olds. We apply the C-R function to individuals ages
18-64 for consistency with other studies estimating impacts to non-elderly adult populations.
The annual national survey results used in this analysis were conducted in the period 1976-
1981. Controlling for PM2.5, two-week average ozone has highly variable association with
RRADs and MRADs. Controlling for ozone, two-week average PM2.5 was significantly linked to
both health effects in most years.
Minor Restricted Activity Days
Using the results of the two-pollutant model, we developed separate coefficients for each
year in the analysis, which were then combined for use in this analysis. The coefficient is a
weighted average of the coefficients in Ostro and Rothschild (1989, Table 4) usingthe inverse
of the variance as the weight:
Equation E-4
f 1981 pj \
2
P =
£ J
i=1976 0p.
1981 1
^/=! 976 °# )
= 0.00741.
BenMAP User Manual v 0.01
E-33
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
The standard error of the coefficient is calculated as follows, assuming that the estimated
year-specific coefficients are independent:
Equation E-5
= var
f i9si n
i=1976
1981 i
^7=1976 °$ J
= var
P,
f 1981
i=1976
7
1981
Z var
P,
2
VP,
This reduces down to:
Equation E-6
Incidence Rate: daily incidence rate for minor restricted activity days (MRAD) = 0.02137
(Ostro and Rothschild, 1989, p. 243)
Population: adult population ages 18 to 64
BenMAP User Manual v 0.01
E-34
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.5.3 Parker et al. (2009)
Parker et al. (2009) investigated the associations between long-term PM2.5 exposure and
respiratory allergies in an unrestricted population of children (aged 3-17 years) sampled from
the United States National Health Interview Survey. Authors obtained symptom data from
participant parents, who reported respiratory allergies on annual surveys. Parker et al. (2009)
placed all study participants reporting symptoms of respiratory allergies or hay fever into a
combined rhinitis group. Parker et al. (2009) then linked annual averages of S02, N02, PM2.5,
and PM2.5-10 and warm season (May to September) 03 averages to participant's addresses
through ambient air pollution and meteorological data (03, S02, N02, PM2i5, and PMi0-2.5)
collected from US EPA Air Quality System monitors. The authors adjusted their logistic
regression models for survey year, poverty-level, race/ethnicity, age, family structure,
insurance coverage, usual source of care, education of adult, urban-rural status, region, and
median county-level income.
Incidence, Hay Fever/Rhinitis
In a multi-pollutant model, the coefficient and standard error were estimated from an odds
ratio of 1.29 (95% CI: 1.07-1.56) for a 10 |ig/m3 increase in PM2i5 concentrations (Parker et al.
2009, Table 4).
BenMAP User Manual v 0.01
E-35
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.6 Asthma-Related Effects
Table E-6 summarizes the health impacts functions used to estimate the relationship
between PM2.5 and asthma exacerbation. Below, we present a brief summary of each of the
studies and any items that are unique to the study.
Table E-21. Core Health Impact Functions for Particulate Matter and Asthma-Related
Effects
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
Asthma
Symptoms,
Albuterol
Rabinovitch
et a I.
2006
Denver,
CO
6-
17
D24HourMean
0.001996
0.001477
Log-
linear
Albuterol
use
use
Separate
Asthma
Onset
Tetreault et
al.
2016
Quebec,
Canada
0-
17
Annual
0.043672
0.000885
Log-
linear
HIFs for
ages 0-4;
5-17
BenMAP User Manual v 0.01
E-36
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.6.1 Rabinovitch et al. (2006)
Rabinovitch et al. (2006) analyzed the relationship between short-term PM2.5 exposure and
asthma exacerbation in children. The study followed children, ages 6 to 13 attending the
Kunsberg School at the National Jewish Medical Research Center with diagnosed asthma for
two consecutive winters from 2001-2003. Authors gave an electronic bronchodilator
(albuterol) to the children to capture the frequency of use within a 24-hour period. The
children also responded to three questions to determine if they may have an upper
respiratory infection (URI), and urine samples were taken to measure urinary leukotriene E4
levels on select days. The authors collected hourly ambient PM2.s levels from the Colorado
Department of Health Air Pollution Control Division's Tapered Element Oscillating
Microbalance (TEOM) monitor, located 2.7 miles west of the school. Additionally, a Federal
Reference Monitor (FRM) located next to the TEOM measured 24-hour PM2.5 levels. The
authors obtained meteorological data from the Colorado Department of Health Air Pollution
Control Division and the National Climatic Data Center. A Poisson regression modeled
albuterol use as a function of the morning (12:00am to 11:00 am) maximum hourly PM2.5 level
or the morning mean hourly PM2.5 level. The model used both the TEOM and FRM data,
individually, incorporated four lag periods (0 to 2 days and 0- to 2-day average), and included
several covariates: temperature, pressure, humidity, time trend, Friday indicator, and URI
indicator. Rabinovitch et al. (2006) found that, although the PM2.5 pollution levels were well
below the National Ambient Air Quality Standards, there is a consistent association between
peak ambient PM2.5 levels and increased albuterol use in asthmatic children.
Asthma Symptoms, Albuterol use
In a single-pollutant model, the coefficient and standard error were estimated from a
percentage of use increase of 1.2% (95% CI: -0.6-2.9%) for a 6 |ig/m3 increase in averaged
daily mean PM2.5Concentration lagged by 0-, 1-, and 2-days (Rabinovitch et al. 2006, Table 4,
pg. 1099).
BenMAP User Manual v 0.01
E-37
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.6.2 Tetreault et al. (2016)
Tetreault et al. (2016) investigated the relationship between childhood asthma onset and
long-term pollution exposure (PM2.5, N02,03) in Quebec, Canada. The authors obtained data
from four medical-administrative databases collectively known as Quebec Integrated Chronic
Disease Surveillance System (QICDSS) between April 1,1996 and March 31,2011. The study
defined the onset of asthma as a hospital discharged diagnosis of asthma or two reports of
asthma from two separate physicians within a two-year period. The authors used Cox
proportional hazard models to estimate the association between asthma onset and pollution
exposure, controlling for demographics and socioeconomic status. Time-varying exposure
models assessed time-varying exposures to the three pollutants in question. Tetreault et al.
(2016) showed that childhood asthma onset may be associated with exposure to PM2.5, N02,
and 03.
As the physiology and etiology of lung development in children is similar in children 6-17, we
apply the 4-12 year age-striated effect estimate from Tetreault et al. (2016) to children ages 4-
17 (Baena-Cagnani et al., 2007, Guerra et al., 2004, Ochs et al., 2004, Sparrow et al., 1991,
Trivedi and Denton, 2019).
Incidence, Asthma
In a single-pollutant time-varying model, the coefficient and standard error were estimated
from a hazard ratio of 1.33 (95% CI: 1.31-1.34) for a 6.53 |ig/m3 (interquartile range) increase in
annual PM2.5concentration at the residential address (Tetreault et al. 2016, Table 5).
BenMAP User Manual v 0.01
E-38
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7 Sensitivity Analysis - General
Table E-7 summarizes the PM2.5 health impacts functions considered by EPA to be sensitivity
analyses. Below, we present a brief summary of each of the studies and any items that are
unique to the study.
BenMAP User Manual v 0.01
E-39
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
Table E-22. Core Health Impact Functions for Particulate Matter Sensitivity Analyses
Effect
Author
Year
Location
Age
Co-Poll
Metric
Beta
Std Err
Form
Notes
Mortality, All
Cause
Di et al.
2017
Nationwide
65-99
Annual
0.008066
0.000118
Log-
linear
Single-pollutant
model
Mortality, All
Cause
Di et al.
2017
Nationwide
65-99
03
Annual
0.005921
0.000096
Log-
linear
Nearest monitor
analysis
Mortality, All
Cause
Di et al.
2017
Nationwide
65-99
03
Annual
0.007789
0.000118
Log-
linear
Cox model with
mixed effects
Hospital
Admissions,
Respiratory
Jones et
al.
2015
New York
State
0-99
D24Hour
Mean
0.000800
0.000170
Logistic
HA, Respiratory-
1
Incidence,
Asthma
McConnell
et al.
2010
13 Southern
California
communities
4-17
Annual
0.029127
0.017732
Log-
linear
Incidence,
Asthma
Nishimura
et al.
2013
5 Urban
regions
7-21
Annual
0.029559
0.069101
Logistic
Black, Hispanic
Mortality, All
Cause
Pope et
al.
2015
Nationwide
30-99
Annual
0.006766
0.000712
Log-
linear
LURBME model
Hospital
Admissions,
All Cardiac
Outcomes
Talbottet
al.
2014
Massachusetts
0-99
03
D24Hour
Mean
0.000499
0.000355
Logistic
Hospital
Admissions,
All Cardiac
Outcomes
Talbottet
al.
2014
New Jersey
0-99
03
D24Hour
Mean
0.001094
0.000227
Logistic
Hospital
Admissions,
All Cardiac
Outcomes
Talbottet
al.
2014
New Mexico
0-99
03
D24Hour
Mean
0.001094
0.001943
Logistic
Hospital
Admissions,
All Cardiac
Outcomes
Talbottet
al.
2014
New York
0-99
03
D24Hour
Mean
0.001094
0.000151
Logistic
Hospital
Admissions,
All Cardiac
Outcomes
Talbottet
al.
2014
Florida
0-99
03
D24Hour
Mean
0.000401
0.000307
Logistic
Hospital
Admissions,
All Cardiac
Outcomes
Talbottet
al.
2014
New
Hampshire
0-99
03
D24Hour
Mean
0.001207
0.001238
Logistic
Hospital
Admissions,
All Cardiac
Outcomes
Talbottet
al.
2014
Washington
0-99
03
D24Hour
Mean
0.000904
0.000540
Logistic
Mortality, All
Cause
Turner et
al.
2016
Nationwide
30-99
Annual
0.005827
0.000963
Log-
linear
Single-pollutant
model
Emergency
Hospital
Admissions,
All Respiratory
Zanobetti
et al.
2009
26 U.S.
communities
65-99
D24Hour
Mean
0.002049
0.000437
Log-
linear
BenMAP User Manual v 0.01
E-40
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.1 Di etal. (2017)
See full study description under Di et al. (2017) in Appendix E, Section E.l.l.
Mortality, All-Cause (Single-Pollutant Model)
In a single-pollutant model, the coefficient and standard error for PM2.5 are estimated from
the hazard ratio (1.084) and 95% confidence interval of (1.081-1.086) associated with a
change in annual mean PM2.5exposure of 10.0 |ig/m3 (Di et al., 2017, Table 2 Single-pollutant
analysis).
Mortality, All-Cause (Nearest Monitor Analysis)
In a two-pollutant model, the coefficient and standard error for PM2.5 are estimated from the
hazard ratio (1.061) and 95% confidence interval of (1.059-1.063) associated with a change in
annual mean PM2.5 exposure of 10.0 |ig/m3 (Di etal., 2017, Table 2 Nearest Monitor Analysis,
Cox PH with GEE).
Mortality, All-Cause (Cox Proportional Hazards Model with Mixed Effects (COXME))
In a two-pollutant model, the coefficient and standard error for PM2.5 are estimated from the
hazard ratio (1.081) and 95% confidence interval of (1.078-1.083) associated with a change in
annual mean PM2.5 exposure of 10.0 |ig/m3 (Di etal., 2017, Table S3 Main Analysis, Cox PH with
mixed effects (COXME)).
BenMAP User Manual v 0.01
E-41
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.2 Jones et al. (2015)
Jones et al. (2015) assessed the impacts of PM2.5 and its chemical constituents (sulfate (S04),
ammonium (NH4), nitrate (N03), elemental carbon (EC), and carbon-only portion of organic
carbon aerosol) on respiratory health. The study encompassed all ages, races, and ethnicities
with a case-crossover analysis in New York state. Analysis used 24-hour average PM2.5
chemical constituent concentrations from the Community Multiscale Air Quality (CMAQ)
model, and meteorological data from the National Climactic Data Center. The authors
assessed hospital discharge data from the New York State Department of Health State
Planning and Research Cooperative System (SPARCS) through principle diagnosis
categorized by ICD-9 code (chronic bronchitis (ICD-9 491), emphysema (ICD-9 492), asthma
(ICD-9 493), and chronic airway obstruction (ICD-9 496)). Authors used a single pollutant
conditional logistic regression model to analyze the respiratory hospital admission and PM2.5
chemical constituent data overtime and by season. The authors calculated hazard ratios
(HRs) using the PHREG procedure in SAS (version 9.2) with 95% confidence intervals from the
regression models. Jones et al. (2015) found that PM2.5 and its chemical constituents showed
significant associations between total PM2.5 mass and hospital admissions in the year-round
model and for all exposure lags (0-4 days). Of all the PM2.5 chemical constituents, sulfate had
the strongest association with respiratory hospital admissions, particularly during the
summer months. Additionally, sulfate was the largest contributor to the PM2.5 total mass
(49.9%).
Hospital Admissions, Respiratory-1 (ICD-9 Codes 491,492,493,496)
In a year-round single-pollutant model, the coefficient and standard error were estimated
from a hazard ratio of 1.006 (95% CI: 1.003-1.008) for a 7.48 |ig/m3 increase in daily mean PM2.5
concentrations, lagged by 4 days. The model was adjusted for season. (Jones et al. 2015,
Figure 2).
BenMAP User Manual v 0.01
E-42
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.3 McConnelletal. (2010)
McConnell et al. (2010) examined the association between long-term traffic-related air
pollution (PM2.5, PM10, 03, and N02) exposure and incident asthma in children. The authors
collected data for three years from a cohort of 2,497 kindergarten and first-grade children
who entered the Southern California Children's Health Study without asthma or wheeze.
McConnell et al. (2010) defined new-onset asthma as physician-diagnosed asthma reported
by parents on a yearly questionnaire. While the primary focus of the study was traffic-related
air pollution from local vehicle emissions, the authors also utilized ambient air pollution
exposure data from central site monitors in each of the 13 communities in the Southern
California Children's Health Study. The authors used a multilevel Cox proportional hazards
model to estimate the association between ambient air pollution exposure and new-onset
asthma, controlling for race/ethnicity, secondhand smoke exposure, and pets in the home.
The authors concluded that traffic-related pollution exposure may contribute to an increased
risk of new-onset asthma in children.
Incidence, Asthma
In a single-pollutant model, the coefficient and standard error were estimated from a hazard
ratio of 1.66 (95% CI: 0.91-3.05) for a 17.4 |ig/m3 (range of exposure in the 13 communities)
increase in annual average PM2.5exposure (McConnell et al. 2010, Table 4).
BenMAP User Manual v 0.01
E-43
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.4 Nishimura etal. (2013)
Nishimura et al. (2013) investigated the relationship between long-term early-life pollution
exposure (PM2.5, PM10, 03, N02, and S02) and asthma onset in Latino and African American
children in five urban areas (Chicago, IL; Bronx, NY; Houston, TX; San Francisco, CA; Puerto
Rico). The authors obtained data from the Genes-environments and Admixture in Latino
Americans (GALA II) Study and the Study of African Americans, Asthma, Genes and
Environments (SAGE II). GALA II and SAGE II are case-control studies that enrolled children
with and without asthma. The studies defined case subjects as children with physician-
diagnosed asthma plus two or more symptoms of coughing, wheezing, orshortness of breath
in the two years before study enrollment while control subjects were children with no
reported history of asthma, lung disease, or chronic illness, and no reported symptoms of
coughing, wheezing, orshortness of breath in the two years before study enrollment. The
authors estimated annual average pollution exposures during the first year of life as well as
the first three years of life from self-reported residential histories by calculating inverse
distance-squared weighted averages from the four closest U.S. EPA Air Quality System
monitoringstations within 50 km. The authors first used regional- and study-specific logistic
regression models to estimate the association between asthma diagnosis and pollution
exposure, controlling for demographics and socioeconomic status and subsequently
combined the regression coefficients into a multi-region estimate using a random-effects
meta-analysis. Nishimura et al. (2013) showed that early-life air pollution exposure may
increase the risk for asthma development in later childhood for Latino and African American
cohorts.
Incidence, Asthma
In a single-pollutant model estimating PM2.5exposure during the first year of life, the
coefficient and standard error were estimated from an odds ratio of 1.03 (95% CI: 0.90-1.18)
BenMAP User Manual v 0.01
E-44
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
for a 1 |ig/m3 increase in average annual PM2.5 levels at the residential address during the first
year of life (Nishimura et al. 2013, Figure 2).
BenMAP User Manual v 0.01
E-45
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.5 Pope et al. (2015)
Pope et al. (2015) evaluated the relationship between long-term exposure to ambient PM2.5
and risk of death from CVD and cardiometabolic disease, including effect modification of the
relationship by pre-existing cardiometabolic risk factors, in the ACS Cancer Prevention Study
II cohort (ages 30+). PM2.5 exposures were estimated at home addresses based on a land use
regression model with Bayesian Maximum Entropy kriging of residuals (LURBME). Pope et al.
utilized a Cox proportional hazards model controlling for individual-level covariates which
included variables that characterized current and former smoking habits, exposure to second
-hand cigarette smoke, workplace PM2.5exposure in each subject's main lifetime occupation,
self -reported exposure to dust and fumes in the workplace, marital status, level of education,
body mass index, consumption of alcohol, and quartile ranges of dietary fat index and
quartile ranges of a dietary vegetable/fruit/fiber index. Ecological covariates included median
household income; percentage of people with <125% of poverty level income; percentage of
unemployed individuals aged >16 years; percentage of adults with <12th grade education;
and percentage of the population who were Black or Hispanic.
Mortality, All-Cause (LURBME)
In a single-pollutant model, the coefficient and standard error are estimated from the hazard
ratio (1.07) and 95% confidence intervals (95% CI: 1.06-1.09) for a 10 |ig/m3 increase in
monthly PM2.5 exposure levels averaged from 1999-2004 (Pope, et al., 2015, Table 1. Cox
model with individual-level plus ecological covariates; exposure based on LUR-BME).
BenMAP User Manual v 0.01
E-46
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.6 Talbott etal. (2014)
Talbott et al. (2014) assessed daily PM2.5 concentrations and hospitalizations for
cardiovascular disease in Florida, Massachusetts, New Hampshire, New Jersey, New Mexico,
New York, and Washington from 2001 to 2008. The authors gathered hospital discharge data
from each state's respective data stewards. Talbott et al. (2014) conducted a time-stratified
case-crossover study using hospitalization data for all cardiovascular disease (ICD-9 390-459)
and for several specific cardiovascular diseases within the ICD-9 390-459 range. Authors used
a downscaling Bayesian space-time modeling approach to combine air monitoring data and
airgridded numerical outputs from the Community Multi-Scale Air Quality Model (CMAQ) to
predict daily PM2.5 concentrations. The authors gathered meteorological data from the CDC
Wonder North America Land Data Assimilation System Daily Air Temperatures and Heat
Index. Talbott et al. (2014) used conditional logistic regression adjusted for03 (same day as
PM2.5) and maximum apparent temperature (same day as admission).
Hospital Admissions, All Cardiac Outcomes (ICD-9 Codes 390-459)
In a two-pollutant multivariable model with 03, the coefficient and standard error are
estimated from an odds ratio of 1.005 (95% CI: 0.998-1.012) for Massachusetts; 1.011 (95% CI:
1.007-1.016) for New Jersey; 1.011 (95% CI: 0.973-1.050) for New Mexico; 1.011 (95% CI: 1.008-
1.014) for New York; 0.996 (95% CI: 0.990-1.002) for Florida; 0.988 (95% CI: 0.965-1.013) for
New Hampshire; and 0.991 (95% CI: 0.981-1.002) for Washington. Each odds ratio is for a 10
|ig/m3 increase in the averaged daily mean PM2.5Concentration 0-, 1-, and 2-day lags (Talbott
etal. 2014, Table 3).
BenMAP User Manual v 0.01
E-47
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.7 Turner etal. (2016)
See full study description underTurneret al. (2016) in Appendix E, Section E.1.2.
Mortality, All-Cause (Single-Pollutant Model)
In a single-pollutant model, the coefficient and standard error for PM2.5 are estimated from
the hazard ratio (1.06) and 95% confidence interval of (1.04-1.08) associated with a change of
10.0 |ig/m3 in the mean PM2.5 exposure level from 1999-2004 (Turner et al., 2016, Table E10
HBM PM2.5,1982-2004).
BenMAP User Manual v 0.01
E-48
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.7.8 Zanobetti etal. (2009)
Zanobetti et al. (2009) examined the relationship between daily PM2.5 levels and emergency
hospital admissions for cardiovascular causes, myocardial infarction, congestive heart
failure, respiratory disease, and diabetes among 26 U.S. communities from 2000-2003. The
authors used meta-regression to examine how this association was modified by season- and
community-specific PM2.5 composition while controlling for seasonal temperature as a
substitute for ventilation. Overall, the authors found that PM2.s mass higher in Ni, As, and Cr as
well as Br and organic carbon significantly increased its effects on hospital admissions. For a
10 |ig/m3 increase in 2-day averaged PM2.5, a 1.89% (95% CI: 1.34-2.45) increase in
cardiovascular disease admissions, a 2.25% (95% CI: 1.10-3.42) increase in myocardial
infarction admissions, a 1.85% (95% CI: 1.19-2.51) increase in congestive heart failure
admissions, a 2.74% (95% CI: 1.30-4.20) increase in diabetes admissions, and a 2.07% (95% CI:
1.20-2.95) increase in respiratory admissions were observed. The relationship between PM2.5
and cardiovascular admissions was significantly modified when the mass of PM2.5 was high in
Br, Cr, Ni, and sodium ions, while mass high in As, Cr, Mn, organic carbon, Ni and sodium ions
modified the myocardial infarction relationship and mass high in As, organic carbon, and
sulfate ions modified the diabetes admission rates.
Emergency Hospital Admissions, All Respiratory (ICD-9 Codes 460-519)
In a single-pollutant model, the coefficient and standard error are estimated from the percent
change in risk (2.07%) and 95% confidence interval (1.2% - 2.95%) for a 10 |ig/m3 increase in
2-day averaged PM2.5 (Zanobetti et al. 2009, Table 3).
BenMAP User Manual v 0.01
E-49
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.8 Sensitivity Analysis - At-Risk Populations
Table E-8 summarizes the PM2.5 health impacts functions considered by EPA to be sensitivity
analyses that characterize risk experienced by certain subpopulations. Below, we present a
brief summary of each of the studies and any items that are unique to the study.
Table E-23. Core Health Impact Functions for Particulate Matter Sensitivity Analyses of
At-Risk Populations
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
ER Visits,
Asthma
Alhanti
et a I.
2016
Atlanta,
Dallas, St.
Louis
0-4
D24HourMean
0.0025
0.0019
Log-
linear
White
ER Visits,
Asthma
Alhanti
et a I.
2016
Atlanta,
Dallas, St.
Louis
0-4
D24HourMean
0.0037
0.0012
Log-
linear
Non-
White
ER Visits,
Asthma
Alhanti
et a I.
2016
Atlanta,
Dallas, St.
Louis
5-
18
D24HourMean
0.0025
0.0016
Log-
linear
White
ER Visits,
Asthma
Alhanti
et a I.
2016
Atlanta,
Dallas, St.
Louis
5-
18
D24HourMean
0.0049
0.0012
Log-
linear
Non-
White
Mortality,
All Cause
Di et
al.
2017
Nationwide
65-
99
03
Annual
0.0061
0.0001
Log-
linear
Non-
Hispanic
White
Mortality,
All Cause
Di et
al.
2017
Nationwide
65-
99
03
Annual
0.0110
0.0008
Log-
linear
Hispanic
White
Mortality,
All Cause
Di et
al.
2017
Nationwide
65-
99
03
Annual
0.0189
0.0004
Log-
linear
Black
Mortality,
All Cause
Di et
al.
2017
Nationwide
65-
99
03
Annual
0.0092
(0.0010)
Log-
linear
Asian
Mortality,
All Cause
Di et
al.
2017
Nationwide
65-
99
03
Annual
0.0095
0.0019
Log-
linear
Native
American
BenMAP User Manual v 0.01
E-50
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.8.1 Alhantietal. (2016)
Alhanti et al. (2016) examined the relationship between daily PM2.5 concentrations and
emergency room visits for asthma (ICD-9 493,786.07) among residents of all ages in Atlanta
(1993-2009), Dallas (2006-2009), and St. Louis (2001-2007). Patient-level ER visit data were
obtained from hospitals in the three cities. Daily 24-hour average PM2.5 concentrations were
estimated using data from all available monitors in the region including monitors from U.S.
EPA AQS in all three cities, as well as the South Eastern Aerosol Research and
Characterization (SEARCH) network and Assessment of the Spatial Aerosol Composition
(ASACA) network in Atlanta. The authors ran city-specific daily time-series Poisson regression
models by age group (0-4,5-18,19-39,40-64,65-99) and performed additional analysis
stratified by race (White, non-White) and sex. Models controlled for temperature, day of the
week, holidays, race, age, and sex.
Emergency Room Visits, Asthma
In single-pollutant models for ages 0-4, the coefficient and standard error are estimated from
the three-city weighted average rate ratio (1.02) and 95% confidence interval (0.99-1.05) for
White children and (1.03) and 95% confidence interval (1.01-1.05) for non-White children for a
8 |ig/m3 increase in three-day moving average PM2.5Concentrations (Alhanti et al. 2016,
Supplemental Table 4).
In single-pollutant models for ages 5-18, the coefficient and standard error are estimated
from the three-city weighted average rate ratio (1.02) and 95% confidence interval (1.00-1.05)
for White children and (1.04) and 95% confidence interval (1.02-1.06) for non-White children
for a 8 |ig/m3 increase in three-day moving average PM2.5 concentrations (Alhanti et al. 2016,
Supplemental Table 4).
BenMAP User Manual v 0.01
E-51
December 2022
-------
Appendix E: Core Particulate Matter Health Impact Functions in BenMAP
E.8.2 Dietal. (2017)
See full study descriptionunder Di et al. (2017) in Appendix E, Section E.l.l.
Mortality, All-Cause
In multi-pollutant models, the coefficient and standard error are estimated from a hazard
ratio of 1.063 (95% CI: 1.060,1.065) for White; 1.208 (95% CI: 1.199,1.217) for Black; 1.096
(95% CI: 1.075,1.117) for Asian; 1.116 (95% CI: 1.100,1.133) for Hispanic; and 1.100 (95% CI:
1.060,1.140) for Native Americans. Each odds ratio is for a 10 |ig/m3 increase in annual mean
PM2.5exposure (Di et al. 2017, Supplementary Table S3 (GEE, By Race)).
BenMAP User Manual v 0.01
E-52
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Appendix F. Core Ozone Health Impact
Functions in BenMAP
In this Appendix, we present the core health impact functions used to estimate ozone-related
adverse health effects, i.e., the functions that, as of the current release, U.S. EPA routinely
uses in its regulatory analyses. Each sub-section has a table with a brief description of each
health impact function and the underlying parameters. Following each table, we present a
brief summary of each of the studies and any items that are unique to the study.
Note that Appendix C mathematically derives the standard types of health impact functions
encountered in the epidemiological literature, such as, log-linear, logistic and linear, so we
simply note here the type of functional form. And Appendix D presents a description of the
sources for the incidence and prevalence data used in the health impact functions.
F.i Short-term Mortality
Table F-l summarizes the core health impacts functions used to estimate the relationship
between ozone and mortality. Below, we present a brief summary of each of the studies and
any items that are unique to the study.
BenMAP User Manual v 0.01
F-l
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Table F-24. Core Health Impact Functions for Ozone and Mortality*
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
Mortality,
Respiratory
Katsouyanni
etal.
2009
90 U.S.
Cities
0-99
DIHou
rMax
0.00
0727
0.0005
67
Log-
linear
Warm season.
