UT Epidemiological Assessment of Construction

COVID-19 in Austin, Texas:
Epidemiological Assessment of
Construction Work
Remy Pasco, Dr. Zhanwei Du, Xutong Wang, Michaela Petty, Dr. Spencer J. Fox,
Dr. Lauren Ancel Meyers
CORRESPONDING AUTHOR
Lauren Ancel Meyers
The University of Texas at Austin
[email protected]
COVID-19 in Austin, Texas:
Epidemiological Assessment of
Construction Work
Remy Pasco, Dr. Zhanwei Du, Xutong Wang, Michaela Petty, Dr. Spencer J. Fox, Dr. Lauren
Ancel Meyers
Corresponding author:
Lauren Ancel Meyers
The University of Texas at Austin
[email protected]
Overview
There are an estimated 50,000 construction workers in the Austin metropolitan area
representing over 4% of the labor force [1], not accounting for undocumented workers.
The Austin Stay Home - Work Safe order that was issued on March 24, 2020 limits
construction work [2]. Since many construction workers live off of weekly income, those
restricted from non-essential worksites may seek work at essential worksites. This may
not only undermine efforts to reduce person-to-person contact, but exacerbate the
individual and city-wide risk by increasing the number of workers in close contact at
single construction sites.
In response to a request from the city of Austin, we projected the epidemiological

impacts of allowing some or all construction workers to resume work. To do so, we
modified the Austin-Round Rock module of our US COVID-19 Pandemic Model to
explicitly include a population subgroup representing construction workers. We
considered several different scenarios in which we varied the contact intensity among
construction workers at worksites and the proportion of workers allowed to work.
As a base case scenario, we assumed that construction workers would maintain typical
workforce contact rates, estimated for 18-49 year olds. As a highest risk scenario, we
assumed that the workers would have double the typical contact rate. This might be the
case if construction workers have higher than average contact rates in general, or if
contacts are elevated by workers migrating from a large number of non-essential to a
smaller number of essential worksites. As a lower risk scenario, we reduce the
workplace contacts by 50%. This might occur if precautionary measures are
implemented as has been discussed by city officials, including thorough cleaning of
equipment between uses, increased use of protective equipment such as gloves and
masks, limits on the number of workers on a given site, and ramped up health
surveillance on worksites including daily temperature readings, rapid COVID-19 testing
with symptoms, contact tracing and isolation of cases and known contacts of workers
who test positive for COVID-19.
The projections suggest that the incremental community risk of allowing construction
work depends on three key factors:
● Efficacy of Stay Home-Work Safe: Perhaps surprisingly, construction work is

most detrimental under scenarios where the social distancing order is highly
effective. If the order is highly effective, then the additional COVID-19
transmission within the construction worker community can undermine the strong
mitigation achieved. If the order is only moderately effective, the additional
transmission caused by construction work may be marginal.
● Size of construction workforce: Generally, the larger the population of active

workers, the faster COVID-19 will spread. However, this effect is only strong
under the scenario of highly effective social distancing.
● Risk of transmission at construction work sites: Generally, the greater the risk of
transmission at construction job sites, the faster COVID-19 will spread. As with
the size of the construction workforce, this is more apparent under the scenario
of highly effective social distancing.
Our projections also suggest that the risk of severe COVID-19 within the construction
workforce will be higher than that in the non-construction working 18-49 year old
populations. Large numbers of workers and high job site risk exacerbate this disparity.
Under a scenario of effective social distancing and a large construction workforce, the
hospitalization risk is expected to be two to three times higher for construction workers
than non-construction workers.
UT COVID-19 Consortium 2 April 5, 2020

Scenarios
We updated the Austin-Round Rock module of our US COVID-19 Pandemic Model to
simulate COVID-19 epidemics under various scenarios that allow partial or all
construction work to proceed during the Stay Home-Work Safe order. The simulations
ran from April 1 through mid-August, 2020. They assume the following initial conditions
and key parameters:
● Starting condition: April 1, 2020 with 346 infected adults
● Epidemic doubling time: 4 days [3]
● Reproduction number: 2.2 [4]
● Average incubation period: 6.9 days [5]
● Proportion of cases asymptomatic: 17.9% [6]
All other model parameters, including age-specific hospitalization and fatality rates are
provided in the Appendix.
All of the scenarios we analyzed assume that Austin’s Stay Home - Work Safe order
has effectively reduced non-household contacts by either 75% or 90%. We estimate the
impact of the Austin population of 50,000 construction workers continuing to work at the
following levels:
● 0%, 25%, 50%, 75% or 100% continue to work at construction sites
● Contact rates between active construction workers are either equal to baseline
contact rates for 18-49 year old workers, half of that baseline (50%) and twice
that baseline (200%)

Projections
Highly effective Stay Home-Work Safe (90%)

Overall increase in hospitalization risk: Assuming that the Stay Home-Work Safe
order is highly effective, allowing all construction work to proceed would be expected to
triple the number of COVID-19 hospitalizations in the Austin-Round Rock Area, under
the scenario that construction job sites have double the transmission risk of a typical
workplace (Figure 1). Measures to reduce risk of transmission at job sites could mitigate
this risk.
Figure 1. Projected cumulative hospitalizations in the Austin-Round

Rock MSA through mid-August under different workforce
scenarios. Bars and error bars indicate minimum, median and maximum
across 100 stochastic simulations. Shading indicates level of risk of
transmission at construction work sites: light gray is half the risk of a
typical workplace; medium gray is typical workplace risk; dark gray is
twice the risk of a typical workplace.

