Modeling to mitigate COVID-19

Modeling to mitigate COVID-19

Lauren Ancel MeyersUT COVID-19 Modeling Consortium

June 17, 2020

https://covid-19.tacc.utexas.edu/

UT COVID-19 Modeling Consortium https://covid-19.tacc.utexas.edu/ [email protected]

Background

Pandemic Exercise Tool (2013)

CDC FluCode - Pandemic Model (2020)

15% 35%

High risk


The key questions

Where and how is the virus spreading today?

Where will the virus be spreading in the future?

How to use limited resources to slow spread and save lives?

Situational awareness

Forecasting

Mitigation


Spread far and fast

-10 0 10 20Serial Interval (days)

0

3

6

9

12

15

Freq

uenc

y


0

3

6

9

12

15

Freq

uenc

y


0

3

6

9

12

15

Freq

uenc

y

Du et al. (2020) Serial Interval of COVID-19 among Publicly Reported Confirmed Cases. Emerging Infectious Diseases

Dec. 1

Dec. 8

Jan.

10

Jan.

22100

105

Cum

mul

ativ

e ca

ses

Cum

ulat

ive

case

s

Du et al. (2020) Risk for Transportation of 2019 Novel Coronavirus Disease from Wuhan to Other Cities in China. Emerging Infectious Diseases

425 cases

reported

12,400 cases

estimated

fast

silent

long


01-17

01-21

01-25

01-29

02-02

02-06

02-10

0

2

4

01-17

01-21

01-25

01-29

02-02

02-06

02-10

0

5

10

15

Rt

case

sEarly action matters

Du et al. (2020) Effects of Proactive Social Distancing on COVID-19 Outbreaks in 58 Cities, China. Emerging Infectious Diseases.

A 1-day delay in intervention prolongs the outbreak by ~2.4 days.

intervention

containment


The key questions

Where and how is it spreading today?

Where will it be spreading in the future?



Forecasting

Mitigation


The IHME Model

Apr 6

900

600

300

0De

aths

Apr 13

Backcast for Spain


Our model

bars grocery

parksmedical

schoolsrestaurants

at home


Ensemble forecasting


The key questions

Where and how is it spreading today?

Where will it be spreading in the future?



Forecasting

Mitigation


Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep

5000

4000

3000

2000

1000

0

Hos

pita

lizat

ions

capacity

How will reopening play out?

Mar Apr May

1501251007550250

If transmission increases by 50% relative to the stay-home period …

over 3500

deaths



3000

2500

2000

1500

1000

500

0

Hos

pita

lizat

ions

capacity

How will reopening play out?If transmission increases by 25% relative to the stay-home period …

~2100 deaths



10,000

8,000

6,000

4,000

2,000

0

Hos

pita

lizat

ions

capacity

How will reopening play out?If transmission doubles relative to the stay-home period …

> 6000 deaths


Decision-support modeling

Goals: Avoid overwhelming surge and avoid stay-home orders

Approach: Multiple stages of risk to allow ‘tapping’ on brakes

What to track?

When to trigger?


Policy thresholds

Use this color-coded alert system to understand the stages of risk. This chart provides recommendations on what people should do to stay safe during the pandemic. Individual risk categories identified pertain to known risks of

complication and death from COVID-19. This chart is subject to change as the situation evolves.

COVID-19: Risk-Based Guidelines

AustinTexas.gov/COVID19 Published: May 13, 2020

Stage 5

Stage 4

Stage 3

Stage 2

Stage 1 • greater than 25

greater than 25

all businesses

essential and re-opened businesses

essential and re-opened businesses

expanded essential businesses

essential businesses only

gathering size TBD

greater than 10

except with precautions



except as essential

except as essential

except as essential

except as essential

except expanded essential

businesses

except as essential

social and

greater than 10

social and

greater than 10

social and

greater than 10

outside of household

outside of household

social and

greater than 2

• • •

• • • •

• • • • •

• • • • •

Practice Good

Hygiene

Stay Home If Sick

Maintain Social

Distancing

Wear Facial Coverings

Higher Risk IndividualsAge over 65, diabetes, high blood

pressure, heart, lung and kidney disease, immunocompromised, obesity

Lower Risk IndividualsNo substantial underlying health

conditions

Workplaces Open

Avoid Gatherings

Avoid GatheringsAvoid Sick

People

Avoid Non-

Essential Travel

Avoid Non-

Essential Travel

Avoid Dining/

Shopping

Avoid Dining/

Shopping

1

Triggers 7-day average

daily COVID-19 hospital admissions

5

20

70



2000

1750

1500

1250

1000

750

500

250

0

Hos

pita

lizat

ions

capacity

stay

-hom

e

restricted relaxed

Projections with stages

~1800 deaths


Data-driven policy in action


Houston

4,500 beds

50179

380

7-day average COVID-19 admissions: 189

189

70450

780

9,000 beds

801140

1180

13,500 beds

189 189


Funding and teamSimon Cauchemez, Institute Pasteur

Ben Cowling, U Hong KongMaytal Dahan, TACC

Oscar Dowson, Northwestern UniversityZhanwei Du, UT

Daniel Duque Villarreal, NorthwesternSpencer Fox, UT

Kelly Gaither, TACCAlison Galvani, Yale

Neo Huang, Precima, LoyaltyOneEmily Javan, UT

Clay Johnston, Dell MedMichael Lachmann, Santa Fe Institute

David Morton, NorthwesternCiara Nugent, UTRemy Pasco, UT

Michaela Petty, UTKelly Pierce, TACC

Michael Pignone, Dell MedJames Scott, UTMauricio Tec, UT

Suzanna Wang, UTSpencer Woody, UT

US Centers for Disease Control and PreventionNational Institutes of Health

National Science FoundationTexas Advanced Computing Center

Tito’s Handmade Vodka

Modeling to mitigate COVID-19

Documents

Transcript of Modeling to mitigate COVID-19