Warm season.
8-hour max
from 1-hour
Mortality,
Respiratory
Katsouyanni
etal.
2009
90 U.S.
Cities
0-99
D8Hou
rMax
0.00
0727
0.0005
67
Log-
linear
max using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.000822.
Long-term
Mortality,
Respiratory
Turner et al.
2016
Nationwi
de
30-
99
PM2.
5,
N02
Annual
(D8Ho
urMax)
0.00
7696
0.0011
76
Log-
linear
Warm season.
Mortality,
Respiratory
Zanobetti
and
Schwartz
2008
48 U.S.
Cities
0-99
D8Hou
rMax
0.00
0827
0.0002
28
Log-
linear
D8HourMean
approximated
as D8HourMax
*Unless otherwise stated, mortality is short-term.
BenMAP User Manual v 0.01
F-2
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.l.l Katsouyanni etal. (2009)
Katsouyanni et al. (2009) used time series methods to examine the relationship between
short-term 03 exposures and mortality across the U.S for all ages. The study utilized mortality
data from the National Center for Health Statistics (www.cdc.gov/nchs) for years 1987
through 1996, excluding accidental deaths (i.e., International Classification of Diseases (ICD]-
9 800). 90 U.S. cities with population sizes varying from about 250,000 to above 9 million with
the largest populations were included. Daily number of deaths ranged from 5 to 198. All 90
cities had daily summer 03 measurements. Investigators conducted extensive simulation
studies to test 1) the choice of the smoothing method and basic functions used to estimate
the smooth function of time in the city-specific models, and 2) the number of degrees of
freedom to be used in the smooth function of time. The investigators also evaluated whether
each city should be assigned the same model specification or whether each city-specific
model should depend on city-specific characteristics. For the former, the same degrees of
freedom (ranging from 1 to 20 df/yearof data) were assigned to the smooth function of time
for every city. The range was determined by choosingthe minimum possible degrees of
freedom per year up to a maximum degrees of freedom per year that essentially removed all
variation in the data beyond time scales of one week. Also, the collective experience of the
investigators indicated that using more than 20 df/year does not substantially affect the risk
estimates. Forthe latter approach, the degrees of freedom forthe smooth function of time
were chosen separately for each city using a fit criterion, such as the Akaike Information
Criterion (AIC), or by minimizing the partial autocorrelation function (PACF) of the residuals.
Nonparametric methods underestimated the standard error of the air pollution regression
coefficient, penalized splines gave relatively small bias, and PACF in combination with
penalized splines performed relatively well in terms of bias. Therefore, the identified risk
estimate was a summer-only penalized spline estimate of respiratory mortality.
BenMAP User Manual v 0.01
F-3
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Short-term Mortality, Respiratory
In a single pollutant model, the coefficient and standard error are based on the summer-only
penalized spline estimate of respiratory mortality of 0.73% (-0.39,1.85%) per 10 ppb increase
in 03 from distributed lag days (Katsouyanni et al. 2009, Table 24: Distributed Lags; Penalized
splines; 03 Results).
The Health Impact Function was adjusted from the daily 1-hour max metric to the daily 8-
hour max metric using a ratio of 1.13 (ratio of 1-hour max to 8-hour max ozone) (Anderson
and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-4
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.1.2 Turner etal. (2016)
Turner et al. (2016) examined the relationship between long-term 03 exposure (1982-2004)
and mortality (all-cause, cause-specific) in American Cancer Society Cancer Prevention Study-
II participants (aged 30-99 years). A hierarchal Bayesian space-time model based on National
Air Monitoring Stations, State and Local Air Monitoring Stations, and Community Multi-Scale
Air Quality model data estimated daily eight-hour maximum ozone concentrations at the
participant's address. The models considered meteorological data and levels of other
ambient pollutants (PM2.5, both regional and near-source, and N02). Turner et al. (2016)
utilized Cox proportional hazard models adjusted a priori for individual, socio-demographic,
and ecological variables. Notably, the study compared annual mortality with warm-season 03
exposures, so full-year baseline incidence rates will be used with risk estimates from this
study.
Long-term Mortality, Respiratory
In a multi-pollutant model, the coefficient and standard error are based on the warm-season
specific hazard ratio of 1.08 (1.06-1.11) per 10 ppb increase in seasonal average of daily 8-
hour maximum 03 concentrations (Turner et al. 2016, Table E9: Diseases of the respiratory
system (cause of death), HBM 03 (multipollutant model data, fully adjusted HR)).
BenMAP User Manual v 0.01
F-5
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.1.3 Zanobetti and Schwartz (2008)
Zanobetti and Schwartz (2008) investigated the effects of short-term 03 exposure on mortality
(all-cause, cardiovascular, stroke, and respiratory) in an unrestricted population of children,
adults, and older adults (aged 0-99 years). Between 1998 and 2000, the authors collected
mortality data from the National Center for Health Statistic in 48 cities across the United
States. Along with eight-hour ozone concentrations and meteorological data obtained from
US EPA's Air Quality System Technology Transfer Network, the authors utilized a generalized
linear model with quasi Poisson link functions to estimate the effects of short-term ozone on
respiratory mortality. The model adjusted for season, day of the week, and temperature.
Since ozone concentrations vary between seasons, the authors decided to restrict their
analysis to ozone warm season (June - August).
Short-term Mortality, Respiratory
In a single pollutant model, the coefficient and standard error are based on the warm season
excess risk estimate of 0.83% (95% CI: 0.38-1.28%) for an increase of 10 ppb in daily 8-hour
mean 03 concentrations over a summed lag structure of zero to three days (Zanobetti and
Schwartz, 2008, Table 1).
The D8HourMean metric is approximated as D8HourMax in this function.
BenMAP User Manual v 0.01
F-6
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.2 Hospital Admissions
Table F-2 summarizes the core health impact functions used to estimate the relationship
between ozone and hospital admissions. Below, we present a brief summary of each of the
studies and any items that are unique to the study.
Table F-25. Core Health Impact Functions for Ozone and Hospital Admissions
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
All
Respiratory
Katsouyanni
eta I.
2009
14 U.S.
Cities
65-
99
DIHourMax
0.000280
0.000176
Log-
linear
Warm season
Warm season.
8-hour max
from 1-hour
All
Respiratory
Katsouyanni
eta I.
2009
14 U.S.
Cities
65-
99
D8HourMax
0.000280
0.000176
Log-
linear
max using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.000316.
BenMAP User Manual v 0.01
F-7
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.2.1 Katsouyanni et al. (2009)
Katsouyanni et al. (2009) used time series methods to examine the relationship between daily
03 concentrations and hospital admissions in North America. For U.S. benefits estimation
purposes, we focus on analyses performed using the U.S hospital admission datasets. These
datasets included 14 cities with populations between 291,000 and 5,377,000 between 1987-
1996 with city-wide daily 1-hour maximum 03 concentrations ranging from -34-60 ppb. The
authors used a first stage analysis protocol that used generalized linear models with either
penalized or natural splines to adjust for seasonality, with varying degrees of freedom. The
number of degrees of freedom were also chosen by minimizing the partial autocorrelation
function of the model's residuals. Model specification approach accounted for seasonal
patterns, weekend and vacation effects, and epistemics of respiratory disease. Data were also
analyzed to detect potential thresholds in the concentration-response relationships. The
second stage analysis used pooling approaches and assessed potential effect modification by
sociodemographic characteristic and indicators of the pollution mixture across study regions.
Hospital Admissions, All Respiratory (ICD-9 Codes 460-519)
In a two-pollutant model including PM10, the coefficient and standard error are based on the
warm season excess risk estimate of 0.28% (-0.07,0.62%) per 10 ppb increase in 03 averaged
over lags 0-1 day (Katsouyanni et al., 2009, Table 40: Average of Lags 0-1 day; Penalized
splines).
The Health Impact Function was adjusted from the daily 1-hour max metric to the daily 8-
hour max metric using a ratio of 1.13 (ratio of 1-hour max to 8-hour max ozone) (Anderson
and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-8
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.3 Emergency Room Visits
Table F-3 summarizes the core health impacts functions used to estimate the relationship
between ozone and emergency room (ER) visits. Below, we present a brief summary of each
of the studies and any items that are unique to the study.
Table F-26. Core Health Impact Functions for Ozone and Emergency Room Visits
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
Respiratory
Barry
2018
Atlanta, GA
0-
D8HourMax
0.00118
0.00040
Log-
All
etal.
99
linear
year
Respiratory
Barry
2018
Birmingham,
0-
D8HourMax
0.00118
0.00059
Log-
All
etal.
AL
99
linear
year
Respiratory
Barry
2018
Dallas, TX
0-
D8HourMax
0.00195
0.00049
Log-
All
etal.
99
linear
year
Respiratory
Barry
2018
Pittsburgh,
0-
D8HourMax
0.00118
0.00040
Log-
All
etal.
PA
99
linear
year
Respiratory
Barry
etal.
2018
St. Louis,
MO-IL
0-
99
D8HourMax
0.00079
0.00030
Log-
linear
All
year
BenMAP User Manual v 0.01
F-9
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.3.1 Barry et al. (2018)
Barry et al. (2018) investigated the effects of short-term ozone exposure on emergency
department visits for respiratory disease (ICD-9 493,786.07,460-466,477,491,492,496,480-
486,466.1,466.11,466.19) in an unrestricted population of children, adults, and older adults
(aged zero-99 years) within five cities (Atlanta, GA, Birmingham, AL, Dallas, TX, Pittsburgh, PA,
and St. Louis, MO-IL) across the United States. Authors obtained individual-level health data
from hospitals and hospital associations in each of the five cities. Models fusing air quality
monitor data with Community Multi-Scale Air Quality modeled data at 12 x 12-km grids were
used to estimate ozone exposure. Barry et al. (2018) assessed associations with short-term
ozone exposure with daily eight-hour maximum ozone concentrations. The authors
implemented Poisson log-linear models to estimate risk values with three day moving
averages.
ER Visits, Respiratory (ICD-9 Codes 493, 786.07,460-466,477,491,492,496,480-486,466.1,
466.11, 466.19)
In single-pollutant models, the coefficient and standard error are based on rate ratios of 1.03
(95% CI: 1.01-1.05) in Atlanta, GA, 1.03 (95% CI: 1.00-1.06) in Birmingham, AL, 1.05 (95% CI:
1.02-1.07) in Dallas TX, 1.03 (95% CI: 1.01-1.05) in Pittsburgh, PA, and 1.02 (95% CI: 1.01-1.04)
in St. Louis, MO-IL for an increase of 25 ppb in full-year 8-hour daily maximum 03
concentrations (three day moving average) (Barry et al. 2018, Table 3).
BenMAP User Manual v 0.01
F-10
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.4 Minor Effects
Table F-4 summarizes the core health impacts functions used to estimate the relationship
between ozone and minor effects. Below, we present a brief summary of each of the studies
and any items that are unique to the study.
Table F-27. Core Health Impact Functions for Ozone and Minor Effects
Ci
Effect
Author
Year
Location
Age P,
Loss Gilliland et
Days, All a I.
Cause
Minor
Restricted Ostroand
Activity Rothschild
Days
Minor
Restricted Ostroand
Activity Rothschild
Days
Hay
Fever/
Rhinitis
Parker et
a I.
Hay
Fever/
Rhinitis
Parker et
a I.
2001
Southern
California
1989 Nationwide
1989 Nationwide
18-
64
5-17
18-
64
PM2
pm2
2009 Nationwide 3-17
2009 Nationwide 3-17
PM2.5,
PM2.5-
10,
N02,
S02
D8HourMax
DIHourMax
D8HourMax
PM2.5,
PM2.5-
10,
N02,
S02
Annual
(D24HourMean)
0.01655
Annual
(D8HourMax)
Std Err Form
0.007824 0.004445
0.002200 0.000658
0.002200 0.000658
0.00390
Log-
linear
Log-
linear
All year, 8-
hour max
from 8-hour
mean.
Log-
linear
0.01655 0.00390 Logistic
8-hour max
from 1-hour
max using
adjustment
factor of
1.14,
resulting in
effective
beta of
0.002508.
Warm
Logistic season; long
term
Warm
season; long
term; 8-hour
max from
24-hour
mean using
adjustment
factor of
0.654,
resulting in
effective
beta of
0.010818.
BenMAP User Manual v 0.01
F-ll
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.4.1 Gillilandetal. (2001)
Gilliland et al. (2001) examined the association between air pollution and school absenteeism
among 4th grade school children (ages 9-10) in 12 southern Californian communities. The
study was conducted from January through June 1996. The authors used school records to
collect daily absence data and parental telephone interviews to identify causes. They defined
illness- related absences as respiratory or non-respiratory. A respiratory illness was defined as
an illness that included at least one of the following: runny nose/sneezing, sore throat, cough,
earache, wheezing, or asthma attack. The authors used 15 and 30 day distributed lag models
to quantify the association between ozone, PMio, and N02 and incident school absences.
Ozone levels were positively associated with all school absence measures and significantly
associated with all illness-related school absences (non-respiratory illness, respiratory illness,
URI and LRI). Neither PM io nor NO2 was significantly associated with illness-related school
absences, but PM10 was associated with non-illness related absences. The health impact
function for ozone is based on the results of the single pollutant model.
School Loss Days
Gilliland et al. (2001) defines an incident absence as an absence that followed attendance on
the previous day and the incidence rate as the number of incident absences on a given day
over the population at risk for an absence on a given day (i.e. those children who were not
absent on the previous day). Since school absences due to air pollution may last longer than
one day, an estimate of the average duration of school absences could be used to calculate
the total avoided school loss days from an estimate of avoided new absences. A simple ratio
of the total absence rate divided by the new absence rate would provide an estimate of the
average duration of school absences, which could be applied to the estimate of avoided new
absences as follows:
BenMAP User Manual v 0.01
F-12
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Equation F-l
T otalAbsences
Duration = — —
NewAbsences
ATotalAbsences = — [incidence x (e~Px03 — l)] x duration x pop
Since the function is log-linear, the baseline incidence rate (in this case, the rate of new
absences) is multiplied by duration, which reduces to the total school absence rate.
Therefore, the same result would be obtained by using a single estimate of the total school
absence rate in the C-R function. Using this approach, we assume that the same relationship
observed between pollutant and new school absences in the study would be observed for
total absences on a given day. As a result, the total school absence rate is used in the function
below. The derivation of this rate is described in the section on baseline incidence rate
estimation.
For all absences, the coefficient and standard error are based on a percent increase of 16.3
percent (95% CI -2.6 percent, 38.9 percent) associated with a 20 ppb increase in 8-hour
average ozone concentration (2001, Table 6, p. 52).
A scaling factor is used to adjust for the number of school days in the ozone season. In the
modeling program, the function is applied to every day in the ozone season (May 1 -
September30), however, in reality, school absences will be avoided only on school days. We
assume that children are in school during weekdays for all of May, two weeks in June, one
week in August, and all of September. This corresponds to approximately 2.75 months out of
the 5 month season, resulting in an estimate of 39.3% of days (2.75/5*5/7).
In addition, not all children are at-risk for a new school absence, as defined by the study. On
average, 5.5% of school children are absent from school on a given day (U.S. Department of
Education, 1996, Table 42-1). Only those who are in school on the previous day are at risk for
BenMAP User Manual v 0.01
F-13
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
a new absence (1-0.055 = 94.5%). As a result, a factor of 94.5% is used in the function to
estimate the population of school children at-risk for a new absence.
Incidence Rate: daily school absence rate = 0.055 (U.S. Department of Education, 1996, Table
42-1)
Population: population of children ages 9-10 not absent from school on a given day = 94.5%
of children ages 9-10 (The proportion of children not absent from school on a given day
(5.5%) is based on 1996 data from the U.S. Department of Education (1996, Table 42-1).)
Scaling Factor: Proportion of days that are school days in the ozone season = 0.393.
(Ozone is modeled for the 5 months from May 1 through September 30. We assume that
children are in school during weekdays for all of May, 2 weeks in June, 1 week in August, and
all of September. This corresponds to approximately 2.75 months out of the 5 month season,
resulting in an estimate of 39.3% of days (2.75/5*5/7).)
BenMAP User Manual v 0.01
F-14
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.4.2 Ostro and Rothschild (1989)
Ostro and Rothschild (1989) estimated the impact of PM2.5 and ozone on the incidence of
minor restricted activity days (MRADs) and respiratory-related restricted activity days (RRADs)
in a national sample of the adult working population, ages 18 to 65, living in metropolitan
areas. The study population is based on the Health Interview Survey (HIS), conducted by the
National Center for Health Statistics. In publications from this ongoing survey, non-elderly
adult populations are generally reported as ages 18-64. From the study, it is not clear if the
age range stops at 65 or includes 65 year olds. We apply the C-R function to individuals ages
18-64 for consistency with other studies estimating impacts to non-elderly adult populations.
The annual national survey results used in this analysis were conducted in 1976-1981.
Controlling for PM2.5, two-week average ozone had a highly variable association with RRADs
and MRADs. Controlling for ozone, two-week average PM2.5 was significantly linked to both
health effects in most years. The C-R function for ozone is based on the co-pollutant model
with PM2.5.
The study is based on a "convenience" sample of non-elderly individuals. Applying the C-R
function to this age group is likely a slight underestimate, as it seems likely that elderly are at
least as susceptible to ozone as individuals under 65. A number of studies have found that
hospital admissions for the elderly are related to ozone exposures (e.g., Schwartz, 1994b;
Schwartz, 1995).
Minor Restricted Activity Days
The coefficient and standard error used in the C-R function are based on a weighted average
of the coefficients in Ostro and Rothschild (1989, Table 4). The derivation of these estimates is
described below.
BenMAP User Manual v 0.01
F-15
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Incidence Rate: daily incidence rate for minor restricted activity days (MRAD) = 0.02137
(Ostro and Rothschild, 1989, p. 243)
Population: adult population ages 18 to 64
The coefficient used in the C-R function is a weighted average of the coefficients in Ostro and
Rothschild (1989, Table 4) using the inverse of the variance as the weight. The calculation of
the MRAD coefficient and its standard error is exactly analogous to the calculation done for
the work-loss days coefficient based on Ostro (1987).
The standard error of the coefficient is calculated as follows, assuming that the estimated
year-specific coefficients are independent:
Equation F-2
f 1981 o "\
7=1976
1981 7~
( 1981 n \ ( 1981
Equation F-3
f 1981 R \
v Pj_
This reduces down to:
Equation F-4
BenMAP User Manual v 0.01
F-16
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.4.3 Parker et al. (2009)
Parker et al. (2009) investigated the associations between long-term 03 exposure and
respiratory allergies in an unrestricted population of children (aged 3-17 years) sampled from
the United States National Health Interview Survey. Authors obtained symptom data from
participant parents, who reported respiratory allergies on annual surveys. Parker et al. (2009)
placed all study participants reporting symptoms of respiratory allergies or hay fever into a
combined rhinitis group. Parker et al. (2009) linked annual averages of S02, N02, PM2.5, and
PM2.5-10 and warm season (May to September) 03 averages to participant's addresses through
ambient air pollution and meteorological data collected from US EPA Air Quality System
monitors. The authors adjusted their logistic regression models for survey year, poverty-level,
race/ethnicity, age, family structure, insurance coverage, usual source of care, education of
adult, urban-rural status, region, and median county-level income.
Incidence, Hay Fever/Rhinitis
In a multi-pollutant model, the coefficient and standard error are based on the odds ratio of
1.18 (95% CI: 1.09-1.27) for a 10 ppb increase in 24-hour mean, warm season 03 (Parker et al.,
2009, Table 4).
The Health Impact Function was adjusted from the daily 24-hour mean metric to the daily 8-
hour max metric using a ratio of 1/1.53 = 0.65359 (inverse of the ratio of 8-hour max to 24-
hour mean ozone) (Anderson and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-17
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.5 Asthma-Related Effects
Table F-5 summarizes the core health impacts functions used to estimate the relationship
between ozone and asthma exacerbation. Below, we present a brief summary of each of the
studies and any items that are unique to the study. Based on advice from the SAB-HES (U.S.
EPA-SAB 2004), regardless of the age ranges included in the source epidemiology studies, we
extend the applied population to ages 6 to 18, reflecting the common biological basis forthe
effect in children in the broader age group.
Table F-28. Core Health Impact Functions for Ozone and Asthma-Related Effects
Effect
Author
Year
Location
Age
Co-
Poll
Metric
Beta
Std Err
Form
Notes
Asthma
Symptoms,
Cough
Lewis et
al.
2013
Detroit,
Ml
5-
12
D8HourMax
0.00708
0.00372
Logistic
All year
Asthma
Symptoms,
Wheeze
Lewis et
al.
2013
Detroit,
Ml
5-
12
D8HourMax
0.00764
0.00410
Logistic
All year
Asthma
Symptoms,
Chest
tightness
Lewis et
al.
2013
Detroit,
Ml
5-
12
D8HourMax
0.01140
0.00505
Logistic
All year
Asthma
Symptoms,
Shortness
of breath
Lewis et
al.
2013
Detroit,
Ml
5-
12
D8HourMax
0.00423
0.00386
Logistic
All year
Asthma
Onset
Tetreault
etal.
2016
Quebec,
Canada
0-
17
Annual
(D8HourMax)
0.02075
0.00146
Log-
linear
Warm
season;
separate
HIFs for
ages 0-
4; 5-17
BenMAP User Manual v 0.01
F-18
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.5.1 Lewis et al. (2013)
Lewis et al. (2013) studied the effects of short-term 03 exposure on frequency of asthma
symptoms in an asthmatic population of primarily lower-income, African American and
Latino children (aged five-12 years) in East and Southwest Detroit, Ml. Authors obtained
health and demographic data through questionnaires filled out by parents or guardians for 14
consecutive days in each studied season. Questionnaires highlighted participant's asthma
symptoms (cough, wheeze, shortness of breath, chest tightness), demographic information,
medication use, and presence of second-hand smoke. The authors acquired maximum one-
hour and maximum 8-hour 03 concentrations and meteorological data from two community-
level monitors placed on East and Southwest Detroit, Ml school rooftops. Lewis et al. (2013)
implemented a combination of generalized estimating equations and alternative logistic
regression models to estimate the associations between short-term 03 exposure and rate of
asthma symptoms. Models adjusted for age, sex, location (Eastside or Southwest), race,
household income, smoker in the home, season, and variables for companion home
intervention study (control or intervention), time (pre- or post-intervention), and the
interaction between intervention group status and time. Lewis et al. (2013) observed positive
associations between short-term 03 exposure and asthma symptoms.
Asthma Symptoms
In single-pollutant models, the coefficient and standard error are based on the all year odds
ratios of 1.12 (95% CI: 0.99-1.25) for cough, 1.13 (95% CI: 0.99-1.28) for wheeze, 1.20 (95% CI:
1.02-1.40) for chest tightness, and 1.07 (95% CI: 0.95-1.21) for shortness of breath, all for a 16
ppb (interquartile range) increase in 8-hour maximum 03 concentrations (five-day average
lag) (Lewis et al. 2013, Figure 1C).
BenMAP User Manual v 0.01
F-19
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.5.2 Tetreault et al. (2016)
Tetreault et al. (2016) investigated the effects of long-term 03 exposure on asthma onset in
children (aged zero-12 years) from Quebec, Canada. The study followed participants from the
Quebec Integrated Chronic Disease Surveillance System open birth cohort between 1999 and
2011. The authors defined new cases of asthma based on hospital discharge reports and
physician diagnoses (two diagnoses within a two-year span). Monitordata (Canadian
National Air Pollution Surveillance network) and land-use mixed effect models estimated
warm season (June to August) 03 exposures. Authors assessed associations with asthma
onset with both time of birth and time-varying exposure models and adjusted for year of
birth, sex, and indices of social and material deprivation. Tetreault et al. (2016) used Cox
proportional hazard models to observe associations between long-term 03 exposure and
asthma onset in children.
As the physiology and etiology of lung development in children is similar in children 6-17
(Baena-Cagnani et al., 2007, Guerra et al., 2004, Ochs et al., 2004, Sparrow et al., 1991, Trivedi
and Denton, 2019), we apply the 4-12 year age-stratified effect estimate from Tetreault et al.
(2016) to children ages 4-17.
Incidence, Asthma
In a single-pollutant time-varying model, the coefficient and standard error were estimated
from a warm-season hazard ratio of 1.07 (95% CI: 1.06-1.08) for a 3.26 ppb (interquartile
range) increase in annual 03 concentrations (Tetreault et al. 2016, Table 5).
BenMAP User Manual v 0.01
F-20
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.6 Sensitivity Analysis - General
Table F-6 summarizes the ozone health impact functions considered by EPA to be sensitivity
analyses. Below, we present a brief summary of each of the studies and any items that are
unique to the study.
Table F-29. Core Health Impact Functions for Ozone Sensitivity Analyses
Effect
Author
Year
Location
Age
Co-Poll
Metric
Beta
Std Err
Form
Notes
Mortality, All
Cause
Di et al.
2017
Nationwide
65-
99
PM2.5
Annual
(D8HourMax)
0.001094
0.000050
Log-
linear
All Cause,
warm season
Incidence,
Asthma
Garcia et al.
2019
12 Southern
California
communities
9-
18
Annual
(D8HourMax)
0.016946
0.010941
Log-
linear
All year
Mortality,
Respiratory
Katsouyanni
et al.
2009
90 U.S. Cities
0-
99
PM10
DIHourMax
0.000985
0.000667
Log-
linear
Multi-
pollutant, lag 1,
warm season
Mortality,
Respiratory
Katsouyanni
et al.
2009
90 U.S. Cities
0-
99
PM10
D8HourMax
0.000985
0.000667
Log-
linear
Multi-
pollutant, lag 1,
warm season,
8-hour max
from 1-hour
max using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.001113.
Mortality,
Respiratory
Katsouyanni
et al.
2009
90 U.S. Cities
0-
99
DIHourMax
0.000767
0.000304
Log-
linear
Single-
pollutant, lag 1,
warm season
Mortality,
Respiratory
Katsouyanni
et al.
2009
90 U.S. Cities
0-
99
D8HourMax
0.000767
0.000304
Log-
linear
Single-
pollutant, lag-
1, warm
season, 8-hour
max from 1-
hour max using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.000867.
Mortality, All
Cause
Turner etal.
2016
Nationwide
30-
99
PM2.5
Annual
(D8HourMax)
0.001980
0.000500
Log-
linear
All Cause,
warm season,
multi-pollutant
Mortality,
Respiratory
Turner etal.
2016
Nationwide
30-
99
Annual
(D8HourMax)
0.013103
0.001791
Log-
linear
Single-
pollutant
Mortality,
Respiratory
Turner etal.
2016
Nationwide
30-
99
PM2.5
Annual
(D8HourMax)
0.011333
0.001823
Log-
linear
Multi-pollutant
BenMAP User Manual v 0.01
F-21
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.6.1 Di etal. (2017)
Di et al. (2017) evaluated the relationship between long-term ozone exposure and all-cause
mortality in nearly 61 million U.S. Medicare enrollees (over the age of 64) through 460 million
person-years of follow-up and roughly 22 million observed deaths. This cohort comprised
approximately 15% of the total U.S. population, included people living in rural areas, and is
one of the largest cohort studies published to date. The authors modeled warm season ozone
exposure across the contiguous U.S. using a hybrid methodology that included land use
regression, satellite data, and monitor data, and resolved estimations to 1 x 1-kilometer
areas. Di et al. (2017) used two-pollutant Cox proportional-hazards models with a generalized
estimating equation. The authors controlled for demographic characteristics, Medicaid
eligibility, and area-level covariates.