Increase in risk among construction workers: Assuming that the Stay Home-Work
Safe order is highly effective, allowing all construction work to proceed would
disproportionately increase risk among construction workers. Allowing all construction
work to proceed would lead to an estimated eight-fold increase in the number of
construction workers that are hospitalized for COVID-19 by mid-August, under the
scenario that construction job sites have double the transmission risk of a typical
workplace (Figure 2). Measures to reduce risk of transmission at job sites could mitigate
but not eliminate this risk.

Rock MSA construction workforce through mid-August under
different workforce scenarios. Bars and points indicate minimum,
median and maximum across 100 stochastic simulations. Shading
indicates level of risk of transmission at construction work sites: light gray
is half the risk of a typical workplace; medium gray is typical workplace risk;
dark gray is twice the risk of a typical workplace.

Projected COVID-19 hospitalizations under various construction workforce
scenarios: Assuming that Stay Home-Work Safe has reduced non-household contacts
by 90%, we estimate that construction work will slightly accelerate pandemic spread. If
all workers are permitted to continue work, we estimate that the cumulative
hospitalizations in the Austin-Round Rock MSA as a whole through mid-August will
increase by 9.2%, 39.5% or 175.5% depending on whether worksite conditions either
reduce COVID-19 transmission by 50%, do not impact onsite transmission, or increase
transmission two-fold, respectively (Figure 3). The most extreme scenario (100%
workforce with two-fold elevated worksite) would be expected to elevate COVID-19
hospitalizations beyond local healthcares capacity by early July. If precautions are taken
to reduce contacts on worksites, such a crisis could be averted.
Figure 3. Projected impact of construction work on COVID-19 hospitalizations in the Austin-Round

Rock MSA through mid-August, assuming that Austin’s Stay Home-Work Safe order has reduced
non-household contacts by 90%. From left to right, the graphs consider three scenarios for onsite
transmission: 50% reduced transmission through precautionary measures; average workplace
transmission rates for adults 18-49y; two-fold increased transmission relative to typical workplace
because of nature of construction work and/or elevated concentration of construction workers at essential
worksites. Colors indicate the fraction of the construction workforce allowed to work. Shaded area
indicates the estimated hospital capacity of 80% of 4299 hospital beds in the Austin-Round Rock MSA.

Moderately effective Stay Home-Work Safe (75%)
Overall increase in hospitalization risk: Assuming that the Stay Home-Work Safe
order is moderately effective, allowing all construction work to proceed would be
expected to increase the number of COVID-19 hospitalizations in the Austin-Round
Rock Area by at least 10,000 by mid-August, under the scenario that construction job
sites have double the transmission risk of a typical workplace (Figure 4). Measures to
reduce risk of transmission at job sites could mitigate this risk.

Rock MSA through mid-August under different workforce scenarios.
Bars and points indicate minimum, median and maximum across 100
stochastic simulations. Shading indicates level of risk of transmission at
construction work sites: light gray is half the risk of a typical workplace;
medium gray is typical workplace risk; dark gray is twice the risk of a
typical workplace.

Increase in risk among construction workers: Assuming that the Stay Home-Work
Safe order is effective, allowing all construction work to proceed would
disproportionately increase risk among construction workers. Allowing all construction
work to proceed would be expected to double the number of construction workers that
are hospitalized for COVID-19 by mid-August, under the scenario that construction job
sites have double the transmission risk of a typical workplace (Figure 5). Measures to
reduce risk of transmission at job sites could mitigate but not eliminate this risk.

Rock MSA construction workforce through mid-August under
different workforce scenarios. Bars and points indicate minimum,
median and maximum across 100 stochastic simulations. Shading
indicates level of risk of transmission at construction work sites: light gray
is half the risk of a typical workplace; medium gray is typical workplace risk;
dark gray is twice the risk of a typical workplace.

Projected COVID-19 hospitalizations under various construction workforce
scenarios: Assuming that Stay Home-Work Safe has reduced non-household contacts
by 75%, we estimate that construction work will slightly accelerate pandemic spread. If
all workers are permitted to continue work, we estimate that the cumulative
hospitalizations in the Austin-Round Rock MSA as a whole through mid-August will
increase by 5.9%, 10.4% or 30.3% depending on whether worksite conditions either
reduce COVID-19 transmission by 50%, do not impact onsite transmission, or increase
transmission two-fold, respectively (Figure 6). If construction work is completely banned,
we would expect COVID-19 hospitalizations to exceed local capacity around July 26,
2020. The most extreme scenario (100% workforce with two-fold elevated worksite)
would accelerate this crisis in healthcare by an estimated two weeks.
Figure 6. Projected impact of construction work on COVID-19 hospitalizations in the Austin-Round

Rock MSA through mid-August, assuming that Austin’s Stay Home-Work Safe order has reduced
non-household contacts by 75%. From left to right, the graphs consider three scenarios for onsite
transmission: 50% reduced transmission through precautionary measures; average workplace
transmission rates for adults 18-49y; two-fold increased transmission relative to typical workplace
because of nature of construction work and/or elevated concentration of construction workers at essential
worksites. Colors indicate the fraction of the construction workforce allowed to work. Shaded area
indicates the estimated hospital capacity of 80% of 4299 hospital beds in the Austin-Round Rock MSA.

