Modeling COVID-19 to Inform State and Local Response





Current R-effective in California

The effective reproduction number (R-effective) is the average number of people each infected person will pass the virus onto and represents the rate at which COVID-19 is spreading.


Latest Estimate of R-effective is:

Low/High Estimates of R-effective:

Download R-eff Values

Statewide Estimates of R-effective

The effective reproductive number (R) is the average number of secondary infected persons resulting from a infected person. If R>1, the number of infected persons will increase. If R<1, the number of infected persons will decrease. At R=1, the number of infected persons remains constant.

Latest Estimate of R-effective is:

Download County R-eff Trend
NOTE: Some counties do not have sufficient case numbers in order for modelers to estimate R-effective.

R-effective Trends by County

Select a county to see how R-effective has changed over time

Download R-eff for Counties

Latest R-effective in California Counties

Using estimates and uncertainty intervals from Covid Act Now, the following graph presents the averaged R-effective over the last 7 days by county; (<1 decreasing spread, >1 increasing spread)

Covid Act Now data has been pulled for further review. Check back soon.

Download R-eff for Counties

Statewide Map of R-effective by County

The following map presents the ensemble R-effective, averaged over the last 3 days by county; (<1 decreasing spread, >1 increasing spread)

Predefined County Regions

Predefined regions include commonly used 'clusters' of counties that either share demographic, economic, planning, or organizational characteristics or purposes in California.
Note: Regional ensembles are population weighted means of the R-effective for counties identified in the region.
Latest Estimate of R-effective is:

County Adjacency

The county adjacency panel allows users to examine 'clusters' of counties by selecting a focal county, and examining all counties that are adjacent, or geographically nearest to the focal county.
Note: Regional ensembles are population weighted means of the R-effective for counties identified in the region.
Latest Estimate of R-effective is:

Custom Regions

The custom regions panel allows users to customize regions by selecting multiple counties.
Note: Regional ensembles are population weighted means of the R-effective for counties identified in the region.
Latest Estimate of R-effective is:

Short-term COVID-19 Forecasts in California

Short-term forecasts take into account the most recent trends in cases, hospitalizations and deaths and apply statistical models to that data to generate anticipated trends in the coming 2-4 weeks.


Current Daily Hospitalizations:

Projected Daily Total:

Download Hospital Forecasts

Statewide Hospitalization Forecasts

The black box (left) represents the current number of hospitalized COVID patients in California. The blue box represents the forecasted number of hospitalized patients at the 30 day mark based on models for California.

California County Hospitalization Forecasts

Select a county to see how modeled number of hospitalizations compare with actual numbers to date and with the number of licensed hospital beds (black box).

Current Daily Hospitalizations:

Projected Daily Total:

Download County Hospital Forecasts

Current Total Deaths:

Projected Total:

Download Total Death Forecasts

Statewide Total Death Forecast

The black box (left) represents the current number of total COVID deaths in California. The blue box represents the COVIDhub (ReichLab) forecasted number of total deaths at the 30 day mark based on models for California.

California County Death Forecasts

Select a county to see how modeled number of cumulative deaths compare with actual numbers to date (black box).

Current Total Deaths:

Projected Total:

Download Total Death Forecasts

Long term Scenarios

Long-term scenarios estimate the effect of various non-pharmaceutical interventions (NPIs)


Note: Detailed model scenario descriptions can be found below the graph or on the Technical Notes tab.


Johns Hopkins Model Options:
IHME Model Options:
RAND Model Options:
Legend

Download Scenarios in Graph
Use slide below graph to adjust date range; will also adjust date range of static plot.
Note: Some modelers do not produce scenario models for all counties.

Selected Model Descriptions - See Technical Notes tab for more information




Nowcasts

Rt.live

rt.live

Rt.live provides a state-level estimate of R-effective, taking the number of cases and the input. It accounts for the delay from infection to onset of symptoms and changes in the amount of testing done.

Raw data
Code repositorty

Covid Act Now

Covid Act Now

Covid Act Now provides statewide and county-level estimates of R-effective, taking mortality and confirmed cases as inputs. Because of potential reporting delays and errors in the data, they perform smoothing, and require 10 preceding days of data. County-level estimates on the "Nowcasts" tab show average R-effective for the last seven (7) days.

Raw data
Code repositorty
Modeling methods

Epi Forecasts

epiforecasts.io

Epi Forecasts provides national and state-level estimates of R-effective, taking the number of cases as an input. It accounts for the delay from infection to onset of symptoms.