All-Cause Mortality
In a two-pollutant model, the coefficient and standard error for ozone are estimated from the
hazard ratio (1.011) and 95% confidence interval of (1.010-1.012) associated with a change in
annual mean ozone exposure of 10.0 ppb (Di et al., 2017, Table 2 Main Analysis).
BenMAP User Manual v 0.01
F-22
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.6.2 Garcia et al. (2019)
Garcia et al. (2019) examined the associations between long-term ozone exposure and
asthma onset in children (aged nine-18 years) with no history of asthma in Southern
California. The authors followed three waves of participants from the Children's Health Study
for eight years between 1993 and 2014. Garcia etal. (2019) obtained health and demographic
data from parents, guardians, or participants, who completed questionnaires annually. In
orderto calculate annual mean, community-level ozone exposure, the authors acquired daily
eight-hour mean ozone concentrations through ambient air pollution monitors. Multi-level
Poisson regression models with one-year lag showed no statistically significant associations
between long-term ozone exposure and asthma onset in children. Models adjusted for
demographic variables as well as factors pertaining to family medical history, environmental
factors, and near-roadway pollution.
Incidence, Asthma
In a single-pollutant, all year model, the coefficient and standard error were estimated from
an incidence rate ratio of 0.86 (95% CI: 0.71-1.04) for an 8.9 ppb decrease in eight-hour mean
ozone concentrations (Garcia et al. 2019, Table 2). For consistency with the other HIFs, we
convert this to a rate ratio for an increase in ozone, by taking the inverse of the reported
incidence rate ratio, giving a rate ratio of 1.163 (95% CI: 0.962-1.408) for an 8.9 ppb increase in
eight-hour mean ozone concentrations.
BenMAP User Manual v 0.01
F-23
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.6.3 Katsouyanni et al. (2009)
See full study descriptionunder Katsouyanni et al. (2009) in Appendix F, Section F.l.l.
Short-term Mortality, Respiratory (Multi-Pollutant)
In a multi-pollutant model, the coefficient and standard error are based on the summer-only
penalized spline estimate of respiratory mortality of 0.99% (-0.33,2.31%) per 10 ppb increase
in 03 lagged by 1 day (Katsouyanni et al. 2009, Table 24: Lag 1; Penalized splines; Controlling
forPMio).
The Health Impact Function was adjusted from the daily 1-hour max metric to the daily 8-
hour max metric using a ratio of 1.13 (ratio of 1-hour max to 8-hour max ozone) (Anderson
and Bell 2010, Table 2).
Short-term Mortality, Respiratory (Single-Pollutant)
In a single-pollutant model, the coefficient and standard error are based on the summer-only
penalized spline estimate of respiratory mortality of 0.77% (0.17,1.37%) per 10 ppb increase
in 03 lagged by 1 day (Katsouyanni et al. 2009, Table 24: Lag 1; Penalized splines; 03 Results).
The Health Impact Function was adjusted from the daily 1-hour max metric to the daily 8-
hour max metric using a ratio of 1.13 (ratio of 1-hour max to 8-hour max ozone) (Anderson
and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-24
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.6.4 Turner etal. (2016)
See full study description underTurneret al. (2016) in Appendix F, Section F.1.2.
Long-term Mortality, All Cause
In a multi-pollutant model, the coefficient and standard error are based on the warm-season
specific hazard ratio of 1.02 (1.01-1.03) per 10 ppb increase in seasonal average of daily 8-
hour maximum 03 concentrations (Turneret al. 2016, Table E9: HBM 03, multipollutant model
data, fully adjusted).
Long-term Mortality, Respiratory (Single-Pollutant)
In a single-pollutant model, the coefficient and standard error are based on the warm-season
specific hazard ratio of 1.14 (1.10-1.18) per 10 ppb increase in seasonal average of daily 8-
hour maximum 03 concentrations (Turner et al. 2016, Table E5: HBM 03,1982-2004, fully
adjusted plus ecological covariate).
Long-term Mortality, Respiratory (Multi-Pollutant)
In a multi-pollutant model, the coefficient and standard error are based on the warm-season
specific hazard ratio of 1.12 (1.08-1.16) per 10 ppb increase in seasonal average of daily 8-
hour maximum 03 concentrations (Turner et al. 2016, Table E7: HBM 03,1982-2004,
multipollutant, fully adjusted).
BenMAP User Manual v 0.01
F-25
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7 Sensitivity Analysis - At-Risk Populations
Table F-7 summarizes the ozone health impact functions considered by EPA to be sensitivity
analyses that characterize risk experienced by certain subpopulations. Below, we present a
brief summary of each of the studies and any items that are unique to the study.
Table F-30. Core Health Impact Functions for Ozone Sensitivity Analyses of At-Risk
Populations
Effect
Author
Year
Location
Age
Metric
Beta
Std Err
Form
Notes
HA, All
Respiratory
Cakmaket
al.
2006
10
Canadian
Cities
0-
99
D24HourMean
0.002033
0.000580
Logistic
Female
HA, All
Respiratory
Cakmaket
al.
2006
10
Canadian
Cities
0-
99
D8HourMax
0.002033
0.000580
Logistic
Female, 8-
hour max
from 24-hour
mean using
adjustment
factor of
0.654,
resulting in
effective beta
of 0.001328.
HA, All
Respiratory
Cakmaket
al.
2006
10
Canadian
Cities
0-
99
D24HourMean
0.002530
0.000519
Logistic
Male
HA, All
Respiratory
Cakmaket
al.
2006
10
Canadian
Cities
0-
99
D8HourMax
0.002530
0.000519
Logistic
Male, 8-hour
max from 24-
hour mean
using
adjustment
factor of
0.654,
resulting in
effective beta
of 0.001653.
Mortality,
Respiratory
Jerrett et al.
2009
Nationwide
US, 96 MS As
30-
99
Annual
(DIHourMax)
0.003922
0.000972
Log-
linear
Female
Mortality,
Respiratory
Jerrett et al.
2009
Nationwide
US, 96 MS As
30-
99
Annual
(D8HourMax)
0.003922
0.000972
Log-
linear
Female, 8-
hour max
from 1-hour
max using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.004432.
Mortality,
Respiratory
Jerrett et al.
2009
Nationwide
US, 96 MS As
30-
99
Annual
(DIHourMax)
0.000995
0.001257
Log-
linear
Male
BenMAP User Manual v 0.01
F-26
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Mortality, Jerrettetal.
Respiratory
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
2009 Nationwide 30-
US, 96 MS As 99
Katsouyanni 2009 Nationwide 0-
etal. US, 90 cities 74
Katsouyanni 2009 Nationwide 0-
etal. US, 90 cities 74
Katsouyanni 2009 Nationwide 75-
etal. US, 90 cities 99
Katsouyanni 2009 Nationwide 75-
etal. US, 90 cities 99
HA, Lower Linetal.
Respiratory
Infection
HA, Lower Linetal.
Respiratory
Infection
2005
2005
Toronto,
Canada
Toronto,
Canada
HA, Lower Linetal.
Respiratory
Infection
HA, Lower Linetal.
Respiratory
Infection
2005
2005
Toronto,
Canada
Toronto,
Canada
Emergency Mar and
Room Koenig
0-
14
0-
14
0-
14
0-
14
2009 Seattle, 0-
Washington 17
Annual
(D8HourMax)
0.000995 0.001257
DIHourMax
D8HourMax
D8HourMax
0.010436
Log-
linear
DIHourMax 0.000698 0.000213 Logistic
D8HourMax 0.000698 0.000213 Logistic
DIHourMax 0.000618 0.000233 Logistic
D8HourMax 0.000618 0.000233 Logistic
DIHourMax 0.007592 0.005232 Logistic
D8HourMax 0.007592 0.005232 Logistic
0.003530 0.004524 Logistic
0.003530 0.004524 Logistic
0.004358 Log-
linear
Male, 8-hour
max from 1-
hour max
using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.001124.
Age <75
Age <75,8-
hour max
from 1-hour
max using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.000788.
Age >75
Age >75,8-
hour max
from 1-hour
max using
adjustment
factor of 1.13,
resulting in
effective beta
of 0.000698.
Female
Female, 8-
hour max
from 1-hour
max using
adjustment
factor of 1.14,
resulting in
effective beta
of 0.008655.
Male
Male, 8-hour
max from 1-
hour max
using
adjustment
factor of 1.14,
resulting in
effective beta
of 0.004025.
Age <18
BenMAP User Manual v 0.01
F-27
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Visits,
Asthma
Emergency Mar and
Room Koenig
Visits,
Asthma
Mortality, Medina-
All Cause Ramon &
Schwartz
Mortality, Medina-
All Cause Ramon &
Schwartz
Mortality, Medina-
All Cause Ramon &
Schwartz
Mortality, Medina-
All Cause Ramon &
Schwartz
Emergency Pauluand
Room Smith
Visits,
Asthma
Emergency Pauluand
Room Smith
Visits,
Asthma
Emergency
Room
Visits,
Asthma
Emergency
Room
Visits,
Asthma
Emergency
Room
Visits,
Asthma
Emergency
Room
Visits,
Asthma
Emergency
Room
Visits,
Asthma
Emergency
Room
Visits,
Asthma
Emergency
Room
Visits,
Asthma
Emergency
Room
Paulu and
Smith
Paulu and
Smith
Villeneuve
et a I.
Villeneuve
et a I.
Villeneuve
et a I.
Villeneuve
et a I.
Villeneuve
et a I.
Villeneuve
et a I.
2009 Seattle, 18-
Washington 99
2008 Nationwide 0-
US, 48 cities 64
2008 Nationwide 65-
US, 48 cities 99
2008 Nationwide 0-
US, 48 cities 99
2008 Nationwide 0-
US, 48 cities 99
2008 Maine 2-
14
2008 Maine 15-
34
2008 Maine 2-
99
2008 Maine 2-
99
2007 Edmonton, 2-4
Canada
2007 Edmonton, 5-
Canada 14
2007 Edmonton, 15-
Canada 14
2007 Edmonton, 45-
Canada 64
2007 Edmonton, 65-
Canada 74
2007 Edmonton, 75-
Canada 99
D8HourMax 0.003922 0.002688 Log-
linear
Age >18
D8HourMean - 0.000102 Logistic
0.000130 Age <65
D8HourMean 0.000965 0.000235 Logistic
D8HourMean 0.000936 0.00024 Logistic
D8HourMean 0.000359 0.000038 Logistic
D8HourMax 0.010436 0.005027 Logistic
D8HourMax 0.014842 0.003524 Logistic
D8HourMax 0.011333 0.002736 Logistic
D8HourMax 0.010436 0.003222 Logistic
Age >65
Female, Age
0-99
Male, AgeO-
99
Age 2-14
Age 15-34
Female, Age
2-99
Male, Age 2-
99
D8HourMax 0.003237 0.003342 Logistic Age 2-4
D8HourMax 0.007279 0.002357 Logistic Age 5-14
D8HourMax 0.005798 0.00179 Logistic Age 15-14
D8HourMax 0.006296 0.003275 Logistic Age 45-64
D8HourMax 0.007279 0.005544 Logistic Age 65-74
D8HourMax - 0.006684 Logistic Age 75-99
0.000558
BenMAP User Manual v 0.01
F-28
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
Visits,
Asthma
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
Mortality,
All Cause
Zanobetti
and
Schwartz
Zanobetti
and
Schwartz
Zanobetti
and
Schwartz
Zanobetti
and
Schwartz
Zanobetti
and
Schwartz
Zanobetti
and
Schwartz
Zanobetti
and
Schwartz
Zanobetti
and
Schwartz
2008 Nationwide 0-
US, 48 cities 20
2008 Nationwide 21-
US, 48 cities 30
2008 Nationwide 31-
US, 48 cities 40
2008 Nationwide 41-
US, 48 cities 50
2008 Nationwide 51-
US, 48 cities 60
2008 Nationwide 61-
US, 48 cities 70
2008 Nationwide 71-
US, 48 cities 80
2008 Nationwide 81-
US, 48 cities 99
D8HourMean 0.00008 0.000252 Logistic Age 0-20
D8HourMean 0.0001 0.000392 Logistic Age 21-30
D8HourMean 0.00007 0.000229 Logistic Age 31-40
D8HourMean 0.00008 0.000178 Logistic Age 41-50
D8HourMean 0.000539 0.000178 Logistic Age 51-60
D8HourMean 0.000379 0.000114 Logistic Age 61-70
D8HourMean 0.000499 0.000089 Logistic Age 71-80
D8HourMean 0.00029 0.000079 Logistic Age 81-99
BenMAP User Manual v 0.01
F-29
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.1 Cakmaketal. (2006)
Cakmak et al. (2006) examined the relationship between daily average 03 concentrations and
hospital admissions for respiratory causes (ICD-9 466,480-486,490,491,492,493,494, and
496) among residents of 10 Canadian cities (Calgary, Edmonton, Halifax, London, Ottawa,
Saint John, Toronto, Vancouver, Windsor, and Winnipeg). Data on 215,544 respiratory
hospitalizations were obtained from the Canadian Institute for Health Information. Daily 24-
hour average 03 concentrations in all seasons were estimated using the average of data from
all available monitors within each city. The authors ran city-specific multi-pollutant Poisson
regression models by sex, education level, and income quartile using time lags of 0 to 5 days.
Models controlled for day of the week, temperature, humidity, and barometric pressure.
Pooled estimates across all 10 cities were calculated by using a random-effects model.
Hospital Admissions, All Respiratory
In single-pollutant models, the coefficient and standard error are estimated from a
percentage increase of 3.6% (95% CI: 1.6-5.7%) for females and 4.5% (95% CI: 2.6-6.3%) for
males for a 17.4 ppb increase in daily 24-hour average 03 concentrations (Cakmak et al. 2006,
Table 3).
The Health Impact Function was adjusted from the daily 24-hour mean metric to the daily 8-
hour max metric using a ratio of 1/1.53 = 0.65359 (inverse of the ratio of 8-hour max to 24-
hour mean ozone) (Anderson and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-30
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.2 Jerrettetal. (2009)
Jerrett et al. (2009) examined the potential contribution of long-term ozone exposure to the
risk of death from cardiopulmonary causes and specifically to death from respiratory causes.
Data from the study cohort of the American Cancer Society Cancer Prevention Study II were
correlated with air-pollution data from 96 metropolitan statistical areas in the United States.
Associations between warm season ozone concentrations and the risk of death were
evaluated with the use of standard and multilevel Cox regression models. In single-pollutant
models, increased concentrations of either PM2.5 or ozone were significantly associated with
an increased risk of death from cardiopulmonary causes. In two-pollutant models, PM2.5 was
associated with the risk of death from cardiovascular causes, whereas ozone was associated
with the risk of death from respiratory causes. The estimated relative risk of death from
respiratory causes that was associated with an increment in ozone concentration of 10 ppb
was 1.040 (95% confidence interval, 1.010 to 1.067). The association of ozone with the risk of
death from respiratory causes was insensitive to adjustment for confounders and to the type
of statistical model used. The authors concluded that they were not able to detect an effect of
ozone on the risk of death from cardiovascular causes when the concentration of PM2.5 was
taken into account. But they did demonstrate a significant increase in the risk of death from
respiratory causes in association with an increase in ozone concentration.
Mortality, Respiratory (ICD-9 code 460-519)
In single-pollutant models, the coefficient and standard error are estimated from a relative
risk of 1.04 (95% CI: 1.03-1.07) for females and 1.01 (95% CI: 0.99-1.04) for males for a 10 ppb
change in ambient ozone concentration measured from April to September during the years
from 1977 to 2000 (Jerrett et al. 2009, Table 4).
The Health Impact Function was adjusted from the daily 1-hour max metric to the daily 8-
hour max metric using a ratio of 1.13 (ratio of 1-hour max to 8-hour max ozone) (Anderson
and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-31
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.3 Katsouyanni et al. (2009)
See full study description under Katsouyanni et al. (2009) in Appendix F, Section F.l.l.
Short-term Mortality, All Cause
In single pollutant models, the coefficient and standard error are based on the summer-only
penalized spline estimate of all-cause mortality of 0.70% (0.28,1.12%) forages <75 and 0.62%
(0.16,1.08%) for ages >75 per 10 ppb increase in 03 from distributed lag days (Katsouyanni et
al. 2009, Table 24: Distributed Lags; Penalized splines; 03 Results).
The Health Impact Function was adjusted from the daily 1-hour max metric to the daily 8-
hour max metric using a ratio of 1.13 (ratio of 1-hour max to 8-hour max ozone) (Anderson
and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-32
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.4 Lin et al. (2005)
Lin et al. (2005) examined the relationship between short term 03 exposures and hospital
admissions for lower respiratory infections (ICD-9 464,466,480-487) in a case-crossover study
among Toronto residents under the age of 15 between 1998 and 2001. Data on 6,782
hospitalizations were obtained from the Discharge Abstract Database. Daily 1-hour maximum
03 concentrations in all seasons were obtained from seven monitoring stations in the
National Air Pollution Surveillance system. The authors ran conditional logistic regression
models controlling for temperature and weather conditions using 1-to 7-day average lags.
Hospital Admissions, All Respiratory
In multi-pollutant models adjusted for PM2.sand PM10-2.5, the coefficient and standard error are
estimated from an odds ratio of 1.18 (95% CI: 0.94-1.47) for females and 1.08 (95% CI: 0.89-
1.31) for males for a 21.8 ppb increase in daily 1-hour maximum 03 concentrations with a 4-
day average lag (Lin etal. 2005, Table 3 (Adjusted B)).
The Health Impact Function was adjusted from the daily 1-hour max metric to the daily 8-
hour max metric using a ratio of 1.13 (ratio of 1-hour max to 8-hour max ozone) (Anderson
and Bell 2010, Table 2).
BenMAP User Manual v 0.01
F-33
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.5 Mar and Koenig (2009)
Mar and Koenig (2009) evaluated the relationship between outdoor ozone in the summer and
asthma aggravation. The authors used hospital data on daily asthma cases from 1998 to 2002
in Seattle with local monitored PM2.5and ozone concentrations to assess the association
between asthma visits to the emergency department and air pollutants. They analyzed 1-
hour and 8-hour max ozone concentrations at two monitors in Greater Seattle. Asthma ED
visits were analyzed at 0 through 5-day lags. The authors found that ozone exposure
exacerbates asthma in people in the Seattle area, especially in children. Authors found that in
adults during the warmer months between May and September, a 10 ppb increase in 8-hour
maximum ozone concentration is associated with relative risk of asthma-related ED visits of
1.08 (1.02,1.14) with a 4-day lag. In children, during the warmer months, a 10 ppb increase in
8-hour maximum ozone concentration is associated with relative risk of asthma-related ED
visits of 1.11 (1.02,1.21) with a 3-day lag. The difference in lag and relative risk between
children and adults suggests that children are more immediately responsive to the adverse
effects of ozone exposure.
Emergency Room Visits, Asthma
In single-pollutant models, the coefficient and standard error are estimated from a relative
risk of 1.11 (95% CI: 1.02-1.21) for age <18 and 1.04 (95% CI: 0.99-1.10) for age >18 for a 10 ppb
increase in daily 8-hour maximum summer ozone concentration with a 3-day average lag
(Mar and Koenig, 2009, Table 5 and Table 6).
BenMAP User Manual v 0.01
F-34
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.6 Medina-Ramon & Schwartz (2008)
Medina-Ramon & Schwartz (2008) evaluated short-term 03 exposure and all-cause mortality
among residents of all ages in 48 U.S. cities from 1989-2000 using a case-only approach. Data
on 2,729,640 non-accidental deaths was obtained from the National Center for Health
Statistics. The authors estimated 8-hour daily mean ozone concentrations (warm season, May
to September) for each city using daily ozone levels reported by the U.S. EPA Aerometric
Retrieval System. The authors ran city-specific conditional logistic regressions controlling for
seasonality, temperature, and day of the week and pooled the results across cities. Results
were presented by socio-demographic characteristics and chronic conditions.
Mortality, All Cause
The Os-mortality risk estimates for at-risk subpopulations reported in Medina-Ramon &
Schwartz (2008) required additional modification in order to use those results to develop
health impact functions. The authors presented excess risk estimates for each subpopulation
as the additional percent change in mortality for persons who have the condition, compared
to persons without the condition. For our populations of interest, these subgroups were
persons age 65 or older compared to those younger than 65, and females relative to males.
However, they did not report the risk estimate for these comparison groups, so in order to
estimate the total excess risk for each exposed at-risk group, we needed to first back-
calculate the excess risk for the comparison group without the factor of interest. We
accomplished this by assuming that the authors' overall reported excess risk for the full
sample of 0.65% (95% confidence interval = 0.38% to 0.93%) could be expressed as a
weighted average of the unreported excess risk ("x") and the full excess risk for the at-risk
group, which would be expressed as the sum of x and the reported excess risk from Medina-
Ramon & Schwartz (2008) Table 2, where the weights are calculated using the total and at-risk
group sample sizes in Table 1 of that paper. For example, to calculate the total excess risks for
the females in the sample, we used the following equation:
BenMAP User Manual v 0.01
F-35
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
ERTotal —
Equation F-5
ERMaie (PoPMale) ERpemaje(Poppemaie)
Popxotal
where ERTotai is the full sample excess risk of 0.65%; ERFemaie is the excess risk of ozone
exposures for females; ERMaie is the excess risk of ozone exposures for males; PopTotai is the
total sample population; and PopFemaie and PopMaie are the size of the female and male subsets
of the sample population, respectively. We also know from Table 2 of that paper that:
Equation F-6
ERFemaie = ERMale + 0.58 %
Substituting and using the available information from Medina-Ramon & Schwartz (2008)
Tables 1 and 2, we can solve for ERMaie and then ERFemaie:
We then used the full excess risk value for the female subpopulation to derive a health impact
function for ozone-related mortality for females. Final calculated excess risks are 0.94% (0.47-
1.42%) for females aged 0-99; 0.36% (0.29-0.44%) for males aged 0-99, -0.13% (-0.33-0.07%)
for both sexes aged 0-64, and 0.97% (0.51-1.44%) for both sexes aged 65-99.
0.65% =
Equation F-7
ERMale(l,365,937) + (0.58% + ERMale)(l,363,703)
2,729,640
ERMaie = 0-36 %
and
ERFemaie = 0.36% + 0.58% = 0.94%
BenMAP User Manual v 0.01
December 2022
F-36
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.7 Paulu and Smith (2008)
Paulu and Smith (2008) conducted a case-crossover analysis to evaluate the relationship
between daily ozone concentrations and emergency room visits for asthma (ICD-9 493)
among Maine residents aged 2 and older from 2000 to 2003. Data on 8,020 asthma-related ER
visits was obtained from the Maine Health Data Organization. Daily 8-hour maximum 03
concentrations were computed from two monitor sites situated in or near Portland, ME and
data was obtained from the Maine Department of Environmental Protection, Bureau of Air
Quality. The authors defined the warm season as May 22-September 10 (2000) and May 23-
September 11 (2001-2003). The authors ran conditional logistic regression models stratified
by sex and age groups (2-14,15-34,35-64, and >65 years) controlling for temperature,
humidity, and day after major holidays as well as PM2.sin co-pollutant models. Paulu and
Smith (2008) found that excess risk was concentrated among females aged 15 to 34 and
males younger than 15.
Emergency Room Visits, Asthma
In single-pollutant models for both sexes, the coefficient and standard error are estimated
from an excess risk of 11% (1-23%) for ages 2-14 and 16% (8-24%) for ages 15-34 for a 10 ppb
increase in average daily 8-hour maximum ozone (lags 0-3 days) (Paulu and Smith, 2008,
Figure 1 text).
In single-pollutant models for ages 2 and above, the coefficient and standard error are
estimated from an excess risk of 12% (6-18%) for females and 11% (4-18%) for males for a 10
ppb increase in average daily 8-hour maximum ozone (lags 0-3 days) (Paulu and Smith, 2008,
Figure 1 text).
BenMAP User Manual v 0.01
F-37
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.8 ViUeneuve etal. (2007)
Villeneuve et al. (2007) evaluated the relationship between short-term ozone exposure and
emergency room visits for asthma (ICD-9 493) among residents of Edmonton, Canada aged 2
and above from 1992 to 2002 in a case-crossover study. Data on 57,912 asthma-related ER
visits was provided by Capital Health. Daily 8-hour maximum 03 concentrations were
obtained from automated fixed-site monitoring stations maintained by Environment Canada
as part of the National Air Pollution Surveillance Network. The monitors measured both warm
season (April-September) and cold season (October-March) ozone concentrations. The
authors ran conditional logistic regression models controlling for temperature, humidity,
influenza, and aeroallergens. Villeneuve et al. (2007) found associations between ozone and
asthma emergency room visits in the warm season and observed the strongest effects in
young children.
Emergency Room Visits, Asthma
In single-pollutant models, the coefficient and standard error are estimated from odds ratios
of 1.06 (0.94-1.19) for ages 2-4; 1.14 (1.05-1.24) for ages 5-14; 1.11 (1.04-1.18) for ages 15-44;
1.12 (1.00-1.26) for ages 45-64; 1.14 (0.94-1.39) for ages 65-74; and 0.99 (0.78-1.25) for ages 75-
99 for a 18 ppb increase in average daily 8-hour maximum warm season ozone (5-day average
lag) (Villeneuve et al., 2007, Tables 4,5,6,7,8,9).
BenMAP User Manual v 0.01
F-38
December 2022
-------
Appendix F: Core Ozone Health Impact Functions in BenMAP
F.7.9 Zanobetti and Schwartz (2008b)
Zanobetti and Schwartz (2008b) evaluated the relationship between short-term ozone
exposure and all-cause mortality across 48 U.S. cities (1989-2000) for all ages in a case-
crossover study. Data on 6,951,395 total deaths was provided by the National Center for
Health Statistics. Daily 8-hour mean 03 concentrations were obtained from the U.S. EPA Air
Quality System Technology Transfer Network for all seasons. The authors ran conditional
logistic regression models by season, month, and age group (0-20,21-30,31-40,41-50,51-60,
61-70,71-80,80+) controlling for temperature, dew point, and day of the week.
Mortality, All Cause
In single-pollutant models, the coefficient and standard error are estimated from an excess
risk of 0.08% (-0.42-0.57%) for ages 0-20; 0.1% (-0.67-0.87%) for ages 21-30; 0.07% (-0.38-
0.52%) for ages 31-40; 0.08% (-0.27-0.43%) for ages 41-50; 0.54% (0.19-0.89%) for ages 51-60;
0.38% (0.16-0.61%) for ages 61-70; 0.5% (0.32-0.67%) for ages 71-80; and 0.29% (0.13-0.44%)
for ages 81-99 for a 10 ppb increase in average daily 8-hour mean all season ozone (Zanobetti
and Schwartz 2008b, Table 2).
BenMAP User Manual v 0.01
F-39
December 2022
-------
Appendix G: Additional Health Impact Functions in BenMAP
Appendix G. Additional Health Impact
Functions in BenMAP
In this Appendix, we present additional health impact functions for estimating PM2.5 and
Ozone-related adverse health effects. Unlike Appendices E and F, these functions are not
currently used by the U.S. EPA in regulatory impact analyses. PLACEHOLDER: Information
on additional functions will be included here as they are added to the tool by EPA.
BenMAP User Manual v 0.01
G-l
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Appendix H. Core Health Valuation
Functions in BenMAP
This appendix presents the core unit values that are available in BenMAP for each of the
health effects included in the current suite of health impact functions. Specifically, this
appendix includes the values currently used by U.S. EPA in regulatory impact analyses. For
the .apvx files summarizing current EPA practices, see:
https://www.epa.gov/benmap/benmap-community-edition
Wherever possible, we present a distribution of the unit value, characterizing the uncertainty
surrounding any point estimate. The mean of the distribution is taken as the point estimate of
the unit value, and the distribution itself is used to characterize the uncertainty surrounding
the unit value, which feeds into the uncertainty surrounding the monetary benefits
associated with reducing the incidence of the health effect. Below we give detailed
descriptions of the derivations of unit values and their distributions, as well as tables listing
the unit values and their distributions, available for each health effect. The definitions of the
distributions and their parameters are given in Table H-l.