Appendix
COVID-19 Epidemic Model Structure and Parameters

The model structure is diagrammed in Figure A1 and described in the equations below.
For each age and risk group, we build a separate set of compartments to model the transitions
between the states: susceptible (S), exposed (E), symptomatic infectious (IY), asymptomatic
infectious (IA), symptomatic infectious that are hospitalized (IH), recovered (R), and deceased
(D). The symbols S, E, IY, IA, IH, R, and D denote the number of people in that state in the given
age/risk group and the total size of the age/risk group is .
The model for individuals in age group and risk group is given by:
A Y H
where A and K are all possible age and risk groups, , , are relative infectiousness of the
I A , I Y , E compartments, respectively, 𝛽 is transmission rate, a,i is the mixing rate between age
A Y H
group a, i ∈ A , , , are the recovery rates for the I A , I Y , I H compartments, respectively, 𝜎
is the exposed rate, 𝜏 is the symptomatic ratio, 𝜋 is the proportion of symptomatic individuals
requiring hospitalization, 𝜂 is rate at which hospitalized cases enter the hospital following
symptom onset, 𝜈 is mortality rate for hospitalized cases, and 𝜇 is rate at which terminal patients
die.
We model stochastic transitions between compartments using the 𝜏-leap method[7,8] with key
parameters given in Table S1. Assuming that the events at each time-step are independent and
do not impact the underlying transition rates, the numbers of each type of event should follow
Poisson distributions with means equal to the rate parameters. We thus simulate the model
according to the following equations:

,
with
and where denotes the force of infection for individuals in age group and risk group and
is given by:
Figure A1. Compartmental model of COVID-19 transmission in a US city. Each subgroup (defined by
age, risk and worker-type) is modeled with a separate set of compartments. Upon infection, susceptible
individuals (S) progress to exposed (E) and then to either symptomatic infectious (IY) or asymptomatic
infectious (IA). All asymptomatic cases eventually progress to a recovered class where they remain
protected from future infection (R); symptomatic cases are either hospitalized (IH) or recover. Mortality (D)
varies by age group and risk group and is assumed to be preceded by hospitalization.

Table A1. Initial conditions, school closures and social distancing policies
Variable Settings
Initial day of simulation 4/1/2020
Initial infection number

346 symptomatic cases in 18-49y age group
in locations
Trigger to close school 4/1/2020
Closure Duration Until start of 2020-2021 school year (8/17/20)
a: social distancing
reduction of other
Two scenarios: [0.75, 0.9]
non-household
contacts
b: proportion
construction workers Five scenarios: [0, 0.25, 0.5, 0.75, 1]
who continue to work
c: contact rates at
work between active
construction workers
are equal to baseline
Three scenarios for scaling factor: [0.5, 1, 2]
contact rates for 18-49
year old works
multiplied by a scaling
factor
work_CW: contact
Work matrices provided in Tables S5.1-S5.4
matrix of construction
work_CW = work - work(1:5, 1:5)
workers
Home, work, other and school matrices provided in Tables S5.1-S5.4

Age-specific and
day-specific contact Weekday = home + (1-a)*(work + other) + b*c*work_CW
rates Weekend = home + (1-a)*(other)
Weekday holiday = home + (1-ɑ)*(other)

Table A2. Model parametersa
Parameters Best guess values Source
R0 2.2 Li et al. [3]
: doubling time 4 days Kraemer et al.[4]
Fitteda to obtain specified

: transmission rate 0.0260
given
: recovery rate on
asymptomatic Equal to
compartment
: recovery rate on
symptomatic
Verity et al. [9]
non-treated
compartment
: symptomatic
82.1 Mizumoto et al.[6]
proportion (%)
: exposed rate Lauer et al.[5]
P: proportion of
Du et al.[10]
pre-symptomatic (%) 12.6
: relative
infectiousness of
infectious individuals
in compartment E

: relative
infectiousness of
Set to mean of
infectious individuals 0.4653
in compartment IA
Overall: [0.0016, 0.0049, 0.0840, 1.0000,

IFR: infected fatality 3.3710]
Age adjusted from Verity et
ratio, age specific Low risk: [0.0009, 0.0022, 0.0339, 0.2520,
al.[9]
(%) 0.6440]
High risk: [0.0092, 0.0218, 0.3388, 2.5197,
6.4402]
Overall: [0.0019, 0.0060, 0.1027, 1.2182,

4.1066]
YFR: symptomatic
Low risk: [0.0011165, 0.0027 , 0.0412,
fatality ratio, age
0.3069, 0.7844]
specific (%)
High risk: [0.0112, 0.0265, 0.4126, 3.0690,
7.8443]
Estimated using 2015-2016

Behavioral Risk Factor
Surveillance System (BRFSS)
h : high-risk data with multilevel regression
proportion, age [8.2825, 14.1121, 16.5298, 32.9912, and poststratification using
specific (%) 47.0568] CDC’s list of conditions that may
increase the risk of serious
complications from
influenza[11–13]
rr : relative risk for

high risk people
Assumption
compared to low risk 10
in their age group
Austin Independent School District

School calendars
calendar (2019-2020, 2020-2021)[14]
a
Values given as five-element vectors are age-stratified with values corresponding to 0-4, 5-17, 18-49, 50-64, 65+ year age groups,
respectively.

