Verisk COVID-19 Projection Tool: Frequently Asked Questions (FAQ)

What is the COVID-19 Projection Tool?

Developed using ArcGIS, a powerful geospatial analytics engine, the Verisk COVID-19 Projection Tool is an interactive dashboard that leverages the AIR Pandemic Model to provide projections of both total cases and deaths for COVID-19. These modeled projections for the next four weeks of the pandemic are available globally at a country level.

Why was the tool developed? Who is it for?

This publicly available tool was developed to enhance understanding the potential number of worldwide COVID-19 cases and deaths by individuals, communities, businesses, governments, and insurers.

Why are the numbers in these projections so much higher than what I see on the news?

Variance in publicly reported estimates stems from a number of factors, not the least of which are misreporting and underreporting of infections and deaths. Underreporting is driven by the fact that as much as 60-80% of people infected with COVID-19 may exhibit no symptoms or only mild symptoms and do not get tested. The cases of COVID-19 that are reported, therefore, represent at best only 20-40% of the total population. Testing more widely in the general population would lead to more accurate numbers of cases and deaths being reported. The projections from the Verisk COVID-19 Projection Tool account for underreporting.

The Verisk COVID-19 Projection Tool provides a range of scenarios for total cases and deaths. The low and high bounds of the scenarios should not be considered best- and worst-case outcomes, but rather should be considered confidence intervals—one might anticipate actual results will fall within these ranges. It is possible, however, that due to significant uncertainty around this event, the actual numbers fall out of the bounds of these estimates.

Please review our Methods and Assumptions document for more details.

Why do some countries show zeroes for cases and deaths? Is this accurate?

We have excluded countries that have fewer than 100 reported cases as the sample size is too small for a stochastic model to provide useful projections. For this set of countries, you will see zeroes for cases and deaths.

Why are the Total Estimated Deaths so much higher than what is shown by other sources?

The Total Estimated Deaths shown in the tool include both those who have died and will die from COVID-19.

There are three dimensions of mortality data that make up the deaths projected in our tool:

  • Reported deaths: These generally make up the largest proportion of total deaths in our projections and will be similar to what you see in the news and other sources.
  • Unreported deaths: These are deaths that are either unreported or are reported as being caused by another disease or condition, such as pneumonia, stroke, ARDS, etc., yet a COVID-19 infection contributed to mortality. Therefore, the cause of death is not listed as COVID-19 and will not be reported as such.
  • Infected and will die: This number includes those who are currently infected with COVID-19 and will die as a result of the disease.

What is a catastrophe model?

Catastrophe models are used to understand the impact of low-frequency, high-severity events, such as pandemics where historical data is limited. Typically, they are used to estimate probabilities of loss to help organizations manage their risk. The outputs provided by these tools consider many variations in data availability and reliability, including underreporting.

Does this tool take mitigation factors into account, such as social distancing? At what resolution?

The modeled projections provided in the Verisk COVID-19 Projection Tool account for the effectiveness of containment measures, including the impact of social distancing, isolation, quarantining, and other mitigation factors. The low, moderate, and high scenarios provided give you the means to understand how effective these containment measures are at mitigating the spread of the disease.

These factors are taken into account at a country level globally.

Please review our Methods and Assumptions document for more details.

Which data sources does AIR use to inform and validate its model?

We integrate data sources that include current epidemiological and scientific literature and societal and governmental factors, then apply our judgment to develop our best estimates of realistic disease spread and severity rates.

Specific examples of our data sources are Johns Hopkins University, the World Health Organization Situation Reports, and data from local health ministries. We do not rely on a single data set because each organization can have a different definition of what constitutes an “infected” individual.

Please review our Methods and Assumptions document for more details.

How often will the data be updated and why?

The Verisk COVID-19 Projection Tool provides rolling, 4-week modeled projections of the progression of the COVID-19 pandemic at any given time. The data will be updated on a daily basis to project one day further into the future. The projections themselves will be refreshed weekly, most likely every Wednesday.

We limit our projections to 4 weeks into the future because uncertainty grows significantly outside that time frame.

Why are some projections so different week to week?

Projections may vary significantly from week to week. Our modeled simulations take into account the latest data available, response/mitigation efforts, changes in hospital capacity, and other factors, which our projections reflect.

Please review our Methods and Assumptions document for more details.

What is the cost?

We are providing this tool free of charge. Additional analyses, projections, and metrics (financial, morbidity, and mortality) can be accessed by licensing the AIR Pandemic Model or through a consulting agreement.

Who should I contact to learn more?

Please click here to contact us with your comments, questions, or suggestions.



You’re almost done.
We need to confirm your email address.
To complete the registration process, please click the link in the email we just sent you.

Unable to subscribe at this moment. Please try again after some time. Contact us if the issue persists.

The email address  is already subscribed.

You are subscribing to AIR Blogs. In order to proceed complete the captcha below.