BenMAP User Manual v 0.01
H-l
December 2022
-------
Table H-31.
Appendix H: Core Health Valuation Functions in BenMAP
Unit Value Uncertainty Distributions and Their Parameters
Normal
Triangular
Standard deviation
Minimum value
Maximum value
Lognormal **
Mean of corresponding
normal distribution
Standard deviation of
corresponding normal
distribution
Uniform
Minimum value
Maximum value
Weibull ***
a
*ln all cases, BenMAP calculates the mean of the distribution, which is used as the "point estimate" of the unit value.
** If Y is a normal random variable, and Y = logeX, then X is lognormally distributed. Equivalently, X is lognormally distributed
if X = eY, where Y is normally distributed.
*** The Weibull distribution has the following probability density function:
This appendix also presents EPA methods for developing income growth adjustment factors
that allow you to adjust the WTP estimates to account forthe growth in income overtime.
The economics literature concerning the appropriate method for valuing reductions in
premature mortality risk is still developing. The adoption of a value forthe projected
reduction in the risk of premature mortality is the subject of continuing discussion within the
economics and public policy analysis communities. Issues such as the appropriate discount
rate and whether there are factors, such as age or the quality of life, that should betaken into
consideration when estimating the value of avoided premature mortality are still under
discussion. BenMAP currently offers a variety of options reflecting the uncertainty
surrounding the unit value for premature mortality.
Equation H-l
H.i Mortality
BenMAP User Manual v 0.01
H-2
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
H.l.l Value of a Statistical Life Based on 26 Studies
The current undiscounted VSL used by EPA is $8.7 million (2015$). This estimate is the mean
of a distribution fitted to 26 "value of statistical life" (VSL) estimates that appear in the
economics literature and that have been identified in the Section 812 Reports to Congress as
"applicable to policy analysis." This represents an intermediate value from a variety of
estimates, and it is a value EPA has frequently used in Regulatory Impact Analyses (RIAs) as
well as in the Section 812 Retrospective and Prospective Analyses of the Clean Air Act.
Accounting for the cessation lag, or the delay between pollutant exposure and death, this VSL
equates to $7.8 million (2015$) using a 3% discount rate and $7.1 million (2015$) using a 7%
discount rate (U.S. EPA, 2014).
The VSL approach mirrors that of Viscusi (1992), and uses the same criteria as Viscusi in his
review of value-of-life studies. The $8.7 million estimate is consistent with Viscusi's
conclusion (updated to 2015$) that "most of the reasonable estimates of the value of life are
clustered in the $5.2 to $12.3 million range." Five of the 26 studies are contingent valuation
(CV) studies, which directly solicit WTP information from subjects; the rest are wage-risk
studies, which base WTP estimates on estimates of the additional compensation demanded
in the labor market for riskier jobs. Because this VSL-based unit value does not distinguish
among people based on the age at their death orthe quality of their lives, it can be applied to
all premature deaths. Table H-2 presents the unit values for the 26 value-of-life studies, the
3% and 7% discounted unit values represent the core EPA values for this effect while the
undiscounted rate represents an additional valuation function.
BenMAP User Manual v 0.01
H-3
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-32. Core Unit Values for VSL based on 26-value-of-life studies
Basis for Estimate*
Age Range at
Death
UnitValue
(VSL)
Distribution of
UnitValue
Parameters of
Distribution
Min
Max
(2015$)
PI
P2
VSL, based on 26 value-of-
life studies
0
99
8,705,114
Weibull
9,648,168
1.50958
8
VSL, based on 26 value-of-
life studies, 3% discount rate
0
99
7,887,356
Weibull
8,741,819
1.36777
7
VSL, based on 26 value-of-
life studies, 7% discount rate
0
99
7,103,778
Weibull
7,873,354
1.23189
4
* The original value of a statistical life was calculated in 1990$. We have used a factor of 1.8134, based on the All-Items CPI-U.
H.2 Hospital Admissions & Emergency Room Visits
This section presents the core values for avoided hospital admissions, as well as avoided
emergency room visits. We assume that hospital admissions due to acute exposure to air
pollution pass through the emergency room. However, the value of hospital admissions that
we have calculated here does not account for the cost incurred in the emergency room visit.
H.2.1 Hospital Admissions
As suggested above, the total value to society of an individual's avoidance of a hospital
admission can be thought of as having two components: (1) the cost of illness (COI) to
society, including the total medical costs plus the value of the lost productivity, as well as (2)
the WTP of the individual, as well as that of others, to avoid the pain and suffering resulting
from the illness.
In the absence of estimates of social WTP to avoid hospital admissions forspecific illnesses
(components 1 plus 2 above), estimates of total COI (component 1) are available for use in
BenMAP as conservative (lower bound) estimates. Because these estimates do not include
the value of avoiding the pain and suffering resulting from the illness (component 2), they are
biased downward. Some analyses adjust COI estimates upward by multiplying by an estimate
BenMAP User Manual v 0.01 December 2022
H-4
-------
Appendix H: Core Health Valuation Functions in BenMAP
of the ratio ofWTP to COI, to better approximate total WTP. Other analyses have avoided
making this adjustment because of the possibility of over-adjusting--that is, possibly
replacing a known downward bias with an upward bias. Based on Science Advisory Board
(SAB) advice, the COI values currently available for use in BenMAP are not adjusted.
Unit values are based on ICD-code-specific estimated hospital charges and opportunity cost
of time spent in the hospital (based on the average length of a hospital stay for the illness).
The opportunity cost of a day spent in the hospital is estimated as the value of the lost daily
wage, regardless of whether or not the individual is in the workforce.
For all hospital admissions effects available in BenMAP, estimates of hospital charges and
lengths of hospital stays were based on discharge statistics provided by the Agency for
Healthcare Research and Quality's Healthcare Utilization Project National Inpatient Sample
(NIS) database (2016). The NIS is the largest inpatient care database in the United States, and
it is the only national hospital database containing charge information on all patients. It
contains data from a very large nationally representative sample of about eight million
hospital discharges, and therefore provides the best estimates of mean hospital charges and
mean lengths of stay available, with negligible standard errors. The sampling frame for the
2016 NIS is a sample of hospitals that comprises approximately 90 percent of all hospital
discharges in the United States. Since the NIS is based on discharge samples, the discharge-
level weight was used to weight discharges in order to produce national estimates. The
principle diagnoses were used to define the health effects.
Since most pollution-related hospital admissions are likely unscheduled, the unit values of
avoided hospital admissions used in BenMAP are based solely on unscheduled
hospitalizations. The total COI for an ICD-code-specific hospital stay lasting n days is
estimated as the mean hospital charge plus n times the daily lost wage.
BenMAP User Manual v 0.01
H-5
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
County-specific median annual income divided by (52*5) was used to estimate county-
specific median daily wage. The data source for median annual income is the 2015 American
Community Survey (ACS). ACS provided data for median annual income for all individuals
over 16 years old in 819 counties. For all other counties, ACS provided a five-year estimate of
median annual income for the years 2010-2014. We calculated the ratio of state-specific
median annual income in 2015 to state-specific median annual income during this five-year
interval (2010-2014). This ratio was then applied to the 2010-2014 county-specific median
annual income to obtain an estimate of 2015 county-specific income forthe 2,323 counties
without 2015 one-year estimates from ACS. Because wage data used in BenMAP are county-
specific, the unit value for a hospital admission varies from one county to another.
Although the data for hospital charges are from year 2016, the default hospital admission unit
values in BenMAP are in year 2015 dollars to be consistent with the unit values of other health
effects in BenMAP. This was done by inflating the medical costs (2016 dollars) to 2015 dollars
using BenMAP's medical inflation index.
The hospital admission outcomes that the EPA uses in its regulatory analyses are given in
Table H-3. Although unit values available for use in BenMAP are county-specific, the national
median daily wage was used to calculate opportunity costs and total costs.
BenMAP User Manual v 0.01
H-6
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-33. Core Unit Values Available for Hospital Admissions
Age Range
Mean
Mean
Total Cost of
Hospital
Length
Illness (Unit
Min
Max
Charge
of Stay
Value in
Health Effect
ICD Codes
(2015 $)
(days)
2015$)*
HA, All Cardiac Outcomes
390-459
0
99
$16,045
5.05
$16,918
HA, All Respiratory
460-519
0
18
$9,075
3.49
$9,678
HA, All Respiratory
460-519
65
99
$35,402
6.07
$36,451
HA, Alzheimer's Disease
331.0
65
99
$10,696
7.95
$12,070
HA, Cardio-, Cerebro- and
Peripheral Vascular Disease
410- 414, 429,
426- 427, 428,
430-438, 440-449
65
99
$14,665
4.82
$15,498
HA, Respiratory-1
491, 492, 493,
496
0
99
$7,676
3.86
$8,343
HA, Respiratory-2
464-466, 480-
487, 490-492, 493
65
99
$9,003
4.66
$9,808
HA, Parkinson's Disease
332
18
99
$12,190
3.83
$12,852
* The opportunity cost of a day spent in the hospital was estimated, for the above exhibit, at the median daily wage of all
workers, regardless of age. The median daily wage was calculated by dividing the median weekly wage ($864 in 2015$) by 5.
The median weekly wages for 2015 were obtained from the U.S. Census Bureau's 2015 American Community Survey,
"Selected Economic Characteristics: 2015 American Community Survey 1-Year Estimates."
Fortwo hospital admission effects, Alzheimer's and Parkinson's Disease, we calculated the
lifetime COI in addition to calculating the costs associated with the initial hospitalization.
Valuation sources of Alzheimer's disease lifetime medical costs were available from the
Alzheimer's Association (2020) report and Jutkowitz etal. (2017). Using Alzheimer's
Association (2020), we first developed an estimate of incremental annual medical expenses
for Medicare beneficiaries living with Alzheimer's Disease. Then, using the estimated life
expectancy duration of 5 years from Jutkowitz et al. (2017), we estimated total discounted
present values for a five-year stream of costs using 3 and 7 percent discount rates (Table H-4).
We note that the median age of Alzheimer's disease onset is after the age of 65. As such, we
exclude any potential lost wages given the low rate of labor force participation in this age
group. Lifetime medical costs, excluding initial hospitalization, are estimated at $156,920
using a 3% discount rate or $145,946 using a 7% discount rate in 2015.
BenMAP User Manual v 0.01
H-7
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-34. Alzheimer's Disease Valuation (2015$)
Year
3% Discount Rate
7% Discount Rate
0
$33,266
$33,266
1
$32,297
$31,090
2
$31,357
$29,056
3
$30,443
$27,155
4
$29,557
$25,379
Total Lifetime Costs
$156,920
$145,946
Estimates of lifetime costs for Parkinson's Disease were provided byYangetal. (2020),
including direct, indirect, and non-medical costs. Using Yang et al. (2020), we first developed
an annual estimate of excess costs associated with living with Parkinson's Disease. Then,
using the estimated life expectancy duration of 14.6 years from De Pablo-Fernandez et al.,
2017, we calculated the present value of lifetime costs over this period using 3 and 7 percent
discount rates (Table H-5). Lifetime medical costs are estimated at $567,285 using a 3%
discount rate or $445,792 using a 7% discount rate in 2015$.
BenMAP User Manual v 0.01
H-8
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-35. Lifetime Parkinson's Disease Valuation Estimate Calculation
Year
3% Discount Rate
7% Discount Rate
0
$44,718
$44,718
1
$43,416
$41,793
2
$42,151
$39,059
3
$40,924
$36,503
4
$39,732
$34,115
5
$38,574
$31,883
6
$37,451
$29,798
7
$36,360
$27,848
8
$35,301
$26,026
9
$34,273
$24,324
10
$33,275
$22,732
11
$32,305
$21,245
12
$31,364
$19,855
13
$30,451
$18,556
14
$29,564
$17,343
14.6
$17,427
$9,992
Total Lifetime Costs (14.6 yr
survival)
$567,285
$445,792
H.2.2 Emergency Room Visits
As with hospital admissions, to value emergency room visits we develop primary COI
estimates using data from the Healthcare Cost and Utilization Project (HCUP). The 2016
Nationwide Emergency Department Sample (NEDS) provides recent, nationally
representative information on medical treatment in emergency departments. In the case of
emergency department visits, valuation estimates include only the medical costs.
The NEDS dataset includes discharge-level observations. That is, each data point represents
one individual being discharged from the emergency department (NEDS). Because
BenMAP User Manual v 0.01
H-9
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
individuals are treated in these settings for a variety of reasons, we use medical billing codes
to extract observations related to each health effect. The epidemiological studies described
in Appendix E, F and G provide ICD-9 codes for each illness; however, recent HCUP datasets
(including NEDS) use ICD-10 codes. Thus, we first crosswalk the relevant ICD-9 codes to
associated ICD-10 codes using a mapping provided by the U.S. Centers for Disease Control.
We then identify all discharges in the HCUP datasets with ICD-10 codes that match to a
study's ICD-9 code(s). Because HCUP datasets often include multiple ICD-10 codes for each
discharge, we focus on the principal diagnosis (i.e., the first-listed ICD-10 code).
In the NIS dataset, we convert total charges (i.e., the amount billed to patients, employers, or
insurance providers) into estimates of total costs (i.e., the final reimbursements for medical
treatment). Unadjusted charges are not suitable for use in regulatory analysis because posted
prices generally do not reflect actual medical costs due, in part, to negotiation between
medical providers and payers (e.g., insurance companies). We assume that adjusted charges
reflect the actual revenue the hospital receives and thus the actual cost of providing care.
This conversion is completed using hospital-specific cost-to-charge (CCR) ratios provided
with NIS. Because CCRs are not available for NEDS, we apply average CCRs for each health
effect in NIS to the same set of ICD-10 codes in NEDS.
For each health effect, mean estimates are calculated using estimation commands for survey
data to account for the sampling design and sample discharge weights of the HCUP data. This
results in estimates of mean costs and a 95% confidence interval, which represents
uncertainty in our valuation estimates of medical costs. The resulting estimates are
presented in Table H-6.
BenMAP User Manual v 0.01
H-10
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-36. Core Unit Values Available for Emergency Department Visits
Health Effect
ICD Codes
Age Range
Mean Unit Value
Min
Max
(2015$)
ER visits, All Cardiac Outcomes
390-459
0
99
$1,161
491-493,460-466,
ER visits, respiratory
477,480-486,496,
0
99
$875
786.07, 786.09
H.2.3 Emergency Room Visits for Asthma
Two unit values are currently available for use in BenMAP for asthma emergency room (ER)
visits. One is $533.69, from Smith et al., 1997, who reported that there were approximately 1.2
million asthma-related ER visits made in 1987, at a total cost of $186.5 million, in 1987$. The
average cost per visit was therefore $155 in 1987$, or $533.69 in 2015$ (using the CPI for
medical care to adjust to 2015$). The uncertainty surrounding this estimate, based on the
uncertainty surrounding the number of ER visits and the total cost of all visits reported by
Smith et al. is characterized by a triangular distribution centered at $533.69, on the interval
[$395.14, $738.19],
A second unit value is $446.52 from Stanford et al. (1999). This study considered asthmatics in
1996-1997, in comparison to the Smith et al. (1997) study, which used 1987 National Medical
Expenditure Survey (NMES) data). In comparing their study, the authors note that the 1987
NMES, used by Smith et al., "may not reflect changes in treatment patterns during the 1990s."
In addition, its costs are the costs to the hospital (or ER) for treating asthma rather than
charges or payments by the patient and/or third party payer. Costs to the ER are probably a
better measure of the value of the medical resources used up on an asthma ER visit (see
above for a discussion of costs versus charges).
BenMAP User Manual v 0.01
H-ll
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
The unit values and the corresponding distributions available in BenMAP for asthma-related
ER visits are summarized in Table H-7.
Table H-37. Core Unit Values Available for Asthma-Related ER Visits
Basis for Estimate
COI: Smith etal. (1997)
COI: Standford et al. (1999)
Age Range UnitValue
Min Max
99
99
(2015$)
$534
$447
Distribution
of Unit
Value
Parameters of
Distribution
Triangular $395
Normal 8.95
$738
H.3 Other Health Effect Occurrence
Monetary valuation estimates for health effect occurrences other than hospital admissions or
emergency department visits are described below, listed in alphabetical order.
H.3.1 Lung Cancer
The unit value for non-fatal lung cancer incidence is derived from the direct medical costs of
lung cancer treatment estimated by Kaye et al. (2018). Lost earnings are assumed to be
negligible because of the low labor force participation rate among the age groups at highest
risk of developing lung cancer (average age of diagnosis is approximately 70 years). Lung
cancer treatment costs depend to a large extent on the phase of care, with costs in the initial
year of treatment ($17,422 for males) far exceeding the continuing costs of treatment in
subsequent years ($3,269 for males). We calculated costs over a five-year span, beginning
with the initial onset which occurs with a delay after exposure. The initial year's treatment
cost is summed with four years of continuing annual costs discounted by 3% and 7%.
Furthermore, Kaye et al. (2018) provides separate treatment cost estimates for men and
women. The distribution of new lung cancer cases by sex in the United States from Siegelet
BenMAP User Manual v 0.01 December 2022
H-12
-------
Appendix H: Core Health Valuation Functions in BenMAP
al. (2019) is approximately 51% male and 49% female. This distribution of new lung cancer
cases was used to weight the sex-specific cost estimates from Kaye et al. (2018) to obtain a
combined five-year cost estimate for both sexes. In order to adjust the cost estimate to 2015$
using a medical cost index, we assume that costs presented by Kaye et al. (2018) are in 2010$
as an approximate midpoint of the data years 2007-2012. Altogether, the cost of non-fatal
lung cancer incidence over a five-year period is estimated to be $33,809 using a 3% discount
rate or $32,548 using a 7% discount rate (Table H-8).
Table H-38. Core Unit Values Available for non-fatal Lung Cancer
Health Effect
Basis for Estimate
Age Range
UnitValue
Min Max
(2015$)
COI: 5 yrs med, 3% DR,
65 99
65 99
$33,809
$32,548
Lung Cancer
Kaye (2018)
COI: 5 yrs med, 7% DR,
Kaye (2018)
For an outcome such as lung cancer, there is an expected time lag between changes in
pollutant exposure in a given year and the total realization of health effect benefits,
commonly referred to in regulatory analyses as the "cessation lag." The time between
exposure and diagnosis can be quite long, on the order of years to decades, to realize the full
benefits of the air quality improvements. This latency period is important in order to properly
discount the economic value of these health benefits.
To estimate the latency period, we performed a literature search using the keywords "non-
fatal lung cancer," "lung cancer," "PM2.5," "latency," and "incidence." Five papers that
estimate the risk of lung cancer incidence from PM2.5 exposure using a latency period were
identified. The latency period length and country of the identified papers are summarized in
Table H-9. Based on estimates of lung cancer latency from the literature, 10 years was the
most common latency period estimate found in the literature (i.e., the mode).
BenMAP User Manual v 0.01
H-13
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-39. Latency Periods Used in Lung Cancer Risk Assessment Papers
Study
Latency Period (years)
Location
Gogna etal., 2019
5
Canada
Bai etal., 2020
4; 10
Canada
Kulhanova et al., 2018
10
France
Coleman et al., 2020
10; 15
US
Harrison etal., 2004
20
US
To account for the latency period between air pollution reductions and avoided lung cancer
diagnoses in our economic valuation estimates, we developed an age-at-diagnosis cessation
lag distribution method based on an approach previously used to estimate avoided cases of
kidney cancer in analyses of water quality rules (U.S. EPA, 2017). The method uses lung and
bronchus cancer diagnosis age-distribution from the Surveillance, Epidemiology, and End
Results Program (SEER). For this model, we assumed that the case reduction distribution
would follow the age-pattern of cancer diagnosis between the age at which the exposure
change occurs and 99 years. Table H-10 shows an example case reduction distribution
calculation for an exposure change experienced at 55. SEER estimates 92.2% of lung and
bronchus cancer cases occur in individuals 55 years and older. Dividing the percentages in the
remaining age bins by 92.2% (the percent of lung and bronchus cancer diagnoses between
the age of exposure change and end of lifetime), we find that there is a 24% chance that the
risk reductions for a 55-year-old occur between ages 55 and 64, a 37% chance that the case
reductions occurs between ages 65 and 74, etc. For distributing avoided cases within an age
bin, we assume an equal incidence distribution across years within each bin.
BenMAP User Manual v 0.01
H-14
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-40. Percent Lung and Bronchus Cancer Incidence by Age and Distribution of
Risk Reduction by Age for an Exposure Change at 55
Age Group
Percent New Cases per Year by Age*
Percent of New Cases Occurring at or After Age 551
0-20
0
NA
20-34
0.2
NA
35-44
0.9
NA
45-54
6.6
NA
55-64
21.8
24
65-74
34.1
37
75-84
26.6
29
85-99
9.7
11
55-99
92.2
100
*May not sum to 100% due to rounding
Calculated as the percentage in column 2 divided by 92.2%, where 92.2% is the percentage of lung and bronchus incidence
between age 55 and 99.
H.3.2 Out of Hospital Cardiac Arrest
The COI for cardiac arrests occurring outside of the hospital is derived from O'Sullivan et al.
(2011), who estimate three-year medical costs associated with cardiovascular disease events
among adults ages 35 and older in the U.S. The authors rely on administrative claims data
from a large U.S. health plan and develop econometric models to predict medical costs for 15
different cardiovascular events, including cardiac arrest, referred to as resuscitated cardiac
arrest. The dataset includes over 20 million commercial and Medical Advantage members
between 2002 and 2006. Cardiac arrests are identified using the ICD-9 code 427.5. The authors
use propensity score matching to develop a control group with which to compare costs
versus individuals that suffered cardiac arrest. Medical costs occurring within the month of
the event were excluded to avoid double counting hospitalization costs, which are separately
captured by the hospitalization valuation functions. Over three years, the total medical costs,
excluding hospitalization, are $36,142 (undiscounted, inflated to 2015$), or $35,753 using a
3% discount rate and $35,282 for a 7% discount rate (Table H-ll).
BenMAP User Manual v 0.01
H-15
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-41. Valuation Estimate for Cardiac Arrests (2015$)
Costs
Cumulative Costs
Annual Costs
Undiscounted
3% Discount Rate
7% Discount Rate
Month of
Event*
$43,904
$43,904
$43,904
$43,904
Year 1
$71,901
$27,997
$27,997
$27,997
Year 2
$74,701
$2,800
$2,718
$2,617
Year 3
$80,046
$5,345
$5,038
$4,668
Years 1-3
$80,046
$36,142
$35,753
$35,282
*Excluded to avoid double-counting with hospitalization costs.
H.3.3 Stroke
The COI of non-fatal stroke incidence is calculated from Mu et al. (2017) estimates of direct
medical costs incurred during initial hospitalization and the 360 days following hospital
discharge. The study identifies individuals experiencing a first-time stroke using ICD-9 codes
434 and 436. The authors analyze medical claims from January 2006 to March 2015 utilizing
the retrospective IMS LifeLink PharMetrics Plus database for individuals ages 18 to 65, and
Medicare Advantage and Medicare Supplemental Claims for individuals above the age of 65.
The authors present acute care and long-term care costs stratified by three discharge
classifications: dead at discharge, discharged with disability, and discharged without
disability. We estimated the average costs for non-fatal cases by weighting the costs for
individuals discharged with disability and without disability by their prevalence (23 and 77
percent, respectively). The resulting COI for non-fatal stroke incidence is $33,962 (2015$). This
value reflects one-year of medical costs following stroke and does not include hospitalization
costs, as these costs are separately captured by hospitalization valuation functions. We
reviewed several studies that estimate longer-term medical costs (Goodwin et al., 2011, Lee
et al., 2007, Luengo-Fernandez et al., 2012, Nicholson et al., 2016) and concluded that roughly
three quarters of costs are incurred in first year afterstroke occurrence.
BenMAP User Manual v 0.01
H-16
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
H.4 Acute Symptoms and Illness Not Requiring
Hospitalization
Several acute symptoms and illnesses have been associated with air pollution, including
acute bronchitis in children, upper and lower respiratory symptoms, and exacerbation of
asthma (as indicated by one of several symptoms whose occurrence in an asthmatic
generally suggests the onset of an asthma episode). In addition, several more general health
effects which are associated with one or more of these acute symptoms and illnesses, such as
minor restricted activity days, school loss days, and work loss days, have also been
associated with air pollution. We briefly discuss the derivation of the unit values for acute
respiratory symptoms (minor restricted activity days), asthma exacerbation, and school loss
days. Tables H-12 and H-13 summarize the values used by EPA in their regulatory impact
analyses.
Table H-42. Additional Unit Values Available for Myocardial Infarction
Basis of Estimate
Age Range
Medical Cost
Opportunity
Total Cost
Min
Max
Cost
0
24
$48,796
$0
$48,796
COI: 3 yrs med, 5 yrs wages,
3% DR, O'Sullivan (2011)
25
44
$48,796
$13,301
$62,097
45
54
$48,796
$19,604
$68,400
55
65
$48,796
$113,316
$162,112
66
99
$48,796
$0
$48,796
0
24
$47,623
$0
$47,623
COI: 3 yrs med, 5 yrs wages,
7% DR, O'Sullivan (2011)
25
44
$47,623
$11,908
$59,531
45
54
$47,623
$17,552
$65,175
55
65
$47,623
$101,451
$149,074
66
99
$47,623
$0
$47,623
BenMAP User Manual v 0.01
H-17
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-43. Core Unit Values Available for Acute Symptoms and Illnesses
Health Effect
Basis for Estimate
Age Range
Distribution of
Unit Value
Parameters of
Distribution
Min
Max
PI
P2
Minor
Restricted
WTP: 1 day, CV studies
18
99
$70
Triangular
28.51
110.62
Activity Days
0
17
$17,232
Normal
441.62
0
COI: lifetime med, lifetime
productivity, 3% DR
4
21
$16,425
Normal
464.38
0
New Onset
35
99
$16,741
Normal
639.75
0
Asthma
0
17
$10,187
Normal
277.28
0
COI: lifetime med, lifetime
productivity, 7% DR
4
21
$9,644
Normal
294.27
0
35
99
$12,594
Normal
505.97
0
Cough; Chest
0
17
$219
LogNormal
5.390
0.078
Tightness;
Shortness of
Breath;
WTP: 1 symptom-day,
Dickie and Mesmen (2004)
18
99
$115
LogNormal
5.390
0.078
Wheeze
Albuterol Use
COI: use of inhaler
0
99
$0.35
None
0
0
Allergic
Rhinitis
COI: 1 yr med costs
0
17
$600
NoneO
0
0
Work Loss
Days *
Median daily wage,
county-specific
18
65
$173
None
N/A
N/A
School Loss
Days
Described in text
0
17
$106
None
N/A
N/A
* Unit values for work loss days are county-specific, based on county-specific median wages. The unit value shown here is the
national median daily wage, given for illustrative purposes only.
H.4.1 Non-Fatal Myocardial Infarctions (Heart Attacks)
In the absence of a suitable WTP value for reductions in the risk of non-fatal heart attacks,
there are a variety of cost-of-illness unit values available for use in BenMAP. These cost-of-
illness unit values incorporate two components: the direct medical costs and the opportunity
cost (lost earnings) associated with the illness event. Because the costs associated with a
heart attack extend beyond the initial event itself, the unit values include costs incurred over
five years.