Table A3 Hospitalization parameters
Parameters Value Source
: recovery rate in 14 day-average from admission

hospitalized 1/14 to discharge (UT Austin Dell
compartment Med)
Overall: [ 0.0487, 0.0487, 3.2876,

11.3373, 17.7330]
YHR: symptomatic
Low risk: [0.0279, 0.0215, 1.3215, Age adjusted from Verity et al.
case hospitalization
2.8563, 3.3873] [9]
rate (%)
High risk: [ 0.2791, 0.2146, 13.2154,
28.5634, 33.8733]
: rate of symptomatic
individuals go to
hospital, age-specific
5.9 day average from symptom

: rate from symptom
0.1695 onset to hospital admission
onset to hospitalized
Tindale et al.[15]
: rate from 14 day-average from admission

1/14
hospitalized to death to death (UT Austin Dell Med)
HFR: hospitalized
fatality ratio, age [4, 12.365, 3.122, 10.745, 23.158]
specific (%)
: death rate on
hospitalized
[0.0390, 0.1208, 0.0304, 0.1049, 0.2269]
individuals, age
specific
Healthcare capacity Hospital beds: 4299 Regional hospitals
a
The parameter is fitted through constrained trust-region optimization in SciPy/Python.[16] Given a
value of , a deterministic simulation is run based on central values for each parameter, from which we
can compute the implied . We (1) track the daily number of new cases (both symptomatic and
asymptomatic) during the exponential growth portion of the epidemic, (2) compute the log of the number
of new cases: and (3) use least squares to fit a line to this curve: . We
then estimate the reproduction number of the simulation for that specific value of as
where is the generation time given by . The optimizing function runs
until the resulting value of does not get closer to the target value.

Model modification to incorporate construction workforce
We assume there are currently 50,000 construction workers in the Austin-Round Rock
MSA, all in the 18-49 year-old age group. The proportion of construction workers at
high-risk of complications from COVID-19 is the same as the overall 18-49y age group
in the Austin-Round Rock MSA.
We extended our US COVID-19 Pandemic Model to include a separate population

subgroup representing construction workers. We moved 50,000 individuals from the
regular 18-49y low risk and high risk compartments into the corresponding construction
compartments. We also extended the contact matrices [17] governing transmission
between age groups to allow us to manipulate the number of construction workers and
intensity of their contacts separately from the rest of the workforce. Initially, we set their
contact rates equal to those of the entire 18-49y population, except that we assume that
all work contacts take place within the subgroup of construction workers. Social
distancing measures reduce work and other contacts for non-construction workers and
other contacts for construction workers (by either 75% or 90%). Tables A4.1-A4.4 give
the original contact matrices and Tables A5.1-A5.4 give the updated contact matrices
assuming 50,000 construction workers.
Let C(X)i,j denote the average daily number of contacts that a person in group i has
with people in group j at location X . Let w denote the proportion of construction
workers in the 18-49y group.
For each age group i the new work (W) contact matrix between groups other than
construction workers is unchanged:
C ′(W )i,j = C(W )i,j for j =/ C onstruction
C ′(W )i,Construction = 0
Construction workers only have contacts among themselves at work so:
C ′(W )Construction,Construction = ∑ C (W )18−49,j

j
C ′(W )Construction,j = 0 for other groups j
For contacts at home, school and other locations we assume that construction workers
have the same contact patterns as any other 18-49 years old individual. Then at those
locations (X) the contacts a person has with individuals in the 18-49y group is simply
split between the existing 18-49y column and the new construction column:
C ′(X)i,18−49 = C(X)i,18−49 * (1 − w)
C ′(X)i,Construction = C(X)i,18−49 * w
C ′(X)Construction,j = C ′(X)18−49,j
Original 5-age groups contact matrices
Table A4.1 Home contact matrix. Daily number contacts by age group at home.
0-4y 5-17y 18-49y 50-64y 65y+
0-4y 0.5 0.9 2.0 0.1 0.0
5-17y 0.2 1.7 1.9 0.2 0.0
18-49y 0.2 0.9 1.7 0.2 0.0
50-64y 0.2 0.7 1.2 1.0 0.1
65y+ 0.1 0.7 1.0 0.3 0.6
Table A4.2 School contact matrix. Daily number contacts by age group at school.
0-4y 5-17y 18-49y 50-64y 65y+
0-4y 1.0 0.5 0.4 0.1 0.0
5-17y 0.2 3.7 0.9 0.1 0.0
18-49y 0.0 0.7 0.8 0.0 0.0
50-64y 0.1 0.8 0.5 0.1 0.0
65y+ 0.0 0.0 0.1 0.0 0.0
Table A4.3 Work contact matrix. Daily number contacts by age group at work.
0-4y 5-17y 18-49y 50-64y 65y+
0-4y 0.0 0.0 0.0 0.0 0.0
5-17y 0.0 0.1 0.4 0.0 0.0
18-49y 0.0 0.2 4.5 0.8 0.0
50-64y 0.0 0.1 2.8 0.9 0.0
65y+ 0.0 0.0 0.1 0.0 0.0
Table A4.4 Others contact matrix. Daily number contacts by age group at other locations.
0-4y 5-17y 18-49y 50-64y 65y+
0-4y 0.7 0.7 1.8 0.6 0.3
5-17y 0.2 2.6 2.1 0.4 0.2
18-49y 0.1 0.7 3.3 0.6 0.2
50-64y 0.1 0.3 2.2 1.1 0.4
65y+ 0.0 0.2 1.3 0.8 0.6
Updated contact matrices with separate subgroup for construction

