May 17th 2020
Estimating the True Case-Fatality Rate of COVID-19
For several months, we have been inundated by reports summarizing incidence and mortality data for COVID-19, on both the national and global level. In a previous ProMED post I cautioned that “reported cases” cannot be equated with “total cases” without inclusion of individuals with asymptomatic of sub-clinical infection that do not seek medical care. [1] If a large segment of the population is found to be seropositive, we might conclude that the true case-fatality ratio of COVID-19 is lower than official data might suggest. A seroprevalence study reported this week seems to provide solid evidence that this is the case. [2]
A national survey of individuals identified IgG antibody toward SARS-CoV-2 in 5.0% of the general population of Spain. At the stated specificity of 100% and sensitivity of 98%, the true seropositivity may be estimated at 5.15%. Although factors which determine seroprevalence rates in Spain need not apply to other countries, the following chart extrapolates the potential impact of a 5.15% population seroprevalence on case-fatality data from high-incidence countries in the European Region, United States and China. [3]
The impact of future SARS-CoV-2 seroprevalence surveys will largely depend in the quality of the test itself, the duration of immunity and protective role of the antibody, possible emergence of newer strains of coronavirus, and other factors. In any event, additional seroprevalence data will play a key role in planning our response to this pandemic going forward.
COVID-19 Reported vs. Estimated cases
| Country | Reported cases | % of population | Reported Deaths | CFR (%)* | Estimated cases** | Estimated CFR (%)*** |
|---|---|---|---|---|---|---|
| Belgium | 55,280 | 0.48% | 9,052 | 16.4 | 596,548 | 1.52 |
| China | 82,947 | 0.0058% | 4,634 | 5.6 | 74,125,174 | 0.0063 |
| France | 179,569 | 0.28% | 28,108 | 15.7 | 3,360,686 | 1.14 |
| Germany | 176,651 | 0.21% | 8,049 | 4.6 | 4,313,197 | 0.19 |
| Israel | 16,617 | 0.19% | 272 | 1.6 | 444,866 | 0.061 |
| Italy | 225,435 | 0.37% | 31,908 | 14.1 | 3,114,329 | 1.02 |
| Spain | 277,719 | 0.59% | 27,650 | 10 | 2,407,759 | 1.14 |
| U.K. | 243,695 | 0.35% | 34,636 | 14.3 | 3,493,878 | 0.99 |
| U.S.A | 1,527,664 | 0.46% | 90,978 | 6 | 17,034,349 | 0.53 |
* CFR = deaths / cases X 100
** True number of cases based on population seroprevalence of 5.15%
*** Adjusted CFR = deaths / estimated true cases X 100
References
- ProMED – What’s the denominator 20200228.7035438
- ProMED – Spain: seroprevalence study 20200516.7342334
- https://www.worldometers.info/coronavirus/ (status as of May 17)



If you search the Internet for countries which have reported COVID-19, an endless variety of sources will describe the status of this disease in 207 countries and their dependencies. Few if any of these sites mention countries where this disease does 




The recent WHO decision to declare the novel coronavirus outbreak a Public Health Emergency of International Concern (PHEIC), while both appropriate and hardly surprising, offers the opportunity to reflect on the previous PHEIC which was declared, namely the Ebola epidemic in Kivu region, Democratic Republic of the Congo (DRC). And you should really say the ongoing Ebola epidemic, as during the time since the declaration in July 2019 through to the present day (March 2020), a total of 3,453 cases have been reported [1].