Oscillatory Dynamics in Infectivity and Death Rates of COVID-19

Abstract

The analysis of systematically collected data for coronavirus disease 2019 (COVID-19) infectivity and death rates has revealed, in many countries around the world, a typical oscillatory pattern with a 7-day (circaseptan) period. Additionally, in some countries, 3.5-day (hemicircaseptan) and 14-day periodicities have also been observed. Interestingly, the 7-day infectivity and death rate oscillations are almost in phase, showing local maxima on Thursdays/Fridays and local minima on Sundays/Mondays. These observations are in stark contrast to a known pattern correlating the death rate with the reduced medical staff in hospitals on the weekends. While we cannot exclude the possibility that a significant portion of the observed oscillations is associated with the reporting of the individual cases, other reasons might contribute at least partly to these data. One possible hypothesis addressing these observations is that they reflect gradually increasing stress with the progressing week, which can trigger the higher death rates on Thursdays/Fridays. Moreover, assuming the weekends provide the likely time for new infections, the maximum number of new cases might fall, again, on Thursdays/Fridays. These observations deserve further study to provide a better understanding of COVID-19 dynamics. IMPORTANCE The infectivity and death rates for COVID-19 have been observed in many countries around the world as well as in the collective data of the whole world. These oscillations show distinct circaseptan periodicity, which could be associated with numerous biological reasons as well as with improper reporting of the data collected. Since very different results are observed in different countries and even continents, such as Sweden (very significant oscillations) or India (almost no oscillations), these data provide a very important message about different conditions under which the disease is spread or is reported, which, in turn, could serve as guidance tools in future epidemics. It is necessary that follow-up studies track the observed differences and fully reliably address their origins.

Impact of vaccination on the COVID-19 pandemic in U.S. states

“…Figure 1 shows a timeline of COVID-19 developments during this period, including important events and vaccination timeline. We aggregated the data to a weekly level in our baseline estimation given the observed weekly cycle 17 , 18 (see Extended Data Table 3 for results using daily data). The dependent variables used to assess the impact of vaccination on the pandemic are the growth rates of total cases and hospitalizations.…”

Predicting the Spread of SARS-CoV-2 in Italian Regions: The Calabria Case Study, February 2020–March 2022

“…We call “seasonality” the 7-days (circaseptan) observed oscillations because it has been treated as a seasonal behavior in the SARIMA model, even if, usually in epidemiology, the seasonality refers to longer periods (such as months). In some countries, additional hemicircaseptan (3.5 days) and 14-days periodicities have been observed [ 23 ]. The circaseptan seasonality of COVID-19 new daily cases, which has not been observed or reported in prior epidemics, is believed to be likely associated with epidemiological and social factors, mainly testing bias and reporting bias [ 23 , 24 , 25 ].…”

“…In some countries, additional hemicircaseptan (3.5 days) and 14-days periodicities have been observed [ 23 ]. The circaseptan seasonality of COVID-19 new daily cases, which has not been observed or reported in prior epidemics, is believed to be likely associated with epidemiological and social factors, mainly testing bias and reporting bias [ 23 , 24 , 25 ].…”

Impact of Vaccination on the COVID-19 Pandemic: Evidence from U.S. States

“…1 shows a timeline of COVID-19 developments during this period, including important events and vaccination timeline. We aggregated the data to a weekly level in our baseline estimation given the observed weekly cycle 17,18 (see Extended Data Table 3 for results using daily data). The dependent variables used to assess the impact of vaccination on the pandemic are the growth rates of total cases and hospitalizations.…”