workers
Table A5.1 Home contact matrix. Daily number contacts by age group at home assuming
50,000 construction workers in Austin MSA.
0-4y 5-17y 18-49y 50-64y 65y+ Construction
0-4y 0.5 0.9 1.9 0.1 0.0 0.1
5-17y 0.2 1.7 1.8 0.2 0.0 0.1
18-49y 0.2 0.9 1.6 0.2 0.0 0.1
50-64y 0.2 0.7 1.2 1.0 0.1 0.1
65y+ 0.1 0.7 0.9 0.3 0.6 0.0
Construction 0.2 0.9 1.6 0.2 0.0 0.1
Table A5.2 School contact matrix. Daily number contacts by age group at school assuming
0-4y 1.0 0.5 0.4 0.1 0.0 0.0
5-17y 0.2 3.7 0.9 0.1 0.0 0.0
18-49y 0.0 0.7 0.8 0.0 0.0 0.0
50-64y 0.1 0.8 0.4 0.1 0.0 0.0
65y+ 0.0 0.0 0.0 0.0 0.0 0.0
Construction 0.0 0.7 0.8 0.0 0.0 0.0
Table A5.3 Work contact matrix. Daily number contacts by age group at work assuming
0-4y 0.0 0.0 0.0 0.0 0.0 0.0
5-17y 0.0 0.1 0.4 0.0 0.0 0.0
18-49y 0.0 0.2 4.5 0.8 0.0 0.0
50-64y 0.0 0.1 2.8 0.9 0.0 0.0
65y+ 0.0 0.0 0.1 0.0 0.0 0.0
Construction 0.0 0.0 0.0 0.0 0.0 5.5
Table A5.4 Others contact matrix. Daily number contacts by age group at other locations
assuming 50,000 construction workers in Austin MSA.
0-4y 0.7 0.7 1.7 0.6 0.3 0.1
5-17y 0.2 2.6 2.0 0.4 0.2 0.1
18-49y 0.1 0.7 3.1 0.6 0.2 0.2
50-64y 0.1 0.3 2.1 1.1 0.4 0.1
65y+ 0.0 0.2 1.2 0.8 0.6 0.1
Construction 0.1 0.7 3.1 0.6 0.2 0.2
Estimation of age-stratified proportion of population at high-risk for

COVID-10 complications
We estimate age-specific proportions of the population at high risk of complications from
COVID-19 based on data for Austin, TX and Round-Rock, TX from the CDC’s 500 cities project
(Figure A2).[18] We assume that high risk conditions for COVID-19 are the same as those
specified for influenza by the CDC.[11] The CDC’s 500 cities project provides city-specific
estimates of prevalence for several of these conditions among adults.[19] The estimates were
obtained from the 2015-2016 Behavioral Risk Factor Surveillance System (BRFSS) data using a
small-area estimation methodology called multi-level regression and poststratification.[12,13] It
links geocoded health surveys to high spatial resolution population demographic and
socioeconomic data.[13]

