For assignment, you will examine the statistical implications of mass COVID-19 testing. You will determine the anticipated PPV and NPV, you will analyze the possible sampling biases in the presented data, and you will identify possible correlations. Finally, you will examine the significance of these data implications on public policy.
Throughout the COVID-19 pandemic, there has been an urgency in many countries, to increase the testing of the general population to determine and contain the spread of the disease. The plan to control an epidemic is to identify and isolate all cases to stop the spread. This plan requires accurate identification of everyone that is infected and/or exposed to the disease. But no test is 100% accurate, so our project is to quantify the implications of testing accuracy during mass testing of COVID-19 and the real-life consequences of our findings on public policy.
1) Review the site below to understand Sensitivity, Specificity, Positive Predictive Value (PPV), and Negative Predictive Value (NPV).
Sensitivity, Specificity, PPV, and NPV
2) Review the accuracy of standard COVID-19 tests to determine their sensitivity and specificity.
The link below provides some information about the sensitivity and specificity of the rapid Covid tests. For our purposes, we will use the following:
Testing a population with symptoms:
· Sensitivity: 92.0%
· Specificity: 99.6%
Testing a population without symptoms:
· Sensitivity: 80.0%
· Specificity: 99.5%
Rapid, point‐of‐care antigen and molecular‐based tests for diagnosis of SARS‐CoV‐2 infection
3) Review the two links below to understand how to calculate false positives and false negatives given sensitivity, specificity, and prevalence.
False Positive Calculator
False Negative Calculator
4) Review the following two instances in time:
– First consider March 15, 2000, when the U.S. decided to shut down. At the time, the CDC reported 383 cases in the U.S. (7-day moving average). For our analysis, please look at the anticipated results of testing 1,000 with symptoms and 1,000,000 without symptoms.
– Second, consider the peak of the outbreak on or about January 11, 2021, with 250,836 reported cases. Since this date is later in the outbreak, please consider testing 10,000 with symptoms and 10,000,000 without symptoms.
– Consider that the U.S. has 330,000,000 people during this time (for our purposes, we can consider the population of the U.S. constant a 330 million). This data will allow calculating the prevalence of the disease at these two points.
The following two resources may also be helpful in these calculations and also considerations of how to implement the results. Additional resources are also encouraged.
Pitfalls of Mass Testing for COVID-19
The Positives and Negatives of Mass Testing for Coronavirus
5) After reviewing the data and learning more about calculations:
• Determine the prevalence of the disease at these two dates. We will assume that the prevalence of those with symptoms is 20 times the general population for our purposes.
• Calculate the true positives, true negatives, false positives, and false negatives.
• Calculate the PPV and NPV for these two times (March and January) for the rapid Covid test using the sensitivity and specificity given above.
• Interpret the results of this statistical analysis and how these results are or are not helpful for public policy to quarantine and contract traces to control the disease.
• Examine how sampling bias (Section 9.4 in the text) could affect the results across large areas and smaller communities.
For assignment, you will examine the statistical implications of mass COVID
–
19
testing. You will determine the anticipated PPV and NPV, you will analyze the possible
sampling biases in the presented data, and you will identify possible correlations.
Finall
y, you will examine the significance of these data implications on public policy.
Throughout
the
COVID
–
19
pandemic,
there
has
been
an
urgency
in
many
countries,
to
increase
the
testing
of
the
general
population
to
determine
and
contain
the
spread
of
the
di
sease.
The
plan
to
control
an
epidemic
is
to
identify
and
isolate
all
cases
to
stop
the
spread.
This
plan
requires
accurate
identification
of
everyone
that
is
infected
and/or
exposed
to
the
disease.
But
no
test
is
100%
accurate,
so
our
project
is
to
quanti
fy
the
implications
of
testing
accuracy
during
mass
testing
of
COVID
–
19
and
the
real
–
life
consequences
of
our
findings
on
public
policy.
1)
Review
the
site
below
to
understand
Sensitivity,
Specificity,
Positive
Predictive
Value
(PPV),
and
Negative
Predict
ive
Value
(NPV).
Sensitivity,
Specificity,
PPV,
and
NPV
2)
R
eview
the
accuracy
of
standard
COVID
–
19
tests
to
determine
their
sensitivity
and
specificity.
The
link
below
provides
some
information
about
the
sensitivity
and
specificity
of
the
rapid
Covid
tests.
For
our
purposes,
we
will
use
the
following:
Testing
a
population
with
symptoms:
·
Sensitivity:
92.0%
·
Specificity:
9
9.6%
Testing
a
population
without
symptoms:
·
Sensitivity:
80.0%
·
Specificity: 99.5%
Rapid,
point
–
of
–
care
antigen
and
molecular
–
based
tests
for
diagnosi
s
of
SARS
–
CoV
–
2
infection
3)
Review
the
two
links
below
to
understand
how
to
calculate
false
positives
and
false
negatives
given
sensitivity,
specificity,
and
prevalence
.
False
Positive
Calculator
False
Negative
Calculator
4)
Review
the
following
two
instances
in
time:
–
First
consider
March
15,
2000,
when
the
U.S.
decided
to
shut
down.
At
the
time,
the
CDC
reported
383
cases
in
the
U.S.
(7
–
day
moving
average).
For
our
analysis,
please
look
at
the
anticipated
results
of
testing
1,000
with
symptoms
and
1,000,000
without
s
ymptoms.
–
Second,
consider
the
peak
of
the
outbreak
on
or
about
January
11,
2021,
with
250,836
reported
cases.
Since
this
date
is
later
in
the
outbreak,
please
consider
testing
10,000
with
symptoms
and
10,000,000
without
symptoms.
For assignment, you will examine the statistical implications of mass COVID-19
testing. You will determine the anticipated PPV and NPV, you will analyze the possible
sampling biases in the presented data, and you will identify possible correlations.
Finally, you will examine the significance of these data implications on public policy.
Throughout the COVID-19 pandemic, there has been an urgency in many countries, to
increase the testing of the general population to determine and contain the spread of
the disease. The plan to control an epidemic is to identify and isolate all cases to stop
the spread. This plan requires accurate identification of everyone that is infected
and/or exposed to the disease. But no test is 100% accurate, so our project is to
quantify the implications of testing accuracy during mass testing of COVID-19 and the
real-life consequences of our findings on public policy.
1) Review the site below to understand Sensitivity, Specificity, Positive Predictive
Value (PPV), and Negative Predictive Value (NPV).
Sensitivity, Specificity, PPV, and NPV
2) Review the accuracy of standard COVID-19 tests to determine their sensitivity
and specificity.
The link below provides some information about the sensitivity and specificity of the
rapid Covid tests. For our purposes, we will use the following:
Testing a population with symptoms:
Sensitivity: 92.0%
Specificity: 99.6%
Testing a population without symptoms:
Sensitivity: 80.0%
Specificity: 99.5%
Rapid, point-of-care antigen and molecular-based tests for diagnosis of SARS-CoV-2
infection
3) Review the two links below to understand how to calculate false positives and
false negatives given sensitivity, specificity, and prevalence.
False Positive Calculator
False Negative Calculator
4) Review the following two instances in time:
– First consider March 15, 2000, when the U.S. decided to shut down. At the time, the
CDC reported 383 cases in the U.S. (7-day moving average). For our analysis, please
look at the anticipated results of testing 1,000 with symptoms and 1,000,000 without
symptoms.
– Second, consider the peak of the outbreak on or about January 11, 2021, with
250,836 reported cases. Since this date is later in the outbreak, please consider
testing 10,000 with symptoms and 10,000,000 without symptoms.
