Case Fatality Rate of COVID-19 by Country

At this point in the COVID-19 pandemic it is difficult to get a reliable estimate of how deadly the disease might be. A useful metric that can be calculated from publicly available data is the case fatality rate (CFR) which is the proportion of persons diagnosed with a condition who die of that condition within a specific time period. For example, if there are 30 deaths among 600 patients diagnosed with disease X then the CFR for the disease is 5% [= 30 (deaths)/600 (cases) * 100)].  The reliability of the CFR as a measure of the severity of disease depends on whether one has a good estimate of the total number of patients with the disease; if one underestimates the cases the CFR will artificially high, if one overestimates the cases the CFR will be artificially low.  The issue of accurate estimation of cases is a particular problem with COVID-19 because of the relatively low rate of testing in the population and the likelihood that large numbers of cases with no or minimal symptoms do not seek medical attention.  Although estimates of the CFR of COVID-19 using currently available data are likely to significantly overestimate its severity, I thought it might be interesting to tabulate the CFR by country for those countries reporting at least 9500 cases on March 29, 2020. Further notes of the sources of the data used in the table and any assumptions made will be found in the Notes on Data paragraph below.

COVID-19 Case Fatality Rates (CFR)

The table shows that there appears to be a wide variation in the severity of the disease in different countries. The adjusted CFR is likely the best estimate of disease severity because the baseline CFR has been adjusted for the rate of testing in the population of eacj country.  Comparatively, COVID-19 infection seems least severe in Germany with an adjusted CFR of 0.79% compared to the worldwide average CFR of 4.68%.  Belgium and the United States also achieve adjusted CFRs of approximately 1% or less. Similarly, the UK, France, and South Korea show less severe infection with an adjusted CFR of approximately 2% or less. The clear outliers at the other end of the severity spectrum are Italy and Spain with adjusted CFRs of 13.38% and 11.03% respectively.  There is no easy explanation for the comparative differences in severity across countries. Total health care expenditure does not seems to be related: the USA and Germany rank number 1st and 5th, respectively, in terms of health care dollars per person in 2016, however Switzerland (2nd) and the Netherlands (8th) do not perform nearly as well as France (14th), the U.K. (17th), or South Korea (24th).  The organization of the individual national health care systems is also an unlikely explanation as those in the table represent the full spectrum from primarily market driven organization in the United States, to a a mixture of public and private insurance in Germany to a universal national health service in the U.K. Population density if also unlikely to be a significant variable as it is much higher in countries with less severe infection such as Belgium (22nd most densely populated), the U.K. (32nd) and Germany (41st) than in Italy (51st) and Spain (89th).  It is possible that the apparent high mortality in Spain and Italy is because the virus in these countries has been disseminated across a great percentage of the population and the prevalence of the virus has not be adequately reflected in the limited testing that has been performed.

Notes on the Data: The principal data in the table (Cases, Cases/Million, and Deaths) were taken from the Coronavirus (COVID-19) Map published on Google (google.com/covid-19-map) on March 29th. These data were filtered to include only those countries that had recorded more than 9500 cases by March 29th.  The Tests/Million data were taken from Wikipedia (en.wikipedia.org/wiki/COVID-19_testing) also on May 29th.  The CFR was = number of Deaths/number of Cases * 100.  I adjusted  the CFR for the density of testing in the population (Tests/Million) normalizing to the testing density in Germany as follows: Adjusted CFR = CFR /[(Tests-Million Germany/Tests-Million Country X)]. No adjustments were made for Worldwide, Iran or China because of unavailable or unreliable data.