Estimating high-risk proportions for adults. To estimate the proportion of adults at high risk
for complications, we use the CDC’s 500 cities data, as well as data on the prevalence of
HIV/AIDS, obesity and pregnancy among adults (Table A6).
The CDC 500 cities dataset includes the prevalence of each condition on its own, rather than
the prevalence of multiple conditions (e.g., dyads or triads). Thus, we use separate co-morbidity
estimates to determine overlap. Reference about chronic conditions[20] gives US estimates for
the proportion of the adult population with 0, 1 or 2+ chronic conditions, per age group. Using
this and the 500 cities data we can estimate the proportion of the population pHR in each age
group in each city with at least one chronic condition listed in the CDC 500 cities data (Table
A6) putting them at high-risk for flu complications.
HIV: We use the data from table 20a in CDC HIV surveillance report[21] to estimate the
population in each risk group living with HIV in the US (last column, 2015 data). Assuming
independence between HIV and other chronic conditions, we increase the proportion of the
population at high-risk for influenza to account for individuals with HIV but no other underlying
conditions.
Morbid obesity: A BMI over 40kg/m2 indicates morbid obesity, and is considered high risk for
influenza. The 500 Cities Project reports the prevalence of obese people in each city with BMI
over 30kg/m2 (not necessarily morbid obesity). We use the data from table 1 in Sturm and
Hattori[22] to estimate the proportion of people with BMI>30 that actually have BMI>40 (across
the US); we then apply this to the 500 Cities obesity data to estimate the proportion of people
who are morbidly obese in each city. Table 1 of Morgan et al.[23] suggests that 51.2% of
morbidly obese adults have at least one other high risk chronic condition, and update our
high-risk population estimates accordingly to account for overlap.
Pregnancy: We separately estimate the number of pregnant women in each age group and
each city, following the methodology in CDC reproductive health report.[24] We assume
independence between any of the high-risk factors and pregnancy, and further assume that half
the population are women.
Estimating high-risk proportions for children. Since the 500 Cities Project only reports data
for adults 18 years and older, we take a different approach to estimating the proportion of
children at high risk for severe influenza. The two most prevalent risk factors for children are
asthma and obesity; we also account for childhood diabetes, HIV and cancer.
From Miller et al.[25], we obtain national estimates of chronic conditions in children. For asthma,
we assume that variation among cities will be similar for children and adults. Thus, we use the
relative prevalences of asthma in adults to scale our estimates for children in each city. The
prevalence of HIV and cancer in children are taken from CDC HIV surveillance report[21] and
cancer research report,[26] respectively.
We first estimate the proportion of children having either asthma, diabetes, cancer or HIV
(assuming no overlap in these conditions). We estimate city-level morbid obesity in children
using the estimated morbid obesity in adults multiplied by a national constant ratio for each age
group estimated from Hales et al.,[27] this ratio represents the prevalence in morbid obesity in
children given the one observed in adults. From Morgan et al.,[23] we estimate that 25% of
morbidly obese children have another high-risk condition and adjust our final estimates
accordingly.
Resulting estimates. We compare our estimates for the Austin-Round Rock Metropolitan Area
to published national-level estimates[28] of the proportion of each age group with underlying
high risk conditions (Table A6). The biggest difference is observed in older adults, with Austin
having a lower proportion at risk for complications for COVID-19 than the national average; for
25-39 year olds the high risk proportion is slightly higher than the national average.
Figure A2. Demographic and risk composition of the Austin-Round Rock MSA. Bars
indicate age-specific population sizes, separated by low risk, high risk, and pregnant. High risk
is defined as individuals with cancer, chronic kidney disease, COPD, heart disease, stroke,
asthma, diabetes, HIV/AIDS, and morbid obesity, as estimated from the CDC 500 Cities
Project,[18] reported HIV prevalence[21] and reported morbid obesity prevalence,[22,23]
corrected for multiple conditions. The population of pregnant women is derived using the CDC’s
method combining fertility, abortion and fetal loss rates.[29–31]

Table A6. High-risk conditions for influenza and data sources for prevalence estimation
Condition Data source
Cancer (except skin) CDC 500 cities[18]
Chronic kidney disease CDC 500 cities[18]
COPD CDC 500 cities[18]
Coronary heart disease CDC 500 cities[18]
Stroke CDC 500 cities[18]
Asthma CDC 500 cities[18]
Diabetes CDC 500 cities[18]
HIV/AIDS CDC HIV Surveillance report[21]
Obesity CDC 500 cities complemented with Sturm and Hattori[22] and
Morgan et al.[23]
Pregnancy National Vital Statistics Reports[29] and abortion data[30]

Table A7: Comparison between published national estimates and Austin-Round Rock MSA
estimates of the percent of the population at high-risk of influenza/COVID-19 complications
Austin Pregnant women
National
(excluding (proportion of age
Age Group estimates[27]
pregnancy) group)
0 to 6 months NA 6.8 -
6 months to 4 years 6.8 7.4 -
5 to 9 years 11.7 11.6 -
10 to 14 years 11.7 13.0 -
15 to 19 years 11.8 13.3 1.7
20 to 24 years 12.4 10.3 5.1
25 to 34 years 15.7 13.5 7.8
35 to 39 years 15.7 17.0 5.1
40 to 44 years 15.7 17.4 1.2
45 to 49 years 15.7 17.7 -
50 to 54 years 30.6 29.6 -
55 to 60 years 30.6 29.5 -
60 to 64 years 30.6 29.3 -
65 to 69 years 47.0 42.2 -
70 to 74 years 47.0 42.2 -
75 years and older 47.0 42.2 -
References
1. Workforce overview. In: Austin Chamber of Commerce [Internet]. [cited 30 Mar 2020].
Available:
https://www.austinchamber.com/economic-development/austin-profile/workforce/overview
2. Stay Home - Work Safe Order Information | AustinTexas.gov. [cited 30 Mar 2020].
Available: https://austintexas.gov/article/stay-home-work-safe-order-information
3. Li Q, Guan X, Wu P, Wang X, Zhou L, Tong Y, et al. Early Transmission Dynamics in

Wuhan, China, of Novel Coronavirus-Infected Pneumonia. N Engl J Med. 2020.
doi:10.1056/NEJMoa2001316
4. Kraemer MUG, Yang C-H, Gutierrez B, Wu C-H, Klein B, Pigott DM, et al. The effect of
human mobility and control measures on the COVID-19 epidemic in China. medRxiv. 2020.
Available: https://www.medrxiv.org/content/10.1101/2020.03.02.20026708v1
5. Lauer SA, Grantz KH, Bi Q, Jones FK, Zheng Q, Meredith HR, et al. The Incubation Period
of Coronavirus Disease 2019 (COVID-19) From Publicly Reported Confirmed Cases:
Estimation and Application. Ann Intern Med. 2020. doi:10.7326/M20-0504
6. Mizumoto K, Kagaya K, Zarebski A, Chowell G. Estimating the Asymptomatic Proportion of

