Fatality rates are at the heart of the debate about reopening the economy. Overestimating the threat of the COVID-19 could cause unnecessary job losses, but underestimating it means more lives lost.

Case fatality rate

How many people with confirmed COVID-19 cases end up dying?

The “best guess” models that US federal health officials are using to project the COVID-19 death toll currently estimate that 0.5% of infected people who show symptoms would die. This percentage is the case fatality rate, or CFR. Four of seven outside experts interviewed by the Center for Public Integrity said that the infection fatality rate estimates in the government models were too low.[1]

Epidemiologists are looking at other countries where we have better data on case fatality rates because of better testing, like Germany, South Korea, Singapore, Taiwan, and also the Diamond Princess cruise ship. That data appears to trend toward a case fatality rate of about 2%.

One of the reasons to look at data from other countries instead of data from the US right now is that estimates for the case fatality rate for an outbreak will evolve over time. You can’t accurately measure this metric when you’re still in the middle of an outbreak.

Infection fatality rate

That’s just accounting for confirmed cases, but what’s the true fatality rate? We’re actually interested in that number. How many infected people end up dying, not just the official cases?

There’s a gap between official cases and true cases. The fatality rate among the true cases is the infection fatality rate, or IFR.

The best data comes from the countries with the highest testing rates. Those places will have the smallest gap between the case fatality rate and the infection fatality rate. Looking at data from those countries, estimating asymptomatic infections at between 20% and 50% of total infections, the infection fatality rate is between 1% and 1.6% [2] One of the most optimistic estimates that only looked at the Diamond Princess put the infection fatality rate at 0.66%, which is still higher than the US federal government’s “best guess” estimate of 0.5%.[3]

We do at least know a lower bound on the infection fatality rate, since more than one in one thousand people in New York have already died from COVID-19. That’s not one in one thousand cases or one in one thousand infections, that’s one in one thousand people. So we know that the infection fatality rate can’t be lower than 0.1%, and that would be assuming that 100% of people in New York have already been infected.

Overall, experts generally think that the infection fatality rate is more likely to end up above 1% than below. That makes herd immunity look much less attractive as a control strategy.

Fatality rates change

It’s crucial to remember that there is no one true case fatality rate or infection fatality rate. These metrics change over time, and they change in different contexts. The fatality rates vary by age, by country, and according to other factors like the availability of health care and whether the health care systems are overwhelmed. The true infection fatality rate will be a lot higher in developing regions than in developed nations. And one of the key findings of the Imperial College model is that the infection fatality rate would be much higher without intervention.[4]

Fatality rates by age

An example of the fatality rate varying is the well-known finding that COVID-19 is much more deadly for older people. We can talk about an overall case fatality rate for all of the people represented in this chart for New York, but the real fatality rates vary for different age brackets.

The worldwide infection fatality rate

The fatality rate data coming from around the world right now is coming from developed nations. Many factors will probably lead to higher fatality rates in lower-income countries, including the availability of health care. The number of hospital beds, ICU beds, and health care workers per capita will have a significant influence on fatality rates.

If the overall average infection fatality rate in developed nations is in the 1% to 2% range, then the worldwide average death toll could be significantly higher.

Implications

Assuming 90% of the world’s population infected during the entire pandemic due to overshoot, a 1% infection fatality rate would result in about 69 million deaths worldwide.

A 2% infection fatality rate would lead to around 138 million deaths.

A 3% rate would lead to 207 million deaths worldwide. And the worldwide average rate could be higher.

Conclusions

These staggering projections for deaths illustrate that herd immunity is not a viable strategy because it would lead to an unprecedented humanitarian disaster.

From a local perspective, if you interact with a lot of people every day, then you’re likely to be exposed to the virus. If you assume that you will be infected, then epidemiologists currently estimate that you might have about a 0.5% to 2% chance of dying.

References

  1. 1. Whyte LE. What Happens If U.S. Reopens Too Fast? Documents Show Federal Coronavirus Projections. NPR; link
  2. 2. Ferretti1 L, Wymant C, Kendall M, et al. Quantifying SARS-CoV-2 transmission suggests epidemic control with digital contact tracing. Science; link
  3. 3. Robert Verity PD, Lucy C Okell PD, Ilaria Dorigatti PD, et al. Estimates of the severity of coronavirus disease 2019: a model-based analysis. The Lancet; link
  4. 4. Walker PGT, Whittaker C, Watson O, et al. Report 12: The Global Impact of COVID-19 and Strategies for Mitigation and Suppression. Imperial College London - MRC Centre for Global Infectious Disease Analysis; 26-April-2020 link; DOI: 10.25561/77735 (Accessed 2020-04-19 11:30)