Raw data
Code repository
Modeling methods

covidestim

covidestim.org

Covidestim calculates state-level effective reproductive numbers, taking cases, deaths and test positivity rates as inputs. It has corrections to account for lags in diagnosis, disease duration and mortality risk.

Code repository
Modeling methods

Youyang Gu's Covid-19 Projections

covid19-projections.com

COVID-19 Projections fits a parameterized S-curve for R-effective to minimize the mean-squared error of historical daily mortality data. COVID-19 Projections provides state-level estimates for R-effective, mortality and the infected population.

Raw data
Code repository
Modeling methods

UCLA Machine Learning Lab

covid19.uclaml.org

The UCLA ML Lab provides state and California county projections of mortality, the number of confirmed cases, and hospitalizations/ICU beds. At the state and county levels in CA, they also publish the current R-effective estimate. UCLA ML uses a modified SEIR model with a compartment for unreported cases. The model parameters learn to minimize the historical prediction error for the number of confirmed cases and deaths.

Raw data
Modeling methods

Imperial College London

Imperial College London COVID-19 State-Level Tracking

ICL provides state-level infection and mortality projections. The reproduction number R_t (which epidemiologists also call R_e, or R-effective) is calibrated using mobility data. Using Bayesian methods, the model calculates backwards from the deaths observed over time to estimate transmission that occurred several weeks prior, and to predict the current rate of transmission.

Raw data
Code repository
Modeling methods

LEMMA

localepi.github.io

LEMMA is an open-source SEIR model with compartments for hospitalization and symptom severity. The model is calibrated to hospitalization, ICU, and death data using Bayesian methods. LEMMA provides credibility intervals and scenarios for future hospitalization, ICU, deaths, and R_t. Users can download the LEMMA package and input their own data and priors for parameters using R or a simple Excel interface.

Raw data
Code repository

Stanford

www.sc-cosmo.org

SC-COSMO is an age-structured, multi-compartment SEIR model calibrated to reported case numbers using a Bayesian approach. The model incorporates contacts patterns by age, the effect of population density, and estimates of the case detection rate. SC-COSMO explicitly considers contacts and transmission in households as well as contacts in work, school, and other settings and the effects of non-pharmaceutical interventions like shelter-in-place and school closures that differentially reduce contacts by venue. Model outputs include number of infections, deaths, and hospitalizations.

Modeling methods

UCSF

ca-covid-r.info

The UCSF model estimates time-varying reproduction numbers (R_t, or R-effective), the average number of cases infected by a given case over the course of that individual’s disease progression, for select Bay Area and California counties/regions. Model inputs are publically available daily counts of COVID-19 cases, archived by the New York Times. The UCSF researchers use the Wallinga-Teunis technique of real-time estimation of reproduction numbers.

Raw data
Modeling methods

Ensemble Nowcast

The ensemble nowcast takes the median of all the nowcasts available on a given date and smooths it with a three-day moving average. The methodology aims to be robust to outliers and to avoid kinks in the median nowcast when an input source for the ensemble estimate is unavailable on a certain date.

Forecasts

Covid Act Now

covidactnow.org

The CovidActNow model is a SEIR model with compartments for disease severity and medical intervention. Each county and state is calibrated separately, and R-effective is inferred using observed data.

Raw data
Code repository
Modeling methods

Institute for Health Metrics and Evaluation

covid19.healthdata.org

IHME is a multi-stage model, where the first stage fits an S-curve to historical daily deaths data, and the second stage is an SEIR compartment model. The SEIR model's R-effective is calibrated using the output of the first stage, but it also incorporates temperature data, population density, local testing capacity, and changes in mobility data. IHME provides projections of mortality, number of infections, and hospital utilization at the state and national level.

Raw data
Modeling methods

Northeastern MOBS GLEAM Project

covid19.gleamproject.org

The Global Epidemic and Mobility Model (GLEAM) uses a individual-based, stochastic spatial epidemic model. The model uses mobility data and travel patterns to simulate spatial contact patterns. The likely ranges of basic parameters, such as R0 and IFR, are inferred from observed data.

Raw data
Modeling methods

MIT Operations Research

covidanalytics.io

MIT DELPHI is a standard SEIR model with compartments for undetected cases and hospitalizations. R-effective is modeled as an S-curve to reflect government interventions and social distancing.

Raw data
Code repository
Modeling methods

UC Berkeley Yu Group in Statistics and EECS

covidseverity.com

A statistical machine learning extrapolation algorithm CLEP which forecasts deaths with MEPI prediction intervals with one week or two in advance by county. The forecast is calculated from an ensemble of linear and exponential predictors (CLEP), some of which pool data across counties or use demographic data.