BenMAP User Manual v 0.01
H-18
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Economic values for acute myocardial infarctions (AMIs, also known as heart attacks) have
been derived from O'Sullivan et al. (2011), which estimate three-year medical costs
associated with cardiovascular disease events among adults ages 35 and older in the U.S. The
authors rely on administrative claims data from a large U.S. health plan and develop
econometric models to estimate medical costs for 15 different cardiovascular events,
including AMIs. The dataset includes over 20 million commercial and Medical Advantage
members between 2002 and 2006. AMIs are identified using the ICD-9 code 410. The authors
use propensity score matching to develop a control group with which to compare costs
versus individuals that suffered AMIs. We exclude medical costs within the month of the event
in an attempt avoid double counting hospitalization costs, which would be captured
separately in the hospitalization valuation. Over three years, the total medical costs,
excluding initial hospitalization, are $49,758 (undiscounted, inflated to 2015$), or $48,796
using a 3% discount rate and $47,623 for a 7% discount rate (Table 25). This study analyzed
costs associated with individuals ages 35 and older. We apply the total medical costs to all
ages from zero to 99, although only a small portion (<10%) of annual AMI incidence occurs in
the age range below 35.
We supplement AMI medical costs with estimates of lost earnings using age-specific
estimates from Cropper and Krupnick (1990). Using a 3% discount rate, we estimated the
following present discounted values in lost earnings over5 years due to a heart attack: 0.219
times annual earnings for someone between the ages of 25 and 44,3.534 times annual
earnings for someone between the ages of 45 and 54, and 1.245 times annual earnings for
someone between the ages of 55 and 65. The corresponding age-specific estimates of lost
earnings using a 7% discount rate are 0.203,3.287, and 1.158 times annual earnings,
respectively. Cropper and Krupnick (1990) does not provide lost earnings for populations
under 25 or over 65. As such we do not include lost earnings in the cost estimates for these
age groups. These costs, along with the total valuation estimates for AMIs, are presented in
Table H-9.
BenMAP User Manual v 0.01
H-19
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
H.4.2 Minor Restricted Activity Days (MRADs)
Two unit values are currently available in BenMAP for MRADs associated with acute
respiratory symptoms. No studies are reported to have estimated WTP to avoid a minor
restricted activity day (MRAD). Although Ostro and Rothschild (1989) estimated the
relationship between PM2.5 and MRADs, rather than MRRADs (a component of MRADs), it is
likely that most of the MRADs associated with exposure to PM2.5 are in fact MRRADs. The
original unit value, then, assumes that MRADs associated with PM exposure may be more
specifically defined as MRRADs, and uses the estimate of mean WTP to avoid a MRRAD.
lEc (1993) derived an estimate of WTP to avoid a MRRAD, using WTP estimates from Tolley et
al. (1986) for avoiding a three-symptom combination of coughing, throat congestion, and
sinusitis. This estimate of WTP to avoid a MRRAD, so defined, is $38.37 (1990 $).
Any estimate of mean WTP to avoid a MRRAD (or any other type of restricted activity day
other than WLD) will be somewhat arbitrary because the effect itself is not precisely defined.
Many different combinations of symptoms could presumably result in some minor or less
minor restriction in activity. Krupnick and Kopp (1988) argued that mild symptoms will not be
sufficient to result in a MRRAD, so that WTP to avoid a MRRAD should exceed WTP to avoid
any single mild symptom. A single severe symptom or a combination of symptoms could,
however, be sufficient to restrict activity. Therefore, WTP to avoid a MRRAD should, these
authors argue, not necessarily exceed WTP to avoid a single severe symptom or a
combination of symptoms. The "severity" of a symptom, however, is similarly not precisely
defined; moreover, one level of severity of a symptom could induce restriction of activity for
one individual while not doing so for another. The same is true for any particular combination
of symptoms.
Given that there is inherently a substantial degree of arbitrariness in any point estimate of
WTP to avoid a MRRAD (or other kinds of restricted activity days), the reasonable bounds on
BenMAP User Manual v 0.01
H-20
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
such an estimate must be considered. By definition, a MRRAD does not result in loss of work.
WTP to avoid a MRRAD should therefore be less than WTP to avoid a WLD. At the other
extreme, WTP to avoid a MRRAD should exceed WTP to avoid a single mild symptom. The
highest lEc midrange estimate of WTP to avoid a single symptom is $20.03 (1999 $), for eye
irritation. The point estimate of WTP to avoid a WLD in the benefit analysis is $83 (1990 $). If
all the single symptoms evaluated by the studies are not severe, then the estimate of WTP to
avoid a MRRAD should be somewhere between $16 and $83. Because the lEc estimate of $38
falls within this range (and acknowledging the degree of arbitrariness associated with any
estimate within this range), the lEc estimate is used as the mean of a triangular distribution
centered at $38, ranging from $16 to $61. Adjusting to 2015$, this is a triangular distribution
centered at $69.58, ranging from $29 to $111.
The estimate forthe MRADs that is used in EPA benefits analyses can be found in Table H-10.
H.4.3 New Onset Asthma
The unit value of new onset asthma is reported by Belova et al. (2020), who estimate the
lifetime costs of asthma using data from the 2002 to 2010 Medical Expenditure Panel Survey
(MEPS). The authors identify all individuals with current asthma (9,409 out of 158,867
respondents) using the ICD-9 code 493 in the MEPS Medical Conditions Files. Additionally,
they identify the date of asthma onset for these individuals. Using the MEPS Medical Events
files, which capture most types of medical expenditures (e.g., hospitalizations, emergency
room visits, outpatient visits, prescriptions), Belova et al., 2020 estimated annual
expenditures by asthma duration and age at onset. The annual healthcare costs for asthma—
as measured by healthcare expenditures by all paying parties—vary from $800 to $2,100 for
children and $900 to $2,500 for adults (2015$). They extrapolate these values to a lifetime cost
stream for an incident chronic asthma case to generate present value estimates by onset age
using discount rates of 3% and 7%. Additionally, the authors consider productivity impacts
BenMAP User Manual v 0.01
H-21
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
that capture 1) the probability of not being able to work due to health reasons, 2) the impact
of asthma on occupational choice, and 3) impact of asthma on weekly earnings.
We adapted the Belova et al. (2020) estimates to align with the age groups 0 to 17,4 to 21, and
35 to 99. This calculation entails weighting the Belova et al. (2020) age groups by their relative
prevalence and propagating the standard errors to derive new uncertainty bounds.
Confidence intervals are not provided for productivity losses because BenMAP is currently
only capable of reflecting uncertainty in one parameter (in this case, medical costs) (Table H-
10).
H.4.4 Asthma Symptoms
The valuation estimates for cough, wheeze, chest tightness, and shortness of breath were
estimated from the Dickie and Messman (2004) analysis of parents' WTP to relieve asthma
symptoms in children and adults. The authors derive the WTP estimates from an attribute-
based, stated-choice question assessing preferences to avoid acute illness as part of a survey
performed in Hattiesburg, Mississippi in 2000. Survey respondents are asked to identify
whether they or their child have experienced the following asthma symptoms in the past
year: cough with phlegm, shortness of breath with wheezing, chest pain on deep inspiration,
and fever with muscle pain and fatigue. Respondents were then assigned one of sixteen
illness profiles varying by symptom, symptom duration, in days, as well as discomfort level.
Dickie and Messman (2004) calculate the WTP for children ages zero to seventeen as $219, for
one avoided mild symptom-day (2015$). The authors also provide WTP estimates by
symptom, however, they represent six avoided symptom-days. Therefore, we apply the same
WTP value, for one avoided mild symptom-day, to each asthma symptom effect.
We calculated the economic value for albuterol use associated with asthma symptoms
through prescription costs for albuterol inhalers. Epocrates and GoodRx provide cost and
actuation information forfour common types of albuterol inhalers in 2020 dollars. Both
BenMAP User Manual v 0.01
H-22
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
online resources utilize published price lists, purchases, claim records, and pharmaceutical
data to provide clinical statistics. Epocrates and the FDA provide cost and actuation
information for one additional, less common, albuterol inhaler.70 We divide the cost of
inhalers by the actuations per inhaler to calculate an average cost per actuation across all
inhaler types. We then adjust the values to 2015$ using the Consumer Price Index (CPI) for
medical care. Since medical cost index data were unavailable for 2020 at the time of these
calculations, we used the most recently available index (2019). The resulting value for asthma
symptoms, albuterol use is $0.35 per actuation (2015$).
Table H-10 summarizes the unit values utilized by EPA for asthma related health effects.
H.4.5 Allergic Rhinitis
We derived COI estimates for allergic rhinits (also referred to as hay fever) from the 2005 unit
cost data presented by Soni (2008). The study utilizes data from the Medical Expenditure
Panel Survey (MEPS) and identifies allergic rhinitis using ICD-9 code 477. Soni (2008) analyzes
medical expenditures stratified by age group for the years 2000 and 2005, and calculate the
cost-of-illness as the mean expenditures per person for ambulatory care, in-patient services,
and prescription medications. The resulting COI for allergic rhinitis is $600 for ages zero to
seventeen (2015$; Table H-10). The COI estimate represents mean annual medical costs for
patients with hay fever. Given that the health impact function for this effect relates to allergic
rhinitis prevalence, these estimates are more applicable than values representing only first-
year costs.
H.4.6 Work Loss Days (WLDs)
Work loss days are valued at a day's wage. BenMAP calculates county-specific median daily
wages from county-specific annual wages by dividing by (52*5), on the theory that a worker's
BenMAP User Manual v 0.01
H-23
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
vacation days are valued at the same daily rate as work days. The unit value for WLDs can be
found in Table H-10.
H.4.7 School Loss Days
We include two costs of school loss days: caregiver costs and loss of learning. We calculate
each separately and then sum them. Caregiver costs are valued at their employers' average
cost for employed caregivers. For unemployed caregivers, the opportunity cost of their time is
calculated as the average take-home pay. The loss of learning is calculated based on the
impact of absences on learning multiplied by the impact of school learning on adult earnings.
The loss of learning estimate is currently only available for middle and high school students.
The two costs are summed.
The caregiver costs assumes that an adult caregiver stays home with the child and loses any
wage income they would have earned that day. For working caregivers, we follow EPA
guidance and value theirtime at the average wage includingfringe benefits and overhead
costs (U.S. EPA, 2020a). The average daily wage in 2021 was $195 (2015$, assumed to be the
average weekly wage divided by 5, US Bureau of Labor Statistics, 2021a), which yields an
average daily labor cost of $340 for employed parents applying average multiplier of 1.46 for
fringe benefits and 1.2 for overhead. For nonworking caregivers, we assume that the
opportunity cost of time is the average after tax earnings. We estimate the income tax rate for
a median household to be 7%, yielding net earnings of $195*0.93 or $181 ($2015). The income
tax rate of 7% is the percentage difference in median post-tax income and median income
from (U.S. Census Bureau, 2021).
The probability that a parent is working is measured with the employment population ratio
among people with their own children under 18 and is 77.2% (US Bureau of Labor Statistics,
2021b). Combining the cost of working and nonworking caregivers yields a caregiver cost of
$305 per school loss day.
BenMAP User Manual v 0.01
H-24
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
To measure the loss of learning, we update the Liu et al., 2021 estimate. Liu et al., 2021
estimated the impact of a school absence on learnings as measured by an end-of-course test
score. We multiply by an estimate of the impact of learning as measured by end-of-course
test scores on adult income from Chetty et al., 2014. This approach yields an estimated
learning loss of $2,230 per school absence (discounted at 3%) or $975 per school absence
(discounted at 7%).
We updated the Chetty et al., 2014 estimate to use 2010 income and to estimate lifetime
incomes discounted at 3% and 7%. Liu et al., 2021 provide an estimate that a school absence
leads to a $1,200 reduction in lifetime earnings, based on the Chetty et al., 2014 estimate that
lifetime earnings are $522,000 (2010$). We use 2010 ACS data from IPUMS to calculate
expected lifetime earnings of $892,579 (discounting at 3%) and $390,393 (discounting at
7%). We then multiply the Liu et al., 2021 estimate of $1200 by (892,579/522,000) and
(390,393/522,000) and convert from 2010 dollars to 2015 dollars based on the Consumer Price
Index for All Urban Consumers.
We use caregiver costs for preschool and elementary school children and the sum of caregiver
costs and loss of learning for middle school and high school students. We calculate that 31%
of children under 18 are middle school and high school ages 13-18, assuming each bin
distributed equally), so the combined average effect is ($305 + $2,230*0.31) or $1000 with 3%
discounting, or ($305 + $975*0.31) or $610 with 7% discounting (U.S. Census Bureau, 2010).
A unit value based on the approach described above is likely to understate the value of a
school loss day in at least four ways:
1. It omits WTP to avoid the symptoms/illness which resulted in the school absence
2. It omits the opportunity cost of time for non-working caregivers' day
3. The approach omits other aspects of school attendance such as social and emotional
development or meals
BenMAP User Manual v 0.01
H-25
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
4. It does not account for deleterious effects on student learning in other subjects.
The unit value for school loss days can be found in Table H-10.
H.5 Developing Income Growth Adjustment Factors
Chapter 4 of the BenMAP User Manual provides instructions forformatting and adding
income growth data. These values are used to adjust WTP estimates for growth in real
income. As discussed in that chapter, evidence and theory suggest that WTP should increase
as real income increases. When reviewing the economic literature to develop income growth
adjustment factors, it is important to have an economist assist. For an overview of valuation,
see Chapter4: Valuation and Discounting.
Adjusting WTP to reflect growth in real income requires three steps:
1. Identify relevant income elasticity estimates from the peer-reviewed literature.
2. Calculate changes in future income.
3. Calculate adjustments to WTP based on changes in future income and income elasticity
estimates.
I. Identifying income elasticity estimates
Income elasticity estimates relate changes in demand for goods to changes in income.
Positive income elasticity suggests that as income rises, demand forthe good also rises.
Negative income elasticity suggests that as income rises, demand forthe good falls. BenMAP
does not adjust Cost-of-lllness (COI) estimates according to changes in income elasticity due
to the fact that COI estimates the direct cost of a health outcome; instead we adjust this
metric using inflation factors described above. BenMAP includes income elasticity estimates
specific to the type of health effect associated with the WTP estimate. BenMAP contains
elasticity estimates for three types of health effects: minor, severe and mortality. Minor health
BenMAP User Manual v 0.01
H-26
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
effects are those of short duration. Severe, or chronic, health effects are of longer duration.
Consistent with economic theory, the peer reviewed literature indicates that income
elasticity varies according to the severity of the health effect. A review of the literature
revealed a range of income elasticity estimates that varied across the studies and according
to the severity of health effect. Table H-14 summarizes the income elasticity estimates for
minor health effects, severe health effects and mortality. Here we have provided a lower-,
upper- and central-elasticity estimate for each type of health effect.
Table H-44. Income Elasticity Estimates
Health Effect
Lower Bound
Central Estimate
Upper Bound
Minor Health Effect
0.04
0.15
0.30
Severe and Chronic
Health Effects
0.25
0.45
0.60
Mortality
0.08
0.40
1.00
2. Calculating changes in future income
The next input to the WTP adjustment is annual changes in income. Historical US Gross
Domestic Product (GDP) data (1990-2016) comes from the U.S. Bureau of Commerce's Bureau
of Economic Analysis (BEA). GDP values were adjusted for inflation using the BEA's price index
for GDP. We divided historical GDP values by populations provided by the BEA to estimate
GDP per capita to maintain internal consistency in the calculation. Future changes in annual
income are based on data presented in the Annual Energy Outlook (AEO) 2020, a report
prepared by the U.S. Energy Information Administration (EIA) (AEO, 2020). AEO published
annual GDP projections through the year 2050, which were adjusted for inflation using the
GDP Chain-type Price Index reported by AEO. We divided projected GDP values by AEO's
population projections to estimate per capita GDP, again maintaining internal consistency in
the calculation.
BenMAP User Manual v 0.01
H-27
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
3. Calculating changes in WTP
The income elasticity estimates from Table H-ll and the estimated changes in future income
may then be used to estimate changes in future WTP for each health effect. The adjustment
formula follows four steps:
Equation H-2
A WTP
1 ~WTP _ (WTP2-WTP1)xQ2+I1)
(i2-i1)x(wtp2+wtp1)
2. sI2WTP2 + eI2WTP1 - eI1WTP2 - sI1WTP1 = I2WTP2 + I1WTP2 - I2WTP1 -
I1WTP1
3. WTP2 x ( s/2 — s/i - I2 - /x) = WTPX x ( s/i - e/2 - h - 12)
4. WTP2 = WTP1 X Efi-gf2-fi-f2
£/2_ £^1" ^2 ~~ ^1
Table H-15 summarizes the income-based WTP adjustments used within BenMAP for minor
health effects, severe health effects, and premature mortality. BenMAP applies the "mid"
income growth adjustment to the WTP for each corresponding health effect. The "low" and
"upper" are provided for bounding the "mid" estimate.
BenMAP User Manual v 0.01
H-28
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
Table H-45. Income-Based WTP Adjustments by Health Effect and Year
Minor Health Endpoint
Severe Health Endpoint
Mortality
Year
Low
Mid
Upper
Low
Mid
Upper
Low
Mid
Upper
1990
1
1
1
1
1
1
1
1
1
1991
0.999425
0.997845
0.995695
0.996411
0.99355
0.991409
0.99885
0.994264
0.985722
1992
1.000278
1.001043
1.002086
1.001738
1.003131
1.004177
1.000556
1.002783
1.006971
1993
1.000845
1.003171
1.006353
1.005291
1.009545
1.012746
1.00169
1.00848
1.021335
1994
1.001941
1.007299
1.014651
1.012194
1.022057
1.029519
1.003886
1.019583
1.049686
1995
1.002529
1.009516
1.019122
1.01591
1.028821
1.038614
1.005064
1.025578
1.065196
1996
1.003545
1.01336
1.026899
1.022366
1.040622
1.054532
1.007103
1.036027
1.092569
1997
1.004809
1.018155
1.036642
1.030442
1.055472
1.074652
1.009642
1.049157
1.127596
1998
1.006098
1.023062
1.046661
1.038734
1.070818
1.095552
1.012234
1.062703
1.164493
1999
1.007498
1.02841
1.057639
1.047803
1.087723
1.1187
1.015052
1.077598
1.205983
2000
1.008676
1.032928
1.066959
1.055489
1.102148
1.138556
1.017428
1.090286
1.242108
2001
1.008675
1.032924
1.06695
1.055482
1.102134
1.138537
1.017426
1.090274
1.242075
2002
1.008984
1.03411
1.069403
1.057504
1.105943
1.143797
1.018048
1.093621
1.251727
2003
1.009739
1.037019
1.075433
1.062468
1.115325
1.15678
1.019574
1.101858
1.275708
2004
1.010863
1.041352
1.084451
1.069884
1.129408
1.176348
1.021844
1.114208
1.312269
2005
1.011864
1.045228
1.092549
1.076534
1.142112
1.194079
1.02387
1.125332
1.345838
2006
1.012602
1.04809
1.09855
1.081456
1.151559
1.207315
1.025363
1.133594
1.371175
2007
1.012957
1.049471
1.101452
1.083835
1.156139
1.213747
1.026084
1.137596
1.383575
2008
1.012532
1.047821
1.097984
1.080992
1.150667
1.206063
1.025223
1.132814
1.368769
2009
1.011166
1.042525
1.086897
1.071894
1.13324
1.181689
1.022457
1.117566
1.322336
2010
1.011843
1.045146
1.092377
1.076393
1.141841
1.1937
1.023827
1.125095
1.345117
2011
1.012165
1.046395
1.094994
1.07854
1.145958
1.199463
1.024479
1.128697
1.356115
2012
1.012764
1.04872
1.099873
1.082541
1.153646
1.210245
1.025692
1.135418
1.376817
2013
1.013213
1.050467
1.103547
1.085551
1.159447
1.2184
1.026602
1.140487
1.392581
2014
1.013911
1.053184
1.109275
1.090241
1.168516
1.231181
1.028017
1.148402
1.417472
2015
1.014754
1.056473
1.116227
1.095927
1.179559
1.246797
1.029727
1.158031
1.4482
2016
1.015112
1.057873
1.119194
1.098352
1.184283
1.253496
1.030454
1.162147
1.461489
2017
1.015291
1.058571
1.120676
1.099562
1.186645
1.25685
1.030817
1.164204
1.468167
2018
1.016092
1.061712
1.127353
1.105013
1.197312
1.272029
1.032446
1.173487
1.498591
2019
1.016945
1.06506
1.134495
1.110837
1.208762
1.288383
1.03418
1.183439
1.531756
2020
1.017494
1.067221
1.139119
1.114604
1.216197
1.299038
1.035298
1.189894
1.553577
2021
1.017858
1.068652
1.142185
1.1171
1.221138
1.306133
1.036037
1.194181
1.568204
2022
1.018224
1.070097
1.145287
1.119624
1.226144
1.313335
1.036784
1.198523
1.583131
2023
1.018585
1.071523
1.148352
1.122117
1.231098
1.320474
1.03752
1.202817
1.598006
2024
1.018991
1.073127
1.151804
1.124923
1.236689
1.328546
1.038347
1.20766
1.61492
2025
1.019446
1.074925
1.155682
1.128074
1.242981
1.337648
1.039273
1.213106
1.634116
2026
1.019915
1.076785
1.159702
1.131337
1.249515
1.347122
1.040231
1.218759
1.654237
2027
1.02041
1.078749
1.163955
1.134787
1.256443
1.357191
1.041242
1.224747
1.675778
2028
1.020937
1.080842
1.168495
1.138469
1.263858
1.367993
1.042318
1.231152
1.699074
2029
1.021436
1.082827
1.17281
1.141964
1.270919
1.378307
1.043337
1.237246
1.721494
2030
1.021959
1.08491
1.17735
1.145639
1.278367
1.389214
1.044406
1.243669
1.745395
2031
1.022479
1.086985
1.181879
1.149303
1.285815
1.400148
1.045469
1.250087
1.76956
2032
1.022969
1.088943
1.186166
1.152768
1.292879
1.410546
1.046472
1.256169
1.792727
2033
1.023487
1.091016
1.190711
1.15644
1.300389
1.421627
1.047532
1.262629
1.817622
2034
1.024004
1.093088
1.195265
1.160115
1.307929
1.432783
1.048592
1.269111
1.842903
2035
1.024493
1.095052
1.199589
1.163603
1.315106
1.443429
1.049594
1.275275
1.867234
2036
1.02498
1.097006
1.203902
1.16708
1.322282
1.454101
1.050592
1.281434
1.891828
2037
1.025478
1.099009
1.208332
1.170647
1.329668
1.465113
1.051613
1.287768
1.917424
2038
1.025984
1.101051
1.212857
1.174289
1.33723
1.476418
1.052652
1.294248
1.943931
2039
1.026492
1.103099
1.217405
1.177947
1.34485
1.487839
1.053694
1.300771
1.970954
2040
1.027004
1.105171
1.222018
1.181654
1.352597
1.499482
1.054748
1.307398
1.998757
2041
1.027508
1.107212
1.226569
1.185309
1.36026
1.511031
1.055784
1.313948
2.026591
2042
1.02803
1.109326
1.231295
1.189101
1.368236
1.523084
1.056857
1.320759
2.055917
2043
1.028555
1.111457
1.236071
1.19293
1.376316
1.535329
1.057937
1.327653
2.086006
2044
1.029075
1.113573
1.240821
1.196736
1.384373
1.547576
1.059008
1.334521
2.116397
2045
1.029605
1.115728
1.245672
1.20062
1.392623
1.560152
1.060099
1.341548
2.147923
2046
1.030116
1.117812
1.250373
1.20438
1.400637
1.572403
1.061152
1.348368
2.178951
2047
1.030638
1.119946
1.255197
1.208236
1.408882
1.585045
1.06223
1.355378
2.211295
2048
1.031135
1.121978
1.259801
1.211913
1.416771
1.597176
1.063255
1.36208
2.242655
2049
1.03162
1.123965
1.264312
1.215513
1.424519
1.609122
1.064256
1.368656
2.273847
2050
1.032088
1.125883
1.268677
1.218994
1.432035
1.620744
1.065222
1.37503
2.304491
BenMAP User Manual v 0.01
H-29
December 2022
-------
Appendix H: Core Health Valuation Functions in BenMAP
H.6 Inflation Indices
Chapter4 of the BenMAP User Manual provides instructions forformatting and adding
inflation data. These values are used to adjust economic values to express monetary units in
a consistent dollaryear. As discussed in that chapter, BenMAP includes inflation factors for
three different types of values. The source for these values is included in Table H-16. These
values were re-indexed to $2015 priorto import in BenMAP.
Table H-46. Inflation Factors
Name
Description
Years
Source
All Goods
Index
Value of generic
goods and services
BLS, Data Series CUUR0000SA0 at
httD://data.bls.eov/cei-bin/survevmost?cu
Medical Cost
Index
Value of medical
expenses
1980-2020
BLS, Data Series CUUR0000SAM at
http://data.bls.eov/cei-bin/survevmost7cu
Wage Index
Value of wages
BLS, Employment Cost Trends. Table 5 at
httD://www.bls.eov/web/eci/ecicois.txt
BenMAP User Manual v 0.01
H-30
December 2022
-------
Appendix I: Additional Health Valuation Functions in BenMAP
Appendix I. Additional Health Valuation
Functions in BenMAP
In this Appendix, we present additional health valuation functions that are included in
BenMAP but are not currently used regularly in U.S. EPA regulatory analyses. PLACEHOLDER:
Information on additional functions will be included here as they are added to the tool
by EPA.
BenMAP User Manual v 0.01
1-1
December 2022
-------
Appendix J: Population & Other Data in BenMAP
Appendix J. Population & Other Data in
BenMAP
This section describes the population, monitor, and demographic data forthe United States
included in BenMAP. It consists of the following three subsections:
Population Data. This describes how BenMAP forecasts population; the block-level and
county-level data underlying the forecasts; and the PopGrid software application, which
aggregates block-level population data to whatever grid definition might be needed. The
population data currently in the web tool through year 2050 relies on the same procedures
described in this Appendix. The web tool adds a population projection for 2055 based on a
linear extrapolation of the projection trend through 2050 from the desktop tool.
Implementation of user-input population data and projections, e.g., using PopGrid tool, will
be included in a future release of the web tool.
Monitor Data. The current version of the web tool does not support the use of monitor data
for air quality surfaces. Support for monitor data will be included in a future release.
Demographic Datasets. This subsection describes the various U.S. datasets in BenMAP related
to demography: household size, poverty rates, and educational attainment. The web tool
currently includes the full set of data described in this appendix; however the variables used
by the tool are limited to those included in the default set of Health Impact Functions.
Broader user access to these variables and more complete integration with user-specified
health impact functions will be implemented in a future version.
BenMAP User Manual v 0.01
J-l
December 2022
-------
Appendix J: Population & Other Data in BenMAP
J.i Population Data for the U.S.
BenMAP calculates health impacts for any desired grid definition, so long as you have a
shapefile forthat grid definition and population data for that grid definition. In this
description, we use the term "population grid cell" to refer to a cell (e.g., county) within a grid
definition. The foundation for calculating the U.S. population in the population grid-cells is
2010 Census block data. A separate application called "PopGrid," described below, combines
the Census block data with any user-specified set of population grid- cells, so long as they are
defined by a GIS shape file. Unfortunately, PopGrid relies on extremely large census files that
are too large to include with BenMAP - hence the need for the separate application. The
PopGrid program is available on the BenMAP website here: www.epa.gov/benmap
Within any given population grid-cell, BenMAP has 304 unique race-ethnicity-gender-age
groups: 19 age groups by 2 ethnic groups by gender by 4 racial groups (19*2*2*4=304). Table
J-l presents the 304 population variables available in BenMAP. As discussed below, these
variables are available for use in developing age estimates in whatever grouping you require.
