Comparing SARS-CoV-2 with SARS-CoV and influenza pandemics

The Personal View by Eskild Petersen and colleagues¹ is a brilliant piece of comparative, historic-epidemiological research. At some important points, however, the contribution of Petersen and colleagues is too vague, although quantitative information is available. We would therefore like to supplement the work with this information.

Petersen and colleagues say the incubation period of the 1918 pandemic influenza is unknown. However, it is mentioned in several contemporary studies, such as the one by Nuzm and colleagues,² which says that, in 1918, the incubation period ranged from a few hours to 2 days.

Petersen and colleagues do not provide an exact estimate for the proportion of patients requiring hospitalisation during the Spanish flu. The statistical yearbooks of Switzerland from 1924, 1925, and 1929 allow estimation of the number of hospitalisations during the Spanish flu.³ As shown in the appendix, Switzerland had an average population of 3·9 million in the years 1917–25 and about 90 000 hospitalisations in non-pandemic times. In 1918, during the first two waves of the Spanish flu, the number of hospitalisations was around 121 000, about 30 000 more than usual. Therefore, by dividing the number of people infected with influenza in 1918 by the population size for that year, we estimate that about 17% of the Swiss population was infected with influenza in 1918. Moreover, by dividing the number of excess hospitalisations by the number infected, we estimate that about 4·5% of infected people were hospitalised.

As stated by Petersen and colleagues, the number of patients treated in an intensive care unit (ICU) during the Spanish flu is unknown because the first ICU worldwide opened in 1953 in Copenhagen, Denmark.⁴ However, people treated in hospitals during the Spanish flu should be regarded as ICU patients because there are many reports that hospitals were overcrowded. For example, in Berlin, Germany, people were admitted to hospital only if they had a fever of at least 41°C.⁵ If we again use the data from Switzerland (appendix), this means that about 30 000 of 665 000 infected people in 1918 required ICU treatment, which is about 4500 per 100 000 infected people and about 4500 times higher than in the 2009 influenza pandemic.¹

Furthermore, Petersen and colleagues speculated that, based on their historical perspective, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) might have a second wave. We have recently published a paper on subsequent waves in viral pandemics and found that more than 90% of epidemic events in influenza pandemics and about half of the epidemic events during the severe acute respiratory syndrome coronavirus pandemic had a second wave.⁶

It is unlikely that data from the Spanish flu in Switzerland will allow any predictions about the current SARS-CoV-2 pandemic. In Switzerland, SARS-CoV-2 resulted in 33 000 confirmed cases within 6 months. However, the Spanish flu resulted in 40 times more cases than SARS-CoV-2 during the same time interval, when adjusted for population sizes.