COVID-19

Global Covid-19 Case Fatality Rates – a lay summary

Media and scientific reports on the coronavirus pandemic refer to estimates of Case Fatality Rate (CFR) and Infection Fatality Rate (IFR). What is the difference, why should the figures vary so much between countries, and how do they help us understand the risk of dying from COVID-19? 

Case Fatality Rate (CFR) refers to the death rate among all those who have actually been tested and diagnosed with COVID-19 (these are known as “cases”).

Infection Fatality Rate (IFR) is an estimate of the death rate among all those infected with SARS-CoV-19, the virus that causes the disease COVID-19. Because not everyone who has symptoms is tested, and not everyone who is infected ever shows any symptoms, you can see that the IFR is very much harder to infer.

Why do Case Fatality Rates (CFRs) vary for different countries?

The team at the Oxford Centre for Evidence Based Medicine examined the latest published data on cases and deaths for more than 100 countries, and found the CFRs varied enormously between countries, ranging from 16% in Belgium to 0.04% in Qatar. The CFR even changed over time within the same countries. This can be explained by a number of factors:

  • Testing strategy In most countries testing is reserved for those with severe disease. In the UK, for example, originally only patients deemed ill enough to require at least one night in hospital met the criteria for a COVID-19 test. Where testing is so limited, it is likely that many who have the disease are never tested,  so the number of “cases” used to calculate the CFR is artificially low, leading to an apparently higher CFR. In contrast, Iceland has tested more than one quarter of its entire 364,000 population, a higher proportion than any other country. This high rate of testing and the smaller population means their estimates of the COVID-19 CFR at 0.56% should be closer to the true figure.
  • Population age The risk of dying from coronavirus infection varies hugely with age. Data from China shows CFRs ranging from zero for cases aged less than nine, to 15% in the over-80’s. Italy has the second-oldest population worldwide (after Japan), and this will doubtless have contributed to an excess death rate there from COVID-19.
  • How healthy the population is to start with Chinese COVID-19 patients who had been otherwise healthy had an average CFR of 0.9%. But risk of death was higher in individuals who also had other conditions. Having cardiovascular disease, for example, makes you more vulnerable in most acute viral infections - whether COVID or seasonal influenza. And the likelihood of having other illnesses is greater in older and the socially disadvantaged. Smoking also seems to be associated with lower survival rates –  in Italy, 24% of the population are current smokers, compared to 15% in the UK.
  • How deaths are attributed to Coronavirus Dying with the disease is not the same as dying from the disease. In Italy, all coronavirus patients who died in hospital were recorded as having died of coronavirus. But a re-evaluation of death certificates by Italy’s National Institute of Health found that only 12% of these deaths had been directly caused by COVID-19.
  • Delay effect Deaths recorded today will be of people who started showing symptoms and became cases possibly two to three weeks ago. If the CFR is calculated is based on the number of deaths divided by the number of confirmed cases on that same day, the resulting figure can be misleading, especially when the infection rate is changing rapidly day by day.

What is the risk of dying from becoming infected (IFR)?

Considering how difficult it is even to establish the death rate per diagnosed case (CFR) in a consistent way, and knowing that many of those infected do not even show symptoms, you can see that estimating the risk of death from just becoming infected with the virus (IFR) is way more of a challenge. Antibody testing will help us understand how many people have been infected so far, and enable us to estimate the IFR more accurately. But until then, taking account of factors such as historical experience, trends in the data, and potential impact of misclassification of deaths, the COVID-19 IFR is currently estimated at somewhere between 0.1% and 0.41%.

Mandy Payne, 22nd May 2020

This is a lay summary of a report prepared for the Oxford COVID-19 Evidence Service by the Oxford Centre for Evidence Based Medicine by Jason Oke and Carl Heneghan and updated on 19th May 2020.

Disclaimer: the article has not been peer-reviewed; it should not replace individual clinical judgement and the sources cited in the original report should be checked. The views expressed in this commentary represent an interpretation by HealthWatch and do not necessarily represent those of Oxford CEBM, the NHS, the NIHR, or the Department of Health and Social Care. The views are not a substitute for professional medical advice.