Table J-47. Demographic Groups and Variables Available in BenMAP
Racial Group
Ethnicity
Age
Gender
White, African
American, Asian,
American Indian
Hispanic,
Non-Hispanic
<1,1-4,5-9,10-14,15-19, 20-24, 25-29,
30-34,35-39, 40-44, 45-49, 50-54,55-59,
60-64, 65-69, 70-74, 75-79, 80-84, 85+
Male, Female
In this section on U.S. population data in BenMAP, we describe:
Forecasting Population. This describes how BenMAP forecasts population.
Data Needed. This section describes the block-level and county-level data underlying the
forecasts.
BenMAP User Manual v 0.01
J-2
December 2022
-------
Appendix J: Population & Other Data in BenMAP
PopGrid. This section reviews the PopGrid software application, which aggregates block-level
population data to whatever grid definition might be need.
J.1.1 How BenMAP Forecasts Population
In calculatingthe population in age groups that may include a portion of one of the pre-
specified demographic groups in Table J-l, BenMAP assumes the population is uniformly
distributed in the age group. For example, to calculate the number of children ages 3 through
12, BenMAP calculates:
To estimate population levels forthe years after the last Census in 2010, BenMAP scales the
2010 Census-based estimate with the ratio of the county-level forecast for the future year of
interest overthe 2010 county-level population level. Woods & Poole (2015) provides the
county-level population forecasts used to calculate the scaling ratios; these data are
discussed in detail below.
In the simplest case, where one is forecasting a single population variable, say, children ages
4 to 9 in the year 2020, BenMAP calculates:
Equation J-l
1
3
a^e3-i2 = - x agex_ 4 + age5_9 + - x agew_u
Equation J-2
aSe4-9,g, 2020 — aSe4-9,g, 2010 X
age 4
-9, county, 2020
age 4
-9, county, 2010
Where the gth population grid-cell is wholly located within a given county.
BenMAP User Manual v 0.01
December 2022
J-3
-------
Appendix J: Population & Other Data in BenMAP
In the case, where the gth grid-cell includes "n" counties in its boundary, the situation is
somewhat more complicated. BenMAP first estimates the fraction of individuals in a given
age group (e.g., ages 4 to 9) that reside in the part of each county within the gth grid-cell.
BenMAP calculates this fraction by simply dividing the population all ages of a given county
within the gth grid-cell by the total population in the gth grid-cell:
Equation J-3
. . . . aSeaU, e in county.
fraction of age nc = ^
aSeaii,g
Multiplying this fraction with the number of individuals ages 4 to 9 in the year 2010 gives an
estimate of the number of individuals ages 4 to 9 that reside in the fraction of the county
within the gth grid-cell in the year 2010:
Equation J-4
age4
—9,gin countyc,2010
= age4
-9,g,2010
x fraction age4_9_giucoimty,c
To then forecast the population in 2020, we scale the 2010 estimate with the ratio of the
county projection for 2020 to the county projection for 2010:
Equation J-5
age* —9, countyc, 2020
age4
—9,s in countyr, 2020
= age4
—9,s in countyr ,2010 ^
aSeA-9, cow,,2010
Combining all these steps for "n" counties within the gth grid-cell, we forecast the population
of persons ages 4 to 9 in the year 2020 as follows:
Equation J-6
n
-9,£,2020 age* -9,£,2010
total pop:iin :ounly age4
—9, countyc , 2020
,=i total popg age4_9 >2010
BenMAP User Manual v 0.01
J-4
December 2022
-------
Appendix J: Population & Other Data in BenMAP
In the case where there are multiple age groups and multiple counties, BenMAP first
calculates the forecasted population level for individual age groups, and then combines the
forecasted age groups. In calculating the number of children ages 4 to 12, BenMAP calculates:
Equation J-7
n
a§e4-9,g,2020 ~ ^"1 a§S4-9 g 7.010
tOtdl pOpg countyc —9, countyc, 2020
*=i total popg age4_9 cmmt},c 20m
aSe\0-\4,g,2020 aSe
total pop:j in conniy age 1Q~1410_14 2020
c=i total popg agel, , ]4 5 2oi o
10-14,£,2010 '
a8e4-12,g,2020 aSe4-9,g,2020 + ^ XClSeiO-14, g
2020
To estimate population for 2055, we extrapolate Woods and Poole projections from the 2045
to 2050 period:
Equation J-8
,.r ... ^2050, i
•^2055,1 — ^2050,i *777
""2045,i
where 1/1/is the growth weight (relative to 2010) and /' is each demographic cell (i.e., unique
combinations of county, gender, ethnicity, race, and age range).
J.1.2 Data Needed for Forecasting
Underlying the population forecasts in BenMAP there are block-level databases used to
provide year 2010 population estimates and a county-level database of forecast ratios. Both
files have the same set of 304 race-ethnicity-gender-age population groups.
BenMAP User Manual v 0.01
J-5
December 2022
-------
Appendix J: Population & Other Data in BenMAP
The block-level data is typically not used directly in BenMAP, and instead is used with the
PopGrid software (described below) to provide year 2010 estimates for a grid definition of
interest (e.g., 12 kilometer CMAQ grid). The output from PopGrid with the year 2010
population estimates can then be loaded into BenMAP.
The county-level data comes pre-installed and is not something that you need to load
yourself. These data are simply county-level ratios of a year (2009,2011-2050) and year 2010
population data for each county and each of the 304 race-ethnicity-gender-age population
groups.
We describe the development of each databases below.
J.1.2.1Block-Level Census 2010
There are about five million "blocks" in the United States, and for each block we have 304
race-ethnicity-gender-age groups. The block-level population database is created separately
for each state, in order to make the data more manageable. (A single national file of block
data would be about six gigabytes.)
The initial block file from the U.S. Census Bureau is not in the form needed. The block data
has 7 racial categories and 23 age groups, as opposed to the 4 and 19 used in BenMAP. Table
J-2 summarizes the initial set of variables and the final desired set of variables.
BenMAP User Manual v 0.01
J-6
December 2022
-------
Appendix J: Population & Other Data in BenMAP
Table J-48. Race, Ethnicity and Age Variables in 2010 Census Block Data
Type
Race
Ethnicity
Gender
Age
Initial
Variables
(SF1 file)
White Alone, Black Alone,
Native American Alone, Asian
Alone, Pacific
Islander/Hawaiian Alone,
Other Alone, Two or More
Alone
Male,
Female
0-4, 5-9,10-14,15-17,18-19,
20, 21, 22-24, 25-29, 30-34,
35-39,40-44,45-49,50-54,
55-59, 60-61, 62-64, 65-66,
67-69, 70-74, 75-79, 80-84,
85+
Final
Desired
Variables
White, African-American,
Asian-American, Native-
American
Hispanic,
Non-
Hispanic
Female,
Male
<1,1-4, 5-9,10-14,15-19, 20-
24, 25-29, 30-34, 35-39, 40-
44,45-49, 50-54, 55-59, 60-
64, 65-69, 70-74, 75-79, 80-
84, 85+
The initial set of input files are as follows.
Census 2010 block-level and tract-level files (Summary File 1)
Data: http://www2.census.gov/census_2010/04-Summary_File_l/
Docs: http://www.census.gov/prod/cen2010/doc/sfl.pdf
Census 2000 MARS national-level summary
Docs: http://www.census.gov/popest/archives/files/MRSF-01-USl.pdf
The SF1 tract-level and MARS data, as described below, are needed to reorganize the
variables that come initially in the block-levelSFl file. (For the sake of completeness, we note
that there exists a county-level Census 2000 MARS file; however, due to major population
count discrepancies between the county-level MARS file and block-level SF1 file, we used only
the nation-level summary table. Tables in MARS documentation file did not have the
discrepancies that the county-level file had. We were unable to get an adequate explanation
of this from the U. S. Census.)
The steps in preparing the data are as follows:
BenMAP User Manual v 0.01
J-7
December 2022
-------
Appendix J: Population & Other Data in BenMAP
1. Adjust Age-classifications:
We combined some age groups in the block-level SF1 data to match the age groups wanted
for BenMAP. For example, we combined age groups 15-17 and 18-19 to create the 15-19 age
group used in BenMAP. Then, in the case of the 0-4 age group, we split it into <1 and 1-4 using
the tract-level SF1 data, which gave us the fraction of 0-4 year-olds who are <1.
2. Fill in Missing Racial-Ethnic Interactions:
We used the tract-level SF1 data to calculate the fraction of Hispanics in each ethnically-
aggregated subpopulation from the block-level data, by age and sex. We used these fractions
to distribute each age-sex-race-block-level datum into Hispanics and non-Hispanics.
3. Assign "Other" and "Multi-Racial" to the Remaining Four Racial Categories:
We assign the "Other" race category in two steps. First, based on the national MARS data, we
estimated how many people in the "multi-racial" category checked off "some other race" as
one of their races, for Hispanics and non-Hispanics separately. In each age-sex-race-block-
level datum, we added those people to "other race" category to create the re-distribution
pool, analogously to the method implemented by Census while creating MARS data (see U.S.
Census Bureau, 2002a, Table 1, below). Second, based on the national re-allocation fractions
for Hispanics and non-Hispanics (derived from the MARS data), we assigned the "Other" race
into the four races of interest and "multi-race".
Afterthe assignment of the "Other" race category, we then assigned "multi-racial" category
to the four racial categories, using state fractions of these races in each age-sex-race-block-
level datum.
BenMAP User Manual v 0.01
J-8
December 2022
-------
Appendix J: Population & Other Data in BenMAP
Exhibit J-3: Summary of Modified Race rind Census 2000 Rate Distributions for the United States
Subject
Modified Race
{'entiii 2000
Numhtr
tVrfiw
Number
Percenr
lOTAl POP! IVttON
291,421.906
moo
2*1,421.906
100.00
One rare
277,524,2 26
91,62
274,595,678
9757
Spocilicd race acly
277,524,236
98.62
259,236.605
92.12
Wfcalc
228,104,485
8105
211 460,626
75.14
Waclc or Atncac American
.15,704,124
12.69
34,6 58 J 90
12J2
IjMut and Alinb Native
2.663,*18
0.95
2.475.956
OA*
A*an
10.589.265
3.76
10.242.998
3.64
Native Hawaiian and Othor Pacific Manler
.162,534
0.16
398.835
0.14
Nor.-specified raoc octy
(X)
15359,073
5,46
Ivie rtcti
3,578,05 3
1.27
MW.075
226
Spocilicd face ody
3,578,053
1.27
3.366,517
120
Specified.and ncei-spraticd raocs
-------
Appendix J: Population & Other Data in BenMAP
Native-American and for all Hispanics. The detailed documentation can be found at
http://www. woodsandpoole.com/pdfs/CED15.pdf. As discussed below, the adjustments
necessary to prepare the data for use in BenMAP are relatively straightforward.
For each non-Hispanicsubset of the population and eachyearfrom 2000-2050, we divided
the Woods and Poole population for that year by the Woods and Poole population for that
subset in 2010. These serve as the growth coefficients for the non-Hispanic subsets of each
race. We used a similar calculation to determine the growth rates for the Hispanic population.
We assume that each Hispanic race grows at the same rate, and use these growth rates for the
Hispanic subsets of each race.14
Matching Age Groups Used in BenMAP
There are 86 age groups, so it is a simple matter of aggregating age groups to match the 19
used in BenMAP.
Matching Counties Used in U.S. Census
The county geographic boundaries used by Woods & Poole are somewhat more aggregated
than the county definitions used in the 2010 Census and those in BenMAP, and the FIPS codes
used by Woods and Poole are not always the standard codes used in the Census. To make the
Woods and Poole data consistent with the county definitions in BenMAP, we disaggregated
the Woods and Poole data and changed some of the FIPS codes to match the U.S. Census.
14 Previous versions of the BenMAP program used a different methodology whereby population estimates for 2000
- 2009 were adjusted using the ratio of 2000 Woods & Poole estimated population and 2000 Census population.
BenMAP User Manual v 0.01
December 2022
J-10
-------
Appendix J: Population & Other Data in BenMAP
Calculating Growth Ratios with Zero Population in 2010
There are a small number of cases were the 2010 county population for a specific
demographic group is zero, so the ratio of any future year to the year 2010 data is undefined.
In these relatively rare cases, we prepared statewide and national totals and used ratios at
the higher levels of geographic aggregation when the more local ratios caused divide-by-zero
errors.
J.2 U.S. Demographic Datasets in BenMAP
BenMAP includes county-level data on household size, poverty status, educational
attainment, unemployment, health insurance coverage, and occupational status. We
describe the data sources and processing methodology for each dataset below. All estimates
were generated at the county level for3,109 counties in the contiguous United States.15
J.2.1 Household Size
To generate average household size for each county, we utilize ACS 5-year estimates for 2012
to 2016. Average household size was provided by ACS at the county level for all counties
except for two, for which data were not available.16 For these counties, we applied the state
level average household size.
15ln 2013, Bedford city, Virginia was removed from the list of counties in the U.S. Due to BenMAP's grid definition,
we continue to include Bedford city (FIPS code 51515) in this update. We impute the value for this county using the
value for the county with which it was combined (Bedford County, FIPS code 51019). In 2015, Oglala Lakota County,
South Dakota (46102) changed name and code from Shannon County (46113). To match BenMAP's grid definition,
we use the old FIPS code (46113) for this county. For further information, please see the Census website:
https://www.census.gov/programs-survevs/geographv/technical-documentation/county-changes.html
16The two counties without data were Shannon County, South Dakota (FIPS Code 46113) and Bedford, Virginia (FIPS
Code 51515).
BenMAP User Manual v 0.01
J-ll
December 2022
-------
Appendix J: Population & Other Data in BenMAP
J.2.2 Educational Attainment
We use data from the ACS to provide county-level summaries of educational attainment.
These data represent 5-year average ACS estimates from 2015 to 2019. Specifically, the data
included in BenMAP span two broad education categories: no high school diploma (termed
"no_hs_degree"), and high school diploma (or equivalency) and above (termed
"hs_degree_plus"). The latter category includes individuals with a high school diploma (or
equivalency), some college, college degree, or post-graduate degree and ranges from 53-
98%.
For both education groups (with/without HS diploma), we estimate the fraction of the total
county population (ages 25 years and above) in each education group. Thus, the two
estimates sum to one for each county.
For each estimate, we generate a coefficient of variation (CV) equal to the ratio of the
standard error to the point estimate. For counties with a CV greater than 0.3 (1.7% percent of
all counties), we impute the county-level estimate with a state-level estimate following
Census guidance, which defines any estimate with a CV greater than 0.3 as low reliability and
to be used with extreme caution (King et al. 2015).
J.2.3 Poverty Status
To determine the poverty status at the county level, we utilize ACS 5-year estimates from
2015 to 2019. The resulting datasets represent the fraction of the total population in the
county that falls below the federal poverty line (termed "below_poverty_line") and the
fraction of the population that falls above the poverty line (termed "above_poverty_line").
The EPA Standard Variables dataset also includes two variables (termed
"below_2x_poverty_line" and "above_2x_poverty_line") representing the fraction of the
county-level population below and above 200% of the poverty line. The county-level
BenMAP User Manual v 0.01
J-12
December 2022
-------
Appendix J: Population & Other Data in BenMAP
proportions below the poverty line range from 3-55% and below 200% of the poverty line
range from 8-75%.
For each estimate, we generate a coefficient of variation (CV) equal to the ratio of the
standard error to the point estimate. For counties with a CV greater than 0.3 (1.7% percent of
all counties), we impute the county-level estimate with a state-level estimate following
Census guidance, which defines any estimate with a CV greater than 0.3 as low reliability and
to be used with extreme caution (King et al. 2015).
J.2.4 Unemployment Rates
BenMAP includes county-level employment variables representing average rates from 2017 to
2021. Importantly, the employment variables are adjusted for use in BenMAP to use total
population as the denominator rather than labor force. This allows you to multiply the rates
by the total population (in a health impact function) to assess populations that are (a)
employed, (b) unemployed, and (c) not part of the labor force (e.g., retirees, students,
discouraged workers).
County-level unemployment rates are from the Bureau of Labor Statistics. We adjusted these
rates using county-level population estimates from the U.S. Census Bureau from 2017 to
2021. The calculations were done for each year individually and then averaged together to
create a five-year average. The value calculated in BenMAP represents the average rate across
all months within this period. The rate of residents not in the labor force was calculated by
subtracting the labor force from the total population and then dividing the remainder by the
total population.
BenMAP User Manual v 0.01
J-13
December 2022
-------
Appendix J: Population & Other Data in BenMAP
J.2.5 Health Insurance
We use data from the Small Area Health Insurance Estimates (SAHIE) collected by the U.S.
Census Bureau from 2015 to 2019 to calculate the percentage of individuals with and without
health insurance in each county. The SAHIE date provides the number of individuals with and
without health insurance by county. Calculations were done for each year individually and
then averaged together to create a five-year average.
J.2.6 Blue Collar Workers
We use five-year estimates (2012-2016) from the ACS to estimate the fraction of each county's
labor force employed in white collar and blue collar occupations. The dataset includes the
number of employed individuals over 16 that work within five occupation categories. We
assign each of these five occupations to either the blue collar or white collar designation, as
shown in Table J-3.
Table J-49. Mapping Occupations to Blue Collar and White Collar Designations
Occupation
Designation
Management, business, science and arts
White collar
Service
White collar
Sales and office
White collar
Natural resources, construction and maintenance
Blue collar
Production, transportation and material moving
Blue collar
We calculate the fraction of each county in blue collar professions by dividing the total
number of individuals employed in blue collar jobs by the total number of employed
individuals within each county. The same calculation is done for white collar professions. We
adjust FIPS codes to match the BenMAP county grid definition, as described in section J.3.5.
BenMAP User Manual v 0.01
J-14
December 2022
-------
Appendix K: Uncertainty & Pooling
Appendix K. Uncertainty & Pooling
This Appendix discusses the treatment of uncertainty in BenMAP, both for incidence changes
and associated dollar benefits. Some background is then given on pooling methodology.
Pooling functionality is not implemented in the current version of the tool, though we
expect it to be addressed in a future version.
K.1 Uncertainty
Although there are several sources of uncertainty affecting estimates of incidence changes
and associated benefits, the sources of uncertainty that are most readily quantifiable in
benefit analyses are uncertainty surrounding the health impact functions and uncertainty
surrounding unit dollar values. The total dollar benefit associated with a given health effect
depends on how much the health effect will change in the control scenario (e.g., how many
premature deaths will be avoided) and how much each unit of change is worth (e.g., how
much a statistical death avoided is worth).
Both the uncertainty about the incidence changes and uncertainty about unit dollar values
can be characterized by distributions. Each "uncertainty distribution" characterizes our
beliefs about what the true value of an unknown (e.g., the true change in incidence of a given
health effect) is likely to be, based on the available information from relevant studies.
Although such an "uncertainty distribution" is not formally a Bayesian posterior distribution,
it is very similar in concept and function (see, for example, the discussion of the Bayesian
approach in Kennedy 1990, pp. 168-172). Unlike a sampling distribution (which describes the
possible values that an estimator of an unknown value might take on), this uncertainty
distribution describes our beliefs about what values the unknown value itself might be.
BenMAP User Manual v 0.01
K-l
December 2022
-------
Appendix K: Uncertainty & Pooling
Such uncertainty distributions can be constructed for each underlying unknown (such as a
particular pollutant coefficient for a particular location) or for a function of several underlying
unknowns (such as the total dollar benefit of a regulation). In either case, an uncertainty
distribution is a characterization of our beliefs about what the unknown (or the function of
unknowns) is likely to be, based on all the available relevant information. Uncertainty
statements based on such distributions are typically expressed as 90 percent credible
intervals. This is the interval from the fifth percentile point of the uncertainty distribution to
the ninety-fifth percentile point. The 90 percent credible interval is a "credible range" within
which, according to the available information (embodied in the uncertainty distribution of
possible values), we believe the true value to lie with 90 percent probability. The uncertainty
surrounding both incidence estimates and dollar benefits estimates can be characterized
quantitatively in BenMAP. Each is described separately below.
K.l.l Characterization of Uncertainty Surrounding Incidence Changes
To calculate point estimates of the changes in incidence of a given adverse health effect
associated with a given set of air quality changes, BenMAP performs a series of calculations at
each grid-cell. First, it accesses the health impact functions needed forthe analysis, and then
it accesses any data needed by the health impact functions. Typically, these include the grid-
cell population, the change in population exposure at the grid-cell, and the appropriate
baseline incidence rate. BenMAP then calculates the change in incidence of adverse health
effects for each selected health impact function. The resulting incidence change is stored,
and BenMAP proceeds to the next grid-cell, where the above process is repeated.
In Latin Hypercube mode, BenMAP reflects the uncertainty surrounding estimated incidence
changes (resulting from the sampling uncertainty surrounding the pollutant coefficients in
the health impact functions used) by producing a distribution of possible incidence changes
ratherthan a single point estimate. To do this, it uses the distribution (DistBeta) associated
with the pollutant coefficient [Beta, or 3), and potentially the point estimate [Beta) and two
BenMAP User Manual v 0.01
K-2
December 2022
-------
Appendix K: Uncertainty & Pooling
parameters (PIBeta, P2Beta). Typically, pollutant coefficients are normally distributed, with
mean Beta and standard deviation PIBeta.
BenMAP uses an N-point Latin Hypercube to represent the underlying distribution of 3 and to
create a corresponding distribution of incidence changes in each population grid cell, where
N is specified by you. The Latin Hypercube method represents an underlying distribution by N
percentile points of the distribution, where the nth percentile point is equal to:
Equation K-l
, „ 100 100
(n-l)x 1
N 2N
The Latin Hypercube method is used to enhance computer processing efficiency. It is a
sampling method that divides a probability distribution into intervals of equal probability,
with an assumption value for each interval assigned according to the interval's probability
distribution. Compared with conventional Monte Carlo sampling, the Latin Hypercube
approach is more precise over a fewer number of trials because the distribution is sampled in
a more even, consistent manner (Decisioneering, 1996, pp. 104-105).
Suppose, for example, that you elect to use a 20-point Latin Hypercube. BenMAP would then
represent the distribution of (3 by 20 percentile points, specifically the 2.5th, 7.5th,..., 97.5th.
To do this, the inverse cumulative distribution function specified by the distribution of P is
called with the input probability equal to each the 20 percentile points. BenMAP then
generates an estimate of the incidence change in a grid-cell for each of these values of (3,
resulting in a distribution of N incidence changes. This distribution is stored, and BenMAP
proceeds to the next population grid-cell, where the process is repeated.
BenMAP User Manual v 0.01
K-3
December 2022
-------
Appendix K: Uncertainty & Pooling
K.1.2 Characterization of Uncertainty Surrounding Dollar Benefits
The uncertainty distribution of the dollar benefits associated with a given health or welfare
effect is derived from the two underlying uncertainty distributions - the distribution of the
change in incidence of the effect (number of cases avoided) and the distribution of the value
of a case avoided (the "unit value"). The derivation of the uncertainty distribution for
incidence change is described above. The distributions used to characterize the uncertainty
surrounding unitvalues are described in detail in the appendix on the Economic Value of
Health Effects. As noted in that Appendix, a variety of distributions have been used to
characterize the uncertainty of unit values, including uniform, triangular, normal, and
Weibull.
To represent the underlying distribution of uncertainty surrounding unit values, a 100-point
Latin Hypercube is generated in the same way described in the previous section for the
distribution of (3. That is, the unit value distribution is represented using the 0.5th, 1.5th,...,
and 99.5th percentile values of its distribution.
A distribution of the uncertainty surrounding the dollar benefits associated with a given
health effect is then derived from Latin Hypercube values generated to represent the change
in incidence and the Latin Hypercube values generated to represent the unit value
distribution. To derive this new distribution, each of the 100 unit values is multiplied by each
of the N incidence change values, yielding a set of 100 * N dollar benefits. These values are
sorted low to high and binned down to a final distribution of N dollar benefit values.
BenMAP User Manual v 0.01
K-4
December 2022
-------
Appendix L: Batch Run Approach
Appendix L. Batch Run Approach
Placeholder: Future versions of the tool will include batch processing functionality, but
this is not yet implemented.
BenMAP User Manual v 0.01
L-l
December 2022
-------
Appendix M: Function Editor
Appendix M. Function Editor
The function editor is used to develop both health impact functions and valuation functions.
This appendix describes the syntax of this editor. This functionality is not implemented in
the current version of the tool, though we expect it to be present in a future version.
BenMAP User Manual v 0.01
M-l
December 2022
-------
References
References
Abbey, D. E., B. L. Hwang, R. J. Burchette, T. Vancuren and P. K. Mills. 1995a. Estimated long-
term ambient concentrations of PMi0 and development of respiratory symptoms in a
nonsmoking population. Arch Environ Health. Vol. 50 (2): 139-52.
Abbey, D. E., B. E. Ostro, F. Petersen and R. J. Burchette. 1995b. Chronic Respiratory
Symptoms Associated with Estimated Long-Term Ambient Concentrations of Fine
Particulates Less Than 2.5 Microns in Aerodynamic Diameter (PM2.5) and Other Air
Pollutants. J Expo Anal Environ Epidemiol. Vol. 5 (2): 137-159.
Abbey, D. E., F. Petersen, P. K. Mills and W. L. Beeson. 1993. Long-Term Ambient
Concentrations of Total Suspended Particulates, Ozone, and Sulfur Dioxide and
Respiratory Symptoms in a Nonsmoking Population. Archives of Environmental Health.
Vol. 48(1): 33-46.
Abt Associates Inc. 2000. Final Heavy Duty Engine/Diesel Fuel Rule: Air Quality Estimation,
Selected Health and Welfare Benefits Methods, and Benefit Analysis Results. Prepared
for U. S. EPA, Office of Air Quality Planning and Standards, Research Triangle Park, NC.
Bethesda, MD. December.
Adams, P. F., G. E. Hendershot and M. A. Marano. 1999. Current Estimates from the National
Health Interview Survey, 1996. Vital Health Stat. Vol. 10 (200): 1-212.
Agency for Healthcare Research and Quality (AHRQ). 2007. Healthcare Cost and Utilization
Project. National Inpatient Sample (NIS) Rockville, Maryland. Available at:
http://www.ahrq. gov/data/hcup/ [accessed February, 2012],
Agency for Healthcare Research and Quality (AHRQ). 2009. HCUPnet, Healthcare Cost and
Utilization Project. Rockville, MD. Available on the Internet at http://hcupnet.ahrq.gov
Alhanti, B.A., Chang, H.H., Winquist, A., Mulholland, J.A., Darrow, L.A. and Sarnat, S.E., 2016.
Ambient air pollution and emergency department visits for asthma: a multi-city
BenMAP User Manual v 0.01
R-l
December 2022
-------
References
assessment of effect modification by age. Journal of exposure science & environmental
epidemiology, 26(2), pp.180-188.
Alzheimer's Association. 2020.2020 Alzheimer's Disease Facts and Figures. Available at:
https://www.alz.org/media/Documents/alzheimers-facts-and-figures.pdf.
American Lung Association. 2002a. Trends in Asthma Morbidity and Mortality. American Lung
Association, Best Practices and Program Services, Epidemiology and Statistics
Unit.http:// www.lungusa.org/data/asthma/ASTHMAdt.pdf.
American Lung Association. 2002b. Trends in Chronic Bronchitis and Emphysema: Morbidity
and Mortality. American Lung Association, Best Practices and Program Services,
Epidemiology and Statistics Unit.http://www.lungusa.org/data/copd/COPDl.pdf.
American Lung Association. 2002c. Trends in Morbidity and Mortality: Pneumonia, Influenza,
and Acute Respiratory Conditions. American Lung Association, Best Practices and
Program Services, Epidemiology and Statistics
Unit. http://www.lungusa.org/data/pi/PI_l.pdf.