2019 Novel Coronavirus onboard the Princess Cruises Ship, 2020. Infectious Diseases
(except HIV/AIDS). medRxiv; 2020. doi:10.1101/2020.02.20.20025866
7. Keeling MJ, Rohani P. Modeling Infectious Diseases in Humans and Animals. Princeton
University Press; 2011. Available:
https://play.google.com/store/books/details?id=LxzILSuKDhUC
8. Gillespie DT. Approximate accelerated stochastic simulation of chemically reacting

systems. J Chem Phys. 2001;115: 1716–1733. doi:10.1063/1.1378322
9. Verity R, Okell LC, Dorigatti I, Winskill P, Whittaker C, Imai N, et al. Estimates of the
severity of COVID-19 disease. Epidemiology. medRxiv; 2020.
doi:10.1101/2020.03.09.20033357
10. Du Z, Xu X, Wu Y, Wang L, Cowling BJ, Meyers LA. The serial interval of COVID-19 from
publicly reported confirmed cases. Epidemiology. medRxiv; 2020.
doi:10.1101/2020.02.19.20025452
11. CDC. People at High Risk of Flu. In: Centers for Disease Control and Prevention [Internet].
1 Nov 2019 [cited 26 Mar 2020]. Available: https://www.cdc.gov/flu/highrisk/index.htm
12. CDC - BRFSS. 5 Nov 2019 [cited 26 Mar 2020]. Available:

https://www.cdc.gov/brfss/index.html
13. Zhang X, Holt JB, Lu H, Wheaton AG, Ford ES, Greenlund KJ, et al. Multilevel regression
and poststratification for small-area estimation of population health outcomes: a case study
of chronic obstructive pulmonary disease prevalence using the behavioral risk factor
surveillance system. Am J Epidemiol. 2014;179: 1025–1033. doi:10.1093/aje/kwu018
14. Calendar of Events. In: Austin ISD [Internet]. [cited 26 Mar 2020]. Available:
https://www.austinisd.org/calendar
15. Tindale L, Coombe M, Stockdale JE, Garlock E, Lau WYV, Saraswat M, et al. Transmission
interval estimates suggest pre-symptomatic spread of COVID-19. Epidemiology. medRxiv;
2020. doi:10.1101/2020.03.03.20029983
16. minimize(method=’trust-constr’) — SciPy v1.4.1 Reference Guide. [cited 28 Mar 2020].

Available: https://docs.scipy.org/doc/scipy/reference/optimize.minimize-trustconstr.html

17. Prem K, Cook AR, Jit M. Projecting social contact matrices in 152 countries using contact
surveys and demographic data. PLoS Comput Biol. 2017;13: e1005697.
doi:10.1371/journal.pcbi.1005697
18. 500 Cities Project: Local data for better health | Home page | CDC. 5 Dec 2019 [cited 19
Mar 2020]. Available: https://www.cdc.gov/500cities/index.htm
19. Health Outcomes | 500 Cities. 25 Apr 2019 [cited 28 Mar 2020]. Available:
https://www.cdc.gov/500cities/definitions/health-outcomes.htm
20. Part One: Who Lives with Chronic Conditions. In: Pew Research Center: Internet, Science
& Tech [Internet]. 26 Nov 2013 [cited 23 Nov 2019]. Available:
https://www.pewresearch.org/internet/2013/11/26/part-one-who-lives-with-chronic-condition
s/
21. for Disease Control C, Prevention, Others. HIV surveillance report. 2016; 28. URL:
http://www cdc gov/hiv/library/reports/hiv-surveil-lance html Published November. 2017.
22. Sturm R, Hattori A. Morbid obesity rates continue to rise rapidly in the United States. Int J
Obes . 2013;37: 889–891. doi:10.1038/ijo.2012.159
23. Morgan OW, Bramley A, Fowlkes A, Freedman DS, Taylor TH, Gargiullo P, et al. Morbid
obesity as a risk factor for hospitalization and death due to 2009 pandemic influenza
A(H1N1) disease. PLoS One. 2010;5: e9694. doi:10.1371/journal.pone.0009694
24. Estimating the Number of Pregnant Women in a Geographic Area from CDC Division of
Reproductive Health. Available:
https://www.cdc.gov/reproductivehealth/emergency/pdfs/PregnacyEstimatoBrochure508.pdf
25. Miller GF, Coffield E, Leroy Z, Wallin R. Prevalence and Costs of Five Chronic Conditions in
Children. J Sch Nurs. 2016;32: 357–364. doi:10.1177/1059840516641190
26. Cancer Facts & Figures 2014. [cited 30 Mar 2020]. Available:
https://www.cancer.org/research/cancer-facts-statistics/all-cancer-facts-figures/cancer-facts
-figures-2014.html
27. Hales CM, Fryar CD, Carroll MD, Freedman DS, Ogden CL. Trends in Obesity and Severe
Obesity Prevalence in US Youth and Adults by Sex and Age, 2007-2008 to 2015-2016.
JAMA. 2018;319: 1723–1725. doi:10.1001/jama.2018.3060
28. Zimmerman RK, Lauderdale DS, Tan SM, Wagener DK. Prevalence of high-risk indications
for influenza vaccine varies by age, race, and income. Vaccine. 2010;28: 6470–6477.
doi:10.1016/j.vaccine.2010.07.037
29. Martin JA, Hamilton BE, Osterman MJK, Driscoll AK, Drake P. Births: Final Data for 2017.
Natl Vital Stat Rep. 2018;67: 1–50. Available:
https://www.ncbi.nlm.nih.gov/pubmed/30707672
30. Jatlaoui TC, Boutot ME, Mandel MG, Whiteman MK, Ti A, Petersen E, et al. Abortion
Surveillance - United States, 2015. MMWR Surveill Summ. 2018;67: 1–45.