Raw data
Code repository
Modeling methods

COVID-19 Forecast Hub

covid19forecasthub.org

The Reich Lab at the UMass-Amherst is an Influenza Forecasting Center of Excellence and the source for the official CDC COVID-19 Forecasting page. Taking other forecasts as the input, this is arithmetic average across eligible models of cumulative deaths forecasts. Forecasts are weekly out to four (4) weeks, at the state and national level.

Raw data
Code repository

Johns Hopkins University Infectious Disease Dynamics Group

www.iddynamics.jhsph.edu

The COVID Scenario Pipeline is a county-level metapopulation model that incorporates commuting patterns and stochastic SEIR disease dynamics. To produce near-term forecasts of deaths and hospitalizations in the population, county-specific transmission and county-specific risks of hospitalization and death were inferred using a novel Bayesian inference algorithm. This algorithm calibrates the model to weekly county-level incident cases and deaths, as reported by USAFacts. Fixed time delays from infection to case confirmation, hospitalization, and death were employed in projecting these outcomes. The estimates reported by this model incorporate uncertainty in baseline R0, the duration of the infectious period, the effectiveness of statewide intervention policies, and process stochasticity.

Code Repository
Scenario Modeling Methods
Inference Methods

LEMMA

localepi.github.io

LEMMA is an open-source SEIR model with compartments for hospitalization and symptom severity. The model is calibrated to hospitalization, ICU and death data using Bayesian methods. LEMMA provides credibility intervals and scenarios for future hospitalization, ICU, deaths and Rt. Users can download the LEMMA package and input their own data and priors for parameters using R or a simple Excel interface.

Raw data
Code repository

Stanford

www.sc-cosmo.org

SC-COSMO is an age-structured, multi-compartment SEIR model calibrated to reported case numbers using a Bayesian approach. The model incorporates contacts patterns by age, the effect of population density, and estimates of the case detection rate. SC-COSMO explicitly considers contacts and transmission in households as well as contacts in work, school, and other settings and the effects of non-pharmaceutical interventions like shelter-in-place and school closures that differentially reduce contacts by venue. Model outputs include number of infections, deaths, and hospitalizations.

Modeling methods

Columbia

Columbia University, Shaman Group

The Columbia model projects nationwide, county-level estimates of R-effective, daily new confirmed case, daily new infection (both reported and unreported), cumulative demand of hospital beds, ICU, and ventilators, as well as daily mortality (2.5, 25, 50, 75 and 97.5 percentiles). County-level case and death data are compiled from Johns Hopkins University and USAFACTS.

Raw data
Modeling methods

Ensemble Forecast

The ensemble forecast takes the median of all the forecasts available on a given date and fits a smoothed spline to the trend. The methodology aims to be robust to outliers, and to avoid artifacts (i.e. abrupt kinks) when the median forecast switches from one source to another.

Scenarios

Johns Hopkins University Infectious Disease Dynamics Group

www.iddynamics.jhsph.edu

The COVID Scenario Pipeline is a county-level metapopulation model that incorporates commuting patterns and stochastic SEIR disease dynamics. To produce long-term planning scenarios, the model is calibrated to weekly county-level incident cases and deaths as reported by USAFacts using a novel Bayesian inference algorithm. The model projects into the future by making assumptions about the effectiveness of scenarios in different interventions, using fixed time delays from infection to case confirmation, hospitalization, and death, and location-specific risks of hospitalization, ICU admission, and death. The estimates reported by this model incorporate uncertainty in baseline R0, the duration of the infectious period, the effectiveness of statewide intervention policies, and process stochasticity.

Code Repository
Scenario Modeling Methods
Inference Methods

Covid Act Now

covidactnow.org

The CovidActNow model is a SEIR model with compartments for disease severity and medical intervention. Each county and state is calibrated separately, and R-effective is inferred using observed data.

Raw data
Code repository
Modeling methods

Institute for Health Metrics and Evaluation

covid19.healthdata.org

IHME is a multistage model, where the first stage fits an S-curve to historical daily deaths data, and the second stage is an SEIR compartment model. The SEIR model's R-effective is calibrated using the output of the first stage, but it also incorporates temperature data, population density, local testing capacity, and changes in mobility data. IHME provides projections of mortality, number of infections, and hospital utilization at the state and national level.

Raw data

RAND

rand.org

RAND is an SEIR model with compartments for symptom severity and hospitalization, stratified by age and health status (to account for vulerable populations with chronic disease). R is estimated from a regression of the disease's growth rate. Policy interventions adjust the matrix contact rates, which account for age group and mode of interaction (such as home or school or work).

Modeling methods