American Lung Association. 2010a. Trends in COPD (Chronic Bronchitis and Emphysema):
Morbidity and Mortality. American Lung Association Epidemiology and Statistics Unit
Research and Program Services Division. February. Table 4. Based on NHIS data (CDC,
2008). Available at http://www.lungusa.org/finding-cures/our-research/trend-
reports/copd- trend-report.pdf.
American Lung Association. 2010b. Trends in Asthma Morbidity and Mortality. American Lung
Association Epidemiology and Statistics Unit Research and Program Services Division.
February. Table 7. Based on NHIS data (CDC, 2008). Available at
http://www.lungusa.org/ finding-cures/our-research/trend-reports/asthma-trend-
report.pdf.
Anderson, G.B. and Bell, M.L. 2010. Does one size fit all? The suitability of standard ozone
exposure metric conversion ratios and implications for epidemiology. J Expo Sci Environ
Epidemiol 20(1): 2-11.
BenMAP User Manual v 0.01
R-2
December 2022
-------
References
Babin, S. M., H. S. Burkom, et al. 2007. "Pediatric patient asthma-related emergency
department visits and admissions in Washington, DC, from 2001-2004, and associations
with air quality, socio-economic status and age group." Environ Health 6:9.
Baena-Cagnani, C, Rossi, GA and Canonica, GW. 2007. Airway remodelling in children: When
does it start? Curr Opin Allergy Clin Immunol 7(2): 196-200.
Barry, V., Klein, M., Winquist, A., Chang, H.H., Mulholland, J.A., Talbott, E.O., Rager, J.R.,
Tolbert, P.E., &Sarnat, S.E. 2018. Characterization of the concentration-response curve
for ambient ozone and acute respiratory morbidity in 5 US cities. J Expo Sci Environ
Epidemiol.
Baxter, L.K., Crooks, J.L. and Sacks, J.D. 2017. Influence of exposure differences on city-to-city
heterogeneity in PM2.5-mortality associations in US cities. Environ Health 16(1): 1.
Bell, M. L., K. Ebisu, et al. 2008. "Seasonal and Regional Short-term Effects of Fine Particles on
Hospital Admissions in 202 US Counties, 1999-2005." American Journal of Epidemiology
168 (11): 1301-1310.
Bell, M. L., F. Dominici and J. M. Samet. 2005. A meta-analysis of time-series studies of ozone
and mortality with comparison to the national morbidity, mortality, and air pollution
study. Epidemiology. Vol. 16 (4): 436-45.
Bell, M. L., A. McDermott, S. L. Zeger, J. M. Samet and F. Dominici. 2004. Ozone and short-term
mortality in 95 US urban communities, 1987-2000. Jama. Vol. 292 (19): 2372-8.
Bell, M. L., Son, J. Y., Peng, R. D., Wang, Y., & Dominici, F. 2015. Ambient PM2.5 and Risk of
Hospital Admissions: Do Risks Differ for Men and Women?. Epidemiology, 26(4): 575-
579.
Belova, A., E. Post and D. McCubbin. 2007. Using Expert Distributions in BenMAP.
Memorandum prepared for Bryan Hubbell at U.S. EPA, Office of Air Quality Planning and
Standards. Prepared by Abt Associates, Bethesda, MD. March 7.
BenMAP User Manual v 0.01
R-3
December 2022
-------
References
Belova, A, Fann, N, Haskell, J, Hubbell, B and Narayan, T. 2020. Estimating Lifetime Cost of
Illness. An Application to Asthma. Ann Am Thorac Soc 17(12): 1558-1569.
Bhattacharyya, N. 2011. Incremental healthcare utilization and expenditures for allergic
rhinitis in the United States. The Laryngoscope 121(9): 1830-1833.
Blumenschein, K. and M. Johannesson. 1998. Relationship between quality of life
instruments, health state utilities, and willingness to pay in patients with asthma. Ann
Allergy Asthma Immunol. Vol. 80 (2): 189-94.
Burnett, R. T., M. Smith-Doiron, D. Stieb, M. E. Raizenne, J. R. Brook, R. E. Dales, J. A. Leech, S.
Cakmak and D. Krewski. 2001. Association between ozone and hospitalization for acute
respiratory diseases in children less than 2 years of age. Am J Epidemiol. Vol. 153 (5):
444-52.
Burnett, R. T., C. Arden Pope III, M. Ezzati, C. Olives, S. S. Lim, S. Mehta, H. H. Shin, G. Singh, B.
Hubbell, M. Brauer, H. R. Anderson, K. R. Smith, J. R. Balmes, N. G. Bruce, H. Kan, F.
Laden, A. Pruss-Ustun, M. C. Turner, S. M. Gapstur, W. R. Diver, and A. Cohen. 2014. An
integrated risk function for estimating the global burden of disease attributable to
ambient fine particulate matter exposure. Environmental Health Perspectives. Vol. 122
(4): 397-403.
Burnett, R., Chen, H., Szyszkowicz, M., Fann, N., Hubbell, B., Pope, C.A., Spte, J.S., Brauer, M.,
Cohen, A., Weichenthal, S., Coggins, J., Qian, D., Brunekreef, B., Frostad, J., Lim, S.S.,
Kan, H., Walker, K.D., Thurston, G.D., Hayes, R.B., Lim, C.C., Turner, M.C., Jerrett, M.,
Krewski, D., Gapstur, M., Diver, W.R., Ostro, B., Goldberg, D., Crouse, D.L., Martin, R.V.,
Peters, P., Pinault, L., Tjepkema, M., van Donkelaar, A., Villeneuve, P.J., Miller, A.B., Yin,
P., Zhou, M., Wang, L., Janssen, N.A.H., Marra, M., Atkinson, R.W., Tsang, H., Thach, T.Q.,
Cannon, J.B., Allen, R.T., Hart, J.E., Laden, F., Cesaroni, G., Forastiere, F., Weinmayr, G.,
Jaensch, A., Nagel, G., Concin, H., Spadaro, J.V. 2018. Global estimates of mortality
associated with long-term exposure to outdoor fine particulate matter. Proceedings of
the National Academy of Sciences. Vol. 115 (38): 9592-9597.
BenMAP User Manual v 0.01
R-4
December 2022
-------
References
Cakmak, S., Dales, R., Judek, S. 2006. Respiratory Health Effects of Air Pollution Gases:
Modification by Education and Income. Archives of Environmental & Occupational
Health, 61(1), 5-10.
Centers for Disease Control and Prevention (CDC). 2008. National Center for Health Statistics.
National Health Interview Survey, 1999-2008.
Chen, L., B. L. Jennison, W. Yang and S. T. Omaye. 2000. Elementary school absenteeism and
air pollution. Inhal Toxicol. Vol. 12 (11): 997-1016.
Chiang, P. L. 1967. Variance and Covariance of Life Table Functions Estimated from a Sample
of Deaths. National Centerfor Health Statistics. Washington, DC. March.
Cohen, A. J., M. Brauer, R. Burnett, H.R. Anderson, J. Frostad, K. Estep, K. Balakrishnan, B.
Brunekreef, L. Dandona, R. Dandona, V. Feigin, G. Freedman, B. Hubbell, A. Jobling, H.
Kan, L. Knibbs, Y. Liu, R. Martin, L. Morawska, C. A. Pope III, H. Shin, K. Straif, G.
Shaddick, M. Thomas, R. van Dingenen, A. van Donkelaar, T. Vos, C. J. L. Murray, M. H.
Forouzanfar. 2017. Estimates and 25-year trends of the global burden of disease
attributable to ambient air pollution: an analysis of data from the Global Burden of
Disease Study 2015. Lancet. Vol. 389:1907-1918.
Collet, D. 1994. Modelling Survival Data in Medical Research. Chapman & Hall: New York.
Cropper, ML and Krupnick, AJ. 1990. Social costs of chronic heart and lung disease. Quality of
the Environment Division, Resources for the Future.
Cropper, Maureen R. and Alan J. Krupnick. 1999. The social costs of chronic heart and lung
disease, in Maureen Cropper, Valuing Environmental Benefits: Selected Essays of
Maureen Cropper, Edward Elgar: Northampton, MA.
Daya, M., R. Schmicker, S. May, and L. Morrison. 2015. Current burden of cardiac arrest in the
United States: Report from the Resuscitation Outcomes Consortium. Paper
commissioned by the Committee on the Treatment of Cardiac Arrest: Current Status and
Future Directions.
BenMAP User Manual v 0.01
R-5
December 2022
-------
References
De Pablo-Fernandez, E, Tur, C, Revesz, T, Lees, AJ, Holton, JL and Warner, TT. 2017.
Association of Autonomic Dysfunction With Disease Progression and Survival in
Parkinson Disease. JAMA Neurol 74(8): 970-976.
Decisioneering. 1996. Crystal Ball: Forecasting and Risk Analysis for Spreadsheet Users: User
Manual. Version 4.0. Denver, CO. www.decisioneering .com.
DerSimonian, R. and N. Laird. 1986. Meta-Analysis in Clinical Trials. Controlled Clinical Trials.
Vol. 7:177-188.
Dewanji, A. and S. H. Moolgavkar. 2000. A Poisson process approach for recurrent event data
with environmental covariates. Environmetrics. Vol. 11: 665-673.
Dewanji, A. and S. H. Moolgavkar. 2002. Choices of stratification in Poisson process analysis of
recurrent event data with environmental covariates. Stat. Med. Vol. 21:3383-3393.
Di, Q., Wang, Y., Zanobetti, A., Wang, Y., Koutrakis, P., Choirat, C., Dominici, F. and Schwartz,
J.D. 2017. Air pollution and mortality in the Medicare population. New Engl J Med
376(26): 2513-2522.
Dickie, M. and S. Gerking. 1987. Reconciling Averting Behavior and Contingent Valuation
Benefit Estimates of Reducing Symptoms of Ozone Exposure (draft), as cited in
Neumann, J. E., M. Dickie, and R.E. Unsworth. 1994. Prepared by Industrial Economics.
Prepared for Jim DeMocker, U.S. EPA, Office of Air and Radiation. March 31.
Dickie, M and Messman, VL. 2004. Parental altruism and the value of avoiding acute illness:
are kids worth more than parents? J Environ Econ Manage 48(3): 1146-1174.
Dickie, M. and V. L. Ulery. 2002. Parental Altruism and the Value of Avoiding Acute Illness: Are
Kids Worth More Than Parents? (Paper to be submitted for publication. Presented at
Association of Environmental and Resource Economists 2001 Workshop, "Assessing and
Managing Environmental and Public Health Risks."). December.
Dockery, D. W., J. Cunningham, A. I. Damokosh, L. M. Neas, J. D. Spengler, P. Koutrakis, J. H.
Ware, M. Raizenne and F. E. Speizer. 1996. Health Effects of Acid Aerosols On North
BenMAP User Manual v 0.01
R-6
December 2022
-------
References
American Children - Respiratory Symptoms. Environmental Health Perspectives. Vol.
104 (5): 500-505.
Dockery, D. W., F. E. Speizer, D. 0. Stram, J. H. Ware, J. D. Spengler and B. G. Ferris, Jr. 1989.
Effects of Inhalable Particles on Respiratory Health of Children. Am Rev Respir Dis. Vol.
139: 587-594.
Eisenstein, E. L., L. K. Shaw, K. J. Anstrom, C. L. Nelson, Z. Hakim, V. Hasselblad and D. B.
Mark. 2001. Assessing the clinical and economic burden of coronary artery disease:
1986-1998. Med Care. Vol. 39 (8): 824-35.
Ensor, K.B., Raun, L.H., & Persse, D. 2013. A Case-Crossover Analysis of Out-of-Hospital
Cardiac Arrest and Air Pollution. Circulation 127:1192-1199.
ESEERCO. 1994. New York State Environmental Externalities Cost Study. Report 2:
Methodology. Empire State Electric Energy Research Corporation.
Fung, K. Y. K., Daniel; Chen, Yue; Burnett, Rick; Cakmak, Sabit. 2003. Comparison of time
series and case-crossover analyses of air pollution and hospital admission data.
International Journal of Epidemiology. Vol. 32:1064-1070.
Garcia, E., Berhane, K.T., Islam, T., McConnell, R., Urman, R., Chen, Z., & Gilliland, F.D. 2019.
Association of Changes in Air Quality With Incident Asthma in Children in California,
1993-2014. JAMA-J Am Med Assoc, 321(19): 1906-1915.
GeoLytics Inc. 2001. CensusCD® 1990 + Maps, Release 4.1. CD-ROM. GeoLytics, Inc.
Gharibvand, L., Shavlik, D., Ghamsary, M., Beeson, W.L., Soret, S., Knutsen, R., & Knutsen, S.F.
2016. The association between ambient fine particulate air pollution and lung cancer
incidence: results from the AHSMOG-2 study. Environ Health Perspect 125 (3): 378-384.
Gilliland, F. D., K. Berhane, E. B. Rappaport, D. C. Thomas, E. Avol, W. J. Gauderman, S. J.
London, H. G. Margolis, R. McConnell, K. T. Islam and J. M. Peters. 2001. The effects of
ambient air pollution on school absenteeism due to respiratory illnesses. Epidemiology.
Vol. 12 (1): 43-54.
BenMAP User Manual v 0.01
R-7
December 2022
-------
References
Glad, J.A., L.L. Brink, E.O. Talbott, P.C. Lee, X. Xu, M. Saul, and J. Rager. 2012. The Relationship
of Ambient Ozone and PM2.5 Levels and Asthma Emergency Department Visits: Possible
Influence of Gender and Ethnicity. Archives of Environmental & Occupational Health. Vol
62 (2): 103-108.
Goodwin, N, Smith, J, Davies, A, Perry, C, Rosen, R, Dixon, A, Dixon, J and Ham, C. 2011. A
report to the Department of Health and the NHS Future Forum; Integrated Care for
Patients and Populations: Improving Outcomes by Working Together.
Greene, W. H. 1997. Econometric Analysis. Prentice Hall: Upper Saddle River, NJ.
Guerra, S, Wright, AL, Morgan, WJ, Sherrill, DL, Holberg, CJ and Martinez, FD. 2004.
Persistence of asthma symptoms during adolescence: role of obesity and age at the
onset of puberty. Am J Respir Crit Care Med 170(1): 78-85.
Haase, N. 2002. Phone conversation. American Heart Association, October.
Huang, Y., F. Dominici and M. L. Bell. 2005. Bayesian hierarchical distributed lag models for
summer ozone exposure and cardio-respiratory mortality. Environmetrics. Vol. 16:547-
562.
Industrial Economics Incorporated (lEc). 1993. Memorandum to Jim DeMocker, U.S.
Environmental Protection Agency, Office of Air and Radiation, Office of Policy Analysis
and Review. September30.
Industrial Economics Incorporated (lEc). 1994. Linkage Between Health Effects Estimation
and Morbidity Valuation in the Section 812 Analysis - Draft Valuation Document.
Memorandum to Jim DeMocker, U.S. Environmental Protection Agency, Office of Air and
Radiation, Office of Policy Analysis and Review. Prepared by J.E. Neumann, M.T. Dickie,
and R.E. Unsworth. March 31.
Industrial Economics Incorporated (IEc).2006. Expanded Expert Judgment Assessment of the
Concentration-Response Relationship Between PM2.5 Exposure and Mortality, Final
Report. Prepared for U.S. Environmental Protection Agency, Office of Air and Radiation,
Office of Policy Analysis and Review. Prepared by lEc. Cambridge, MA. September 21.
BenMAP User Manual v 0.01
R-8
December 2022
-------
References
Ito, K. 2003. Associations of Particulate Matter Components with Daily Mortality and Morbidity
in Detroit, Michigan. In: Revised Analyses of Time-Series Studies of Air Pollution and
Health. Health Effects Institute. Boston, MA. May.
Ito, K., S. F. De Leon and M. Lippmann. 2005. Associations between ozone and daily mortality:
analysis and meta-analysis. Epidemiology. Vol. 16 (4): 446-57.
Ito, K., Ross, Z., Zhou, J., Nadas, A., Lippmann, M. and Thurston, G.D. 2013. National Particle
Component Toxicity (NPACT) initiative: Study 3. Time-series analysis of mortality,
hospitalizations, and ambient PM2.5 and its components. Boston, MA, Health Effects
Institute: 95-125.
Ito, K. and G. D. Thurston. 1996. Daily PMi0/mortality associations: an investigations of at-risk
subpopulations. Journal of Exposure Analysis and Environmental Epidemiology. Vol. 6
(1): 79-95.
Ito, K., G. D. Thurston and R. A. Silverman. 2007. Characterization of PM2.s, gaseous pollutants,
and meteorological interactions in the context of time-series health effects models. J
Expo Sci Environ Epidemiol. Vol. 17 Suppl 2: S45-60. Jerrett, M, Turner, MC, Beckerman,
BS, Pope, CA, van Donkelaar, A, Martin, RV, Serre, M, Crouse, D, Gapstur, SM, Krewski, D,
Diver, WR, Coogan, PF, Thurston, GD and Burnett, RT. 2017. Comparing the health
effects of ambient particulate matter estimated using ground-based versus remote
sensing exposure estimates. Environ Health Perspect 125(4): 552-559.
Jerrett, M., R. Burnett, A. I. Pope, K. Ito, G. Thurston, D. Krewski, Y. Shi, E. Calle and M. Thun.
2009. Long-Term Ozone Exposure and Mortality. The New England Journal of Medicine.
Vol.360 (11): 1085-95.
Jones, R.R., Hogrefe, C., Fitzgerald, E.F., Hwang, S.A., Ozkaynak, H., Garcia, V.C., & Lin, S. 2015.
Respiratory hospitalizations in association with fine PM and its components in New York
State. J Air Waste Manag Assoc, 65(5): 559-569.
Judge, G. G., W. E. Griffiths, R. C. Hill, H. Lutkepohl and T.-C. Lee. 1985. The Theory and
Practice of Econometrics. John Wiley and Sons: New York.Katsouyanni, K and J.M.
BenMAP User Manual v 0.01
R-9
December 2022
-------
References
Samet. 2009. Air Pollution and Health: A European and North American Approach
(APHENA). Health Effects Institute, Research Report 142.
Jutkowitz, E, Kane, RL, Gaugler, JE, MacLehose, RF, Dowd, B and Kuntz, KM. 2017. Societal
and Family Lifetime Cost of Dementia: Implications for Policy. J Am Geriatr Soc 65(10):
2169-2175.
Katsouyanni, K., Samet, J. M., Anderson, H. R., Atkinson, R., Tertre, A. L., Medina, S., et al.
2009. Air Pollution and Health: A European and North American Approach (APHENA):
Health Effects Institute.
Kaye, DR, Min, HS, Herrel, LA, Dupree, JM, Ellimoottil, C and Miller, DC. 2018. Costs of Cancer
Care Across the Disease Continuum. Oncologist 23(7): 798-805.
Kennedy, P. 1990. A Guide to Econometrics. MIT Press: Cambridge, MA.
King, K., Starsinic, M., Hernandez Viver, A., Beaghen, M. 2015, "The Reliability of ACS 5-Year
Estimates of Race Groups and American Indian and Alaska Native Populations", U.S.
Census Bureau.
Kioumourtzoglou, M.A., Schwartz, J.D., Weisskopf, M.G., Melly, S.J., Wang, Y., Dominici, F., &
Zanobetti, A. 2016. Long-term PM2.5 exposure and neurological hospital admissions in
the northeastern United States. Environ Health Perspect, 124(1): 23-29.
Kloog, I., B.A. Coull, A. Zanobetti, P. Koutrakis, J.D. Schwartz. 2012. Acute and Chronic Effects
of Particles on Hospital Admissions in New-England. PLoS ONE. Vol 7 (4): 1-8.
Krall, J. R., Mulholland, J. A., Russell, A. G., Balachandran, S., Winquist, A., Tolbert, P. E.,
Waller, L.A., & Sarnat, S. E. 2016. Associations between Source-Specific Fine Particulate
Matter and Emergency Department Visits for Respiratory Disease in Four U.S. Cities.
Environ Health Perspect, 125(1), 97-103.
Krewski, D., M. Jerrett, R. Burnett, R. Ma, E. Hughes, Y. Shi, M. C. Turner, C. A. I. Pope, G.
Thurston, E. Calle and M. J. Thun. 2009. Extended follow-up and spatial analysis of the
BenMAP User Manual v 0.01
R-10
December 2022
-------
References
American Cancer Society study linking particulate air pollution and mortality. HEI
Research Report, 140, Health Effects Institute, Boston, MA.
Krupnick, A. J. and M. L. Cropper. 1992. The Effect of Information On Health Risk Valuations.
Journal of Risk and Uncertainty. Vol. 5 (1): 29-48.
Laden, F., J. Schwartz, F. E. Speizer and D. W. Dockery. 2006. Reduction in Fine Particulate Air
Pollution and Mortality: Extended follow-up of the Harvard Six Cities Study. Am J Respir
Crit Care Med. Vol. 173 (6): 667-72.
Lee, WC, Christensen, MC, Joshi, AV and Pashos, CL. 2007. Long-term cost of stroke subtypes
among Medicare beneficiaries. Cerebrovascular Diseases 23(1): 57-65.
Levy, J. I., S. M. Chemerynski and J. A. Sarnat. 2005. Ozone exposure and mortality: an empiric
bayes metaregression analysis. Epidemiology. Vol. 16 (4): 458-68.
Lewis, T.C., Robins, T.G., Mentz, G.B., Zhang, X., Mukherjee, B., Lin, X., Keeler, G.J., Dvonch,
J.T., Yip, F.Y., O'Neill, M.S., Parker, E.A., Israel, B.A., Max, P.T., Reyes, A., & Committee,
CAAACS. 2013. Air pollution and respiratory symptoms among children with asthma:
Vulnerability by corticosteroid use and residence area. Sci Total Environ 448:48-55.
Lin, M., Stieb, D.M., Chen, Y. 2005. Coarse Particulate Matter and Hospitalization for
Respiratory Infections in Children Youngerthan 15 Years in Toronto: A Case-Crossover
Analysis. American Academy of Pediatrics, 116,235-240.
Lippmann, M., K. Ito, A. Nadas and R. Burnett. 2000. Association of Particulate Matter
Components with Daily Mortality and Morbidity in Urban Populations. Health Effects
Institute. Number 95. August.
Loehman, E. T., S. V. Berg, A. A. Arroyo, R. A. Hedinger, J. M. Schwartz, M. E. Shaw, R. W.
Fahien, V. H. De, R. P. Fishe, D. E. Rio, W. F. Rossley and A. E. S. Green. 1979.
Distributional Analysis of Regional Benefits and Cost of Air Quality Control. Journal of
Environmental Economics and Management. Vol. 6: 222-243.
BenMAP User Manual v 0.01
R-ll
December 2022
-------
References
Luengo-Fernandez, R, Gray, AM and Rothwell, PM. 2012. A population-based study of hospital
care costs during 5 years after transient ischemic attack and stroke. Stroke 43(12): 3343-
3351.
Mar, T.F., J.Q. Koenig. 2009. Relationship between visits to emergency departments for
asthma and ozone exposure in greater Seattle, Washington. Annals of Allergy, Asthma &
Immunology, Vol. 103(6):474-479.
Mar, T. F., J. Q. Koenig, et al. 2010. Associations between asthma emergency visits and
particulate matter sources, including diesel emissions from stationary generators in
Tacoma, Washington. Inhal Toxicol 22 (6): 445-8.
Mar, T. F., T. V. Larson, et al. 2004. An analysis of the association between respiratory
symptoms in subjects with asthma and daily air pollution in Spokane, Washington."
Inhal Toxicol 16 (13): 809-15.
McConnell, R., K. Berhane, F. Gilliland, S. J. London, H. Vora, E. Avol, W. J. Gauderman, H. G.
Margolis, F. Lurmann, D. C. Thomas and J. M. Peters. 1999. Air pollution and bronchitic
symptoms in Southern California children with asthma. Environ Health Perspect. Vol.
107 (9): 757-60.
McConnell, R., Islam, T., Shankardass, K., Jerrett, M., Lurmann, F., Gilliland, F., Gauderman, J.,
Avol, E., Kunzli, N., Yao, L., Peters, J., & Berhane, K. 2010. Childhood Incident Asthma and
Traffic-Related Air Pollution at Home and School. Environ Health Perspect, 118(7), 1021-
1026.
Medina-Ramon, M. and Schwartz, J. 2008. Who is more vulnerable to die from ozone air
pollution? Epidemiology, 19(5), 672-679.
Moolgavkar, S. H. 2000a. Air Pollution and Hospital Admissions for Chronic Obstructive
Pulmonary Disease in Three Metropolitan Areas in the United States. Inhalation
Toxicology. Vol. 12 (Supplement 4): 75-90.
Moolgavkar, S. H. 2000b. Air pollution and hospital admissions for diseases of the circulatory
system in three U.S. metropolitan areas. J Air Waste Manag Assoc. Vol. 50 (7): 1199-206.
BenMAP User Manual v 0.01
R-12
December 2022
-------
References
Moolgavkar, S. H. 2003. Air Pollution and Daily Deaths and Hospital Admissions in Los Angeles
and Cook Counties. In: Revised Analyses of Time-Series Studies of Air Pollution and
Health. Health Effects Institute. Boston, MA. May.
Moolgavkar, S. H., E. G. Luebeck and E. L. Anderson. 1997. Air pollution and hospital
admissions for respiratory causes in Minneapolis St. Paul and Birmingham.
Epidemiology. Vol. 8 (4): 364-370.
Moolgavkar, S. H., E. G. Luebeck, T. A. Hall and E. L. Anderson. 1995. Air Pollution and Daily
Mortality in Philadelphia. Epidemiology. Vol. 6 (5): 476-484.
Mortimer, K. M., L. M. Neas, D. W. Dockery, S. Redline and I. B. Tager. 2002. The effect of air
pollution on inner-city children with asthma. Eur Respir J. Vol. 19 (4): 699-705.
Mrozek, J. R. and L. O.Taylor. 2002. What Determines the Value of Life? A Meta-Analysis.
Journal of Policy Analysis and Management. Vol. 21: 253-270.
Mu, F, Hurley, D, Betts, K, Messali, A, Paschoalin, M, Kelley, C and Wu, E. 2017. Real-world costs
of ischemic stroke by discharge status. Current medical research and opinion 33(2): 371-
378.
National Center for Health Statistics. 1999. National Vital Statistics Reports. U.S. Department
of Health and Human Services, Centers for Disease Control and Prevention, National
Centerfor Health Statistics. Washington, DC. Volume 47, Number 19. June 30.
NCI. 2015. SEER, Surveillance, Epidemiology, and End Results Program. National Institutes of
Health, National Cancer Institute (NCI). Bethesda, MD.
Nelder, J. A. and R. Mead. 1965. A simplex algorithm for function minimization. Computer
Journal. Vol. 7:308-313.
Nicholson, G, Gandra, SR, Halbert, RJ, Richhariya, A and Nordyke, RJ. 2016. Patient-level costs
of major cardiovascular conditions: a review of the international literature.
ClinicoEconomics and outcomes research: CEOR 8:495.
BenMAP User Manual v 0.01
R-13
December 2022
-------
References
Nishimura, K.K., Galanter, J.M., Roth, L.A., Oh, S.S., Thakur, N., Nguyen, E.A., Thyne, S., Farber,
H.J., Serebrisky, D., Kumar, R., Brigino-Buenaventura, E., Davis, A., LeNoir, M.A., Meade,
K., Rodriguez-Cintron, W., Avila, P.C., Borrell, L.N., Bibbins-Domingo, K., Rodriguez-
Santana, J.R., Sen, S., Lurmann, F., Balmes, J.R., & Burchard, E.G. 2013. Early-Life Air
Pollution and Asthma Risk in Minority Children: The GALA II and SAGE II Studies. Am J
Respir Crit Care Med Vol, 188(3), 309-318.