doi:10.15585/mmwr.ss6713a1
31. Ventura SJ, Curtin SC, Abma JC, Henshaw SK. Estimated pregnancy rates and rates of
pregnancy outcomes for the United States, 1990-2008. Natl Vital Stat Rep. 2012;60: 1–21.
Available: https://www.ncbi.nlm.nih.gov/pubmed/22970648

UT Epidemiological Assessment of Construction

Uploaded by

Copyright:

Available Formats

UT Epidemiological Assessment of Construction

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

UT Epidemiological Assessment of Construction

Uploaded by

Copyright:

Available Formats

COVID-19 in Austin, Texas:

In response to a request from the city of Austin, we projected the epidemiological

● Efficacy of Stay Home-Work Safe​: Perhaps surprisingly, construction work is

● Size of construction workforce​: Generally, the larger the population of active

UT COVID-19 Consortium 2 April 5, 2020

● Starting condition: April 1, 2020 with 346 infected adults

● Epidemic doubling time: 4 days ​[3]

● Reproduction number: 2.2 ​[4]

● Average incubation period: 6.9 days ​[5]

● Proportion of cases asymptomatic: 17.9% ​[6]

● 0%, 25%, 50%, 75% or 100% continue to work at construction sites

UT COVID-19 Consortium 3 April 5, 2020

Highly effective Stay Home-Work Safe (90%)

Figure 1. Projected cumulative hospitalizations in the Austin-Round

UT COVID-19 Consortium 4 April 5, 2020

Figure 2. Projected cumulative hospitalizations in the Austin-Round

UT COVID-19 Consortium 5 April 5, 2020

Figure 3. Projected impact of construction work on COVID-19 hospitalizations in the Austin-Round

UT COVID-19 Consortium 6 April 5, 2020

Figure 4. Projected cumulative hospitalizations in the Austin-Round

UT COVID-19 Consortium 7 April 5, 2020

Figure 5. Projected cumulative hospitalizations in the Austin-Round

UT COVID-19 Consortium 8 April 5, 2020

Figure 6. Projected impact of construction work on COVID-19 hospitalizations in the Austin-Round

UT COVID-19 Consortium 9 April 5, 2020

COVID-19 Epidemic Model Structure and Parameters

UT COVID-19 Consortium 10 April 5, 2020

UT COVID-19 Consortium 11 April 5, 2020

Initial day of simulation 4/1/2020

Initial infection number

Trigger to close school 4/1/2020

Closure Duration Until start of 2020-2021 school year (8/17/20)

Home, work, other and school matrices provided in Tables S5.1-S5.4

UT COVID-19 Consortium 12 April 5, 2020

Parameters Best guess values Source

R0​ 2.2 Li et al. ​[3]

: doubling time 4 days Kraemer et al.​[4]

Fitted​a​ to obtain specified

: exposed rate Lauer et al.​[5]

UT COVID-19 Consortium 13 April 5, 2020

Overall: [0.0016, 0.0049, 0.0840, 1.0000,

Overall: [0.0019, 0.0060, 0.1027, 1.2182,

Estimated using 2015-2016

rr : relative risk for

Austin Independent School District

UT COVID-19 Consortium 14 April 5, 2020

Parameters Value Source

: recovery rate in 14 day-average from admission

Overall: [ 0.0487, 0.0487, 3.2876,

5.9 day average from symptom

: rate from 14 day-average from admission

Healthcare capacity Hospital beds: 4299 Regional hospitals

UT COVID-19 Consortium 15 April 5, 2020

We extended our ​US COVID-19 Pandemic Model​ to include a separate population

C ′(W )Construction,Construction = ∑ C (W )18−49,j

0-4y 0.5 0.9 2.0 0.1 0.0

5-17y 0.2 1.7 1.9 0.2 0.0

18-49y 0.2 0.9 1.7 0.2 0.0

50-64y 0.2 0.7 1.2 1.0 0.1

● Efficacy of Stay Home-Work Safe: Perhaps surprisingly, construction work is

● Size of construction workforce: Generally, the larger the population of active

● Epidemic doubling time: 4 days [3]

● Reproduction number: 2.2 [4]

● Average incubation period: 6.9 days [5]

● Proportion of cases asymptomatic: 17.9% [6]

R0 2.2 Li et al. [3]

: doubling time 4 days Kraemer et al.[4]

Fitteda to obtain specified

: exposed rate Lauer et al.[5]

We extended our US COVID-19 Pandemic Model to include a separate population

Cancer (except skin) CDC 500 cities[18]

Chronic kidney disease CDC 500 cities[18]

COPD CDC 500 cities[18]

Coronary heart disease CDC 500 cities[18]

Stroke CDC 500 cities[18]

Asthma CDC 500 cities[18]

Diabetes CDC 500 cities[18]

HIV/AIDS CDC HIV Surveillance report[21]

Pregnancy National Vital Statistics Reports[29] and abortion data[30]