Norris, G., S. N. YoungPong, J. Q. Koenig, T. V. Larson, L. Sheppard and J. W. Stout. 1999. An
association between fine particles and asthma emergency department visits for
children in Seattle. Environ Health Perspect. Vol. 107 (6): 489-93.
Ochs, M, Nyengaard, JR, Jung, A, Knudsen, L, Voigt, M, Wahlers, T, Richter, J and Gundersen,
HJ. 2004. The number of alveoli in the human lung. Am J Respir Crit Care Med 169(1):
120-124.
O'Connor, R. M. and G. C. Blomquist. 1997. Measurement of Consumer-Patient Preferences
Using a Hybrid Contingent Valuation Method. Journal of Health Economics. Vol. 16:667-
683.
O'Connor, G.T., L. Neas, B. Vaughn, M. Kattan, H. Mitchell, E. F. Crain, R. Evans III, R. Gruchalla,
W. Morgan, J. Stout, G. K. Adam, and M. Lippmann. 2008. Acute respiratory health effects
of air pollution on children with asthma in US inner cities. Journal of Allergy and Clinical
Immunology, Vol. 121(5): 1133-1139.
O'Sullivan, AK, Rubin, J, Nyambose, J, Kuznik, A, Cohen, DJ and Thompson, D. 2011. Cost
estimation of cardiovascular disease events in the US. Pharmacoeconomics 29(8): 693-
704.
Ostro, B., M. Lipsett, J. Mann, H. Braxton-Owens and M. White. 2001. Air pollution and
exacerbation of asthma in African-American children in Los Angeles. Epidemiology. Vol.
12 (2): 200-8.
Ostro, B. D. 1987. Air Pollution and Morbidity Revisited: A Specification Test. Journal of
Environmental Economics and Management. Vol. 14: 87-98.
BenMAP User Manual v 0.01
R-14
December 2022
-------
References
Ostro, B. D., M. J. Lipsett, M. B. Wiener and J. C. Seiner. 1991. Asthmatic Responses to Airborne
Acid Aerosols. Am J Public Health. Vol. 81 (6): 694-702.
Ostro, B., Malig, B., Hasheminassab, S., Berger, K., Chang, E. & Sioutas, C. 2016. Associations
of Source-Specific Fine Particulate Matter With Emergency Department Visits in
California. Am J Epidemiol 184(6): 450-459.
Ostro, B., Roth, L., Malig, M., & Marty, M. 2009. The Effects of Fine Particle Components on
Respiratory Hospital Admissions in Children. Environ Health Perspect, 117(3): 475-480.
Ostro, B. D. and S. Rothschild. 1989. Air Pollution and Acute Respiratory Morbidity - an
Observational Study of Multiple Pollutants. Environ Res. Vol. 50 (2): 238-247. Owings, M.
F. and L. Lawrence. 1999. Detailed Diagnoses and Procedures, National Hospital
Discharge Survey, 1997. National Center for Health Statistics. Hyattsville, MD. Vital
Health Statistics, Series 13, No. 145. December.
Parker, J.D., Akinbami, L. J., & Woodruff, T.J. 2009. Air Pollution and Childhood Respiratory
Allergies in the United States. Environ Health Perspect 117(1): 140-147.
Paulu, C. and Smith, A.E. 2008. Tracking Associations Between Ambient Ozone and Asthma-
Related Emergency Department Visits Using Case-Crossover Analysis. Journal of Public
Health Management and Practice, 14(6), 581-591.
Peel, J. L., P. E. Tolbert, M. Klein, K. B. Metzger, W. D. Flanders, K. Todd, J. A. Mulholland, P. B.
Ryan and H. Frumkin. 2005. Ambient air pollution and respiratory emergency
department visits. Epidemiology. Vol. 16 (2): 164-74.
Peng, R. D., M. L. Bell, et al. 2009. Emergency admissions for cardiovascular and respiratory
diseases and the chemical composition of fine particle air pollution. Environ Health
Perspect, Vol. 117 (6): 957-63.
Peng, R. D., H. H. Chang, et al. 2008. "Coarse particulate matter air pollution and hospital
admissions for cardiovascular and respiratory diseases among Medicare patients."
Jama, Vol. 299 (18): 2172-9.
BenMAP User Manual v 0.01
R-15
December 2022
-------
References
Peters, A., D. W. Dockery, J. E. Muller and M. A. Mittleman. 2001. Increased particulate air
pollution and the triggering of myocardial infarction. Circulation. Vol. 103 (23): 2810-5.
http:// www.circulationaha.org/cgi/content/fuU/103/23/2810.
Pope, C. A., 3rd, J. B. Muhlestein, et al. 2006. "Ischemic heart disease events triggered by
short-term exposure to fine particulate air pollution." Circulation, Vol. 114 (23): 2443-8.
Pope, C. A., 3rd, R. T. Burnett, M. J. Thun, E. E. Calle, D. Krewski, K. Ito and G. D. Thurston.
2002. Lung cancer, cardiopulmonary mortality, and long-term exposure to fine
particulate air pollution. Jama. Vol. 287 (9): 1132-41.
Pope, C. A., D. W. Dockery, J. D. Spengler and M. E. Raizenne. 1991. Respiratory Health and
PMio Pollution - a Daily Time Series Analysis. American Review of Respiratory Disease.
Vol. 144 (3): 668-674.
Pope, C.A., Turner, M.C., Burnett, R., Jerrett, M., Gapstur, S.M., Diver, W.R., Krewski, D. and
Brook, R.D. 2015. Relationships between fine particulate air pollution, cardiometabolic
disorders and cardiovascular mortality. Circul Res 116(1): 108-U258.
Pope, C.A., J.S. Lefler, M. Ezzati, J.D. Higbee, J.D. Marshall, S. Kim, M. Bechele, K.S. Gilliat, S.E.
Vernon, A.L. Robinson, and R.T. Burnett. 2019. Mortality Risk and Fine Particulate Air
Pollution in a Large, Representative Cohort of U.S. Adults. Environmental Health
Perspectives 127(7): 077077.
Popovic, J. R. 2001.1999 National Hospital Discharge Survey: annual summary with detailed
diagnosis and procedure data. Vital Health Stat 13. (151): i-v, 1-206.
Rabinovitch, N., Strand, M., & Gelfand, E.W. 2006. Particulate Levels Are Associated with Early
Asthma Worsening in Children with Persistent Disease. Am J Respir Crit Care Med, 173:
1098-1105.
Ransom, M. R. and C. A. Pope. 1992. Elementary school absences and PMio pollution in Utah
Valley. Environ Res. Vol. 58 (2): 204-19.
BenMAP User Manual v 0.01
R-16
December 2022
-------
References
RogerV.L., A.S. Go, D.M. Lloyd-Jones, E.J. Benjamin, J.D. Berry, W.B. Borden, D.M. Bravata, S.
Dai, E.S. Ford, C.S. Fox, H.J. Fullerton, C. Gillespie, S.M. Hailpern, J.A. Heit, V.J. Howard,
B.M. Kissel, S.J. Kittner, D.T. Lackland, J.H. Lichtman, L.D. Lisabeth, D.M. Makuc, G.M.
Marcus, A. Marelli, D.B. Matchar, C.S. Moy, D. Mozaffarian, M.E. Mussolino, G. Nichol, N.P.
Paynter, E.Z. Soliman, P.D. Sorlie, N. Sotoodehnia, T.N. Turan, S.S. Virani, N.D. Wong, D.
Woo and M.B. Turner; on behalf of the American Heart Association Statistics Committee
and Stroke Statistics Subcommittee. Heart disease and stroke statistics—2012 update: a
report from the American Heart Association. Circulation. 2012;125:e2- e220.
Rosamond, W., G. Broda, E. Kawalec, S. Rywik, A. Pajak, L. Cooper and L. Chambless. 1999.
Comparison of medical care and survival of hospitalized patients with acute myocardial
infarction in Poland and the United States. Am J Cardiol. Vol. 83 (8): 1180-5.
Rosenthal, F.S., Carney, J.P., & Olinger, M.L. 2008. Out-of-Hospital Cardiac Arrest and Airborne
Fine Particulate Matter: A Case-Crossover Analysis of Emergency Medical Services Data
in Indianapolis, Indiana. Environ Health Perspect 116(5): 631-636.
Rowe, R. D. and L. G. Chestnust. 1986. Oxidants and Asthmatics in Los Angeles: A Benefits
Analysis - Executive Summary. Prepared for U.S. Environmental Protection Agency,
Office of Policy Analysis. Prepared by Energy and Resource Consultants, Inc.
Washington, DC. EPA-230-09-86-018. March.
Russell, M. W., D. M. Huse, S. Drowns, E. C. Hamel and S. C. Hartz. 1998. Direct medical costs of
coronary artery disease in the United States. Am J Cardiol. Vol. 81 (9): 1110-5.
Sarnat, J. A., S. E. Sarnat, W. D. Flanders, H. H. Chang, J. Mulholland, L. Baxter, V. Isakob, and
H. Ozkaynak. 2013. Spatiotemporally resolved air exchange rate as a modifier of acute
air pollution-related morbidity in Atlanta. Journal of Exposure Science and
Environmental Epidemiology 23,606-615.
Samet, J. M., S. L. Zeger, J. E. Kelsall, J. Xu and L. S. Kalkstein. 1997. Air Pollution, Weather,
and Mortality in Philadelphia 1973-1988. Health Effects Institute. Cambridge, MA. March.
BenMAP User Manual v 0.01
R-17
December 2022
-------
References
Schildcrout, J. S., L. Sheppard, T. Lumley, J. C. Slaughter, J. Q. Koenig and G. G. Shapiro. 2006.
Ambient air pollution and asthma exacerbations in children: an eight-city analysis. Am J
Epidemiol. Vol. 164 (6): 505-17.
Schwartz, J. 1994a. Air Pollution and Hospital Admissions For the Elderly in Detroit, Michigan.
American Journal of Respiratory and Critical Care Medicine. Vol. 150 (3): 648-655.
Schwartz, J. 1994b. PM(10) Ozone, and Hospital Admissions Forthe Elderly in Minneapolis St
Paul, Minnesota. Archives of Environmental Health. Vol. 49 (5): 366-374.
Schwartz, J. 1995. Short term fluctuations in air pollution and hospital admissions of the
elderly for respiratory disease. Thorax. Vol. 50 (5): 531-538.
Schwartz, J. 2005. How sensitive is the association between ozone and daily deaths to control
for temperature? Am J Respir Crit Care Med. Vol. 171 (6): 627-31.
Schwartz, J., C. Spix, G. Touloumi, L. Bacharova, T. Barumamdzadeh, A. le Tertre, T. Piekarksi,
A. Ponce de Leon, A. Ponka, G. Rossi, M. Saez and J.P. Schouten. 1996. Methodological
issues in studies of air pollution and daily counts of deaths or hospital admissions. J.
Epidemiol. Community Health. Vol. 50 Suppl. 1: S3-11.
Schwartz, J., D. W. Dockery, L. M. Neas, D. Wypij, J. H. Ware, J. D. Spengler, P. Koutrakis, F. E.
Speizerand B. G. Ferris. 1994. Acute Effects of Summer Air Pollution On Respiratory
Symptom Reporting in Children. Am J Respir Crit Care Med. Vol. 150 (5): 1234-1242.
Schwartz, J. and L. M. Neas. 2000. Fine particles are more strongly associated than coarse
particles with acute respiratory health effects in schoolchildren. Epidemiology. Vol. 11
(1): 6-10.
Sheppard, L. 2003. Ambient Air Pollution and Nonelderly Asthma Hospital Admissions in
Seattle, Washington, 1987-1994. In: Revised Analyses of Time-Series Studies of Air
Pollution and Health. Health Effects Institute. Boston, MA. May.
BenMAP User Manual v 0.01
R-18
December 2022
-------
References
Sheppard, L., D. Levy, G. Norris, T. V. Larson and J. Q. Koenig. 1999. Effects of ambient air
pollution on nonelderly asthma hospital admissions in Seattle, Washington, 1987-1994.
Epidemiology. Vol. 10 (1): 23-30.
Siegel, RL, Miller, KD and Jemal, A. 2019. Cancer statistics, 2019. CA Cancer J Clin 69(1): 7-34.
Slaughter, J. C., E. Kim, et al. 2005. "Association between particulate matter and emergency
room visits, hospital admissions and mortality in Spokane, Washington." J Expo Anal
Environ Epidemiol 15(2): 153-9.
Samet, J. M., S. L. Zeger, J. E. Kelsall, J. Xu and L. S. Kalkstein. 1997. Air Pollution, Weather,
and Mortality in Philadelphia 1973-1988. Health Effects Institute. Cambridge, MA. March.
Schildcrout, J. S., L. Sheppard, T. Lumley, J. C. Slaughter, J. Q. Koenig and G. G. Shapiro. 2006.
Ambient air pollution and asthma exacerbations in children: an eight-city analysis. Am J
Epidemiol. Vol. 164 (6): 505-17.
Schwartz, J. 1994a. Air Pollution and Hospital Admissions Forthe Elderly in Detroit, Michigan.
American Journal of Respiratory and Critical Care Medicine. Vol. 150 (3): 648-655.
Schwartz, J. 1994b. PM(10) Ozone, and Hospital Admissions Forthe Elderly in Minneapolis St
Paul, Minnesota. Archives of Environmental Health. Vol. 49 (5): 366-374.
Schwartz, J. 1995. Short term fluctuations in air pollution and hospital admissions of the
elderly for respiratory disease. Thorax. Vol. 50 (5): 531-538.
Schwartz, J. 2005. How sensitive is the association between ozone and daily deaths to control
for temperature? Am J Respir Crit Care Med. Vol. 171 (6): 627-31.
Schwartz, J., C. Spix, G. Touloumi, L. Bacharova, T. Barumamdzadeh, A. le Tertre, T. Piekarksi,
A. Ponce de Leon, A. Ponka, G. Rossi, M. Saez and J.P. Schouten. 1996. Methodological
issues in studies of air pollution and daily counts of deaths or hospital admissions. J.
Epidemiol. Community Health. Vol. 50 Suppl. 1: S3-11.
BenMAP User Manual v 0.01
R-19
December 2022
-------
References
Schwartz, J., D. W. Dockery, L. M. Neas, D. Wypij, J. H. Ware, J. D. Spengler, P. Koutrakis, F. E.
Speizerand B. G. Ferris. 1994. Acute Effects of Summer Air Pollution On Respiratory
Symptom Reporting in Children. Am J Respir Crit Care Med. Vol. 150 (5): 1234-1242.
Schwartz, J. and L. M. Neas. 2000. Fine particles are more strongly associated than coarse
particles with acute respiratory health effects in schoolchildren. Epidemiology. Vol. 11
(1): 6-10.
Sheppard, L. 2003. Ambient Air Pollution and Nonelderly Asthma Hospital Admissions in
Seattle, Washington, 1987-1994. In: Revised Analyses of Time-Series Studies of Air
Pollution and Health. Health Effects Institute. Boston, MA. May.
Sheppard, L., D. Levy, G. Norris, T. V. Larson and J. Q. Koenig. 1999. Effects of ambient air
pollution on nonelderly asthma hospital admissions in Seattle, Washington, 1987-1994.
Epidemiology. Vol. 10 (1): 23-30.
Silverman, R.A., Ito, K., Freese, J., Kaufman, B.J., De Claro, D., Braun, J., & Prezant, D.J. 2010.
Association of Ambient Fine Particles With Out-of-Hospital Cardiac Arrests in New York
City. Am J Epidemiol, 172(8): 917-923.
Slaughter, J. C., E. Kim, et al. 2005. "Association between particulate matter and emergency
room visits, hospital admissions and mortality in Spokane, Washington." J Expo Anal
Environ Epidemiol 15(2): 153-9.
Smith, D. H., D. C. Malone, K. A. Lawson, L. J. Okamoto, C. Battista and W. B. Saunders. 1997. A
national estimate of the economic costs of asthma. Am J Respir Crit Care Med. Vol. 156
(3 Pt 1): 787-93.
Soni, A. 2008. Allergic Rhinitis: Trends in use and expenditures, 2000 and 2005. Medical
Expenditure Panel Survey, Agency for Healthcare Research and Quality.
Sparrow, D, O'Connor, GT, Rosner, B, Segal, MR and Weiss, ST. 1991. The influence of age and
level of pulmonary function on nonspecific airway responsiveness. The Normative Aging
Study. Am Rev Respir Dis 143(5 Pt 1): 978-982.
BenMAP User Manual v 0.01
R-20
December 2022
-------
References
Stanford, R., T. McLaughlin and L. J. Okamoto. 1999. The cost of asthma in the emergency
department and hospital. Am J Respir Crit Care Med. Vol. 160 (1): 211-5.
Sullivan, J., L. Sheppard, et al. 2005. "Relation between short-term fine-particulate matter
exposure and onset of myocardial infarction." Epidemiology 16 (1): 41-8.
Talbott, E.O., Rager, J.R., Benson, S., Brink, L.A., Bilonick, R.A., & Wu, C. 2014. A case-crossover
analysis of the impact of PM2.5 on cardiovascular disease hospitalizations for selected
CDC tracking states. Environ Res 143:455-465.
Tetreault, L.F., Doucet, M., Gamache, P., Fournier, M., Brand, A., Kosatsky, T., & Smargiassi, A.
2016. Childhood exposure to ambient air pollutants and the onset of asthma: an
administrative cohort study in Quebec. Environ Health Perspect 124(8):1276-1282.
Thurston, GD, Ahn, J, Cromar, KR, Shao, Y, Reynolds, HR, Jerrett, M, Lim, CC, Shanley, R, Park,
Y and Hayes, RB. 2016. Ambient particulate matter air pollution exposure and mortality
in the NIH-AARP Diet and Health Cohort. Environ Health Perspect 124(4): 484-490.
Tolley, G. S. and et al. 1986. Valuation of Reductions in Human Health Symptoms and Risks.
Prepared for U.S. Environmental Protection Agency. January.
Trivedi, M. and Denton, E. 2019. Asthma in Children and Adults-What Are the Differences and
What Can They Tell us About Asthma? Front Pediatr7:256.
Turner, M.C., Jerrett, M., Pope, A., Ill, Krewski, D., Gapstur, S.M., Diver, W.R., Beckerman, B.S.,
Marshall, J.D., Su, J., Crouse, D.L. and Burnett, R.T. 2016. Long-term ozone exposure and
mortality in a large prospective study. Am J Respir Crit Care Med 193(10): 1134-1142.
U.S. Bureau of the Census. 1997. Statistical Abstract of the United States: 1997. Washington,
DC.
U.S. Bureau of the Census. 2002a. Modified Race Data Summary File 2000 Technical
Documentation. Census of Population and Housing: Washington DC.
http://www.census.gov/ popest/archives/files/MRSF-01-USl.pdf.
BenMAP User Manual v 0.01
R-21
December 2022
-------
References
U.S. Bureau of the Census. 2002b. Statistical Abstract of the United States: 2001. Washington
DC.
US Bureau of Labor Statistics. 2015. Current Population Survey. 2015 Annual Social and
Economic Supplement. Available at: https://www.bls.gov/opub/reports/womens-
databook/2016/home.htm.
U.S. Department of Education. 1996. The Condition of Education 1996, Indicator 42: Student
Absenteeism and Tardiness. National Center for Education Statistics. Washington DC.
U.S. EPA. 1997. The Benefits and Costs of the Clean Air Act: 1970 to 1990. U.S. EPA, Office of Air
and Radiation, Office of Policy, Planning and Evaluation. Washington, DC. EPA 410- R-97-
002. October, http://www.epa.gov/airprogm/oar/sect812/index.html.
U.S. EPA. 1999a. The Benefits and Costs of the Clean Air Act: 1990 to 2010: EPA Report to
Congress. U.S. EPA, Office of Air and Radiation, Office of Policy. Washington, DC. EPA
410-R-99-001. November, http://www.epa.gov/airprogm/oar/sect812/index.html.
U.S. EPA. 1999b. OAQPS Economic Analysis Resource Document. Prepared by U.S. EPA, Office
of Air Quality Planning and Standards, Innovative Strategies and Economics Group.
Durham, NC. April, http://www.epa.gov/ttn/ecas/analguid.html.
U.S. EPA. 2000. Guidelines for Preparing Economic Analyses. Office of the Administrator.
Washington, DC. EPA#: 240-R-00-003. September.
http://yosemite.epa.gov/ee/epa/eed.nsf/
webpages/Guidelines.html/Sfile/Guidelines.pdf.
U.S. EPA. 2004. Final Regulatory Analysis: Control of Emissions from Nonroad Diesel Engine.
U.S. EPA, Office of Transportation and Air Quality. Washington, DC. EPA420-R-04-007.
April.
U.S. EPA. 2006. Final Regulatory Impact Analysis: PM2.5 NAAQS. Office of Air and Radiation,
Office of Air Quality Planning and Standards. Research Triangle Park, NC. http://
www.epa.gov/ttn/ecas/ria.html.
BenMAP User Manual v 0.01
R-22
December 2022
-------
References
U.S. EPA. 2008a. Final Ozone NAAQS Regulatory Impact Analysis. Prepared by Office of Air and
Radiation, Office of Air Quality Planning and Standards. Research Triangle Park, NC.
http://www.epa.gov/ttn/ecas/ria.html.
U.S. EPA. 2008b. Regulatory Impact Analysis: Control of Emissions of Air Pollution from
Locomotive Engines and Marine Compression Ignition Engines Less than 30 Liters Per
Cylinder. Office of Transportation and Air Quality, Assesssment and Standards Division.
Washington, DC. EPA420-R-08-001. March, http://www.epa.gov/otaq/locomotv.htm.
U.S. EPA. 2012. Regulatory Impact Analysis for the Final Revisions to the National Ambient Air
Quality Standards for Particulate Matter. EPA-452/R-12-003. Office of Air Quality
Planning and Standards, Health and Environmental Impacts Division. December.
Available at http://www.epa.gov/pm/2012/finalria.pdf.
U.S. EPA Science Advisory Board (U.S. EPA-SAB). 2004. Advisory on Plans for Health Effects
Analysis in the Analytical Plan for EPA's Second Prospective Analysis—Benefits and
Costs of the Clean Air Act, 1990-2020. EPA-SAB-COUNCIL-ADV-04-002. March. Available
at
http://vosemite.epa.gov/sab%5CSABPRQDUCT.NSF/08E1155AD24F871C85256E5400433
D5D/$File/council adv 04002.pdf.
U.S. EPA. 2017. Trichloroethylene (TCE); Regulation of Use in Vapor Degreasing Under TSCA
Section 6(a). 40 CFR 751. Washington, DC. Available at:
https://www.federalregister.gov/documents/2017/01/19/2017-Q1229/trichloroethylene-
tce-regulation-of-use-in-vapor-degreasing-under-tsca-section-6a.
Villeneuve, P.J., Chen, L., Rowe, B.H., Coates, F. 2007. Outdoor air pollution and emergency
department visits for asthma among children and adults: A case-crossover study in
northern Alberta, Canada. Environmental Health, 6(40).
Viscusi, W. K. 1992. Fatal Tradeoffs: Public and Private Responsibilities for Risk. Oxford
University Press: New York.
BenMAP User Manual v 0.01
R-23
December 2022
-------
References
Viscusi, W. K. and J. E. Aldy. 2003. The Value of a Statistical Life: A Critical Review of Market
Estimates throughout the World. AEI-Brookings Joint Center for Regulatory Studies.
Washington, DC. January.
Viscusi, W. K., W. A. Magat and J. Huber. 1991. Pricing Environmental Health Risks - Survey
Assessments of Risk - Risk and Risk - Dollar Trade-Offs For Chronic Bronchitis. Journal of
Environmental Economics and Management. Vol. 21 (1): 32-51.
Wei, Y., Y. Wang, Q. Di, C. Choirat, Y. Wang, P. Koutrakis, A. Zanobetti, F. Dominici, and J.D.
Schwartz. 2019. Short term exposure to fine particulate matter and hospital admission
risks and costs in the Medicare population: time stratified, case crossover study. BMJ
367:16258
Wilson, A. M., C. P. Wake, T. Kelly and J. C. Salloway. 2005. Air pollution, weather, and
respiratory emergency room visits in two northern New England cities: an ecological
time- series study. Environ Res. Vol. 97 (3): 312-21.
Winer, RA, Qin, X, Harrington, T, Moorman, J and Zahran, H. 2012. Asthma incidence among
children and adults: findings from the Behavioral Risk Factor Surveillance system
asthma call-back survey—United States, 2006-2008. J Asthma 49(1): 16-22.
Wittels, E. H., J. W. Hay and A. M. Gotto, Jr. 1990. Medical costs of coronary artery disease in
the United States. Am J Cardiol. Vol. 65 (7): 432-40.
Woodruff, T.J., Darrow, L.A., & Parker, J.D. 2008. Air Pollution and Postneonatal Infant
Mortality in the United States, 1999-2002. Environ Health Perspect 116(1): 110-115.
Woodruff, T. J., J. Grillo and K. C. Schoendorf. 1997. The relationship between selected causes
of postneonatal infant mortality and particulate air pollution in the United States.
Environmental Health Perspectives. Vol. 105 (6): 608-612.
Woodruff, T. J., J. D. Parker and K. C. Schoendorf. 2006. Fine particulate matter (PM2.5) air
pollution and selected causes of postneonatal infant mortality in California.
Environmental Health Perspectives. Vol. 114:786-790.
BenMAP User Manual v 0.01
R-24
December 2022
-------
References
Woods & Poole Economics Inc. 2012. Complete Demographic Database. Washington, DC.
http://www.woodsandpoole.com/index.php.
Wu, X., D. Braun, J. Schwartz, M.A. Kioumourtzoglou, and F. Dominici. 2020. Evaluating the
impact of long-term exposure to fine particulate matter on mortality among the elderly.
Science Advances 6: eaba5692
Yang, W, Hamilton, JL, Kopil, C, Beck, JC, Tanner, CM, Albin, RL, Ray Dorsey, E, Dahodwala, N,
Cintina, I, Hogan, P and Thompson, T. 2020. Current and projected future economic
burden of Parkinson's disease in the U.S. NPJ Parkinsons Dis 6:15.
Yao, J., Ghosh, K., Perraillon, M.C., Cutler, D.M, Fang, M.C. 2019. Trends and Racial Differences
in First Hospitalization for Stroke and 30-Day Mortality in the US Medicare Population
From 1988 to 2013. Medical Care 57(4): 262-269.
Zanobetti, A., M. Franklin, et al. 2009. Fine particulate air pollution and its components in
association with cause-specific emergency admissions. Environmental Health 8:58-60.
Zanobetti, A, Dominici, F, Wang, Y and Schwartz, JD. 2014. A national case-crossover analysis
of the short-term effect of PM2.5 on hospitalizations and mortality in subjects with
diabetes and neurological disorders. Environ Health 13(1): 38.
Zanobetti A. and Schwartz, J. 2006. Air pollution and emergency admissions in Boston, MA. J
Epidemiol Community Health 60 (10): 890-5.
Zanobetti, A. and Schwartz, J., 2008. Mortality displacement in the association of ozone with
mortality: an analysis of 48 cities in the United States. American Journal of Respiratory
and Critical Care Medicine 177(2): 184-189.
Zanobetti, A. and Schwartz, J. 2008b. Is there adaptation in the ozone mortality relationship:
A multi-city case-crossover analysis. Environmental Health, 7(22).
BenMAP User Manual v 0.01
R-25
December 2022
------- |