Figure 1. Time series plots of daily confirmed COVID-19 cases (in black), COVID-19 deaths (in red), and influenza positive cases in 2015-2019 (in light blue) and in 2020 (in dark blue).
We obtained the daily confirmations of COVID-19 cases and deaths for 216 countries or regions, during February 2020 - August 2020, and weekly influenza confirmations during 2015 to current week, from World Health Organization [1,2]. We showed data of four countries (two from Europe and two from Asia which caught a lot of attention in the early phase of the pandemic) in Figure 1. Research of other 49 countries or regions are listed in the supplementary. The time series (in common log scale on vertical axis) of confirmed cases showed two waves. Based on this observation, we divided the transmission of the disease into two phases: Phase I before June 1 and Phase II after June 1 for confirmed cases; and Phase I before June 10 and Phase II after June 10 for deaths. For deaths, we choose the truncated time ten days later to account for the delay between confirmation and deaths [3]. We compared the raw case fatality rate (CFR) of Phase I and Phase II for all countries or regions.
In our supplementary data in the appendix, for each country, we break down the data for Phase I or Phase II. The column “case_pre” is the total of confirmed cases before June 1 and the column “case_post” is the total of confirmed cases after June 1 up to July 26. Accordingly, the column of “death_pre” is the infection death before June 10 and the column “death_post” is the infection death after June 10 up to August 6. We define the raw case fatality rates (CFR) as\(r_{1}\)=\(\frac{death\_pre}{case\_pre}\) and\(r_{2}\)=\(\frac{death\_post}{case\_post}\). Then the change in CFR is reduction= \(\frac{r_{1}-r_{2}}{r_{1}}\). Based on our analysis, among all 53 hardest-hit countries or regions (supplementary Table 1), 43 of them had an apparent reduction in CFR. Only ten remaining countries or regions had an increase in CFR (negative reduction). The decrease in the CFR might indicate the severity of the global pandemic is becoming better. The potential reason for such decrease is worth further investigation. We propose the following hypotheses that could contribute to the decrease of CFR in the second phase. First, the apparent higher CFR in the first phase could be a harvest effect, namely a large number of elderly and individual with health conditions (the group at risk) likely died in the first phase, especially in these countries with a high infection rate, and this risk group run low in the second phase. If a country or region (such as Hong Kong) was spared from the first phase, of course it is not surprising to see an increase in CFR. Second, the age structure of infected changed due to a variety of reasons, e.g. social movement in many countries might involve more healthier young individuals. Third, the virus might evolve such that young health adults become more susceptible, thus lead to a reduced CFR. Forth, favourable climate might lead to reduced CFR (e.g. warmer weather in north hemisphere and improved air quality due to city lockdown [4-6]). Last but not least, improved timely treatment and enhanced massive testing could reduce the deaths and increase the number of cases, thus a reduced CFR in the second phase.
We show time series plots of eight countries in supplementary figure, where only Iran faces an increase in the raw CFR and a table summarizes results of 53 countries or regions. The weekly influenza laboratory confirmations for the previous five years may be used as a proxy of the weather, since it is well known that influenza seasonality is driven by temperature and humidity. Thus we may wonder whether favourable weather may contribute to a reduce CFR for COVID-19. The sharp drop in influenza cases in 2020 (dark bold curve), compared to previous years, may be due to social distancing and possible interference with COVID-10 infection. Thus it is informative to compare the COVID-19 and influenza in these plots. Individual data or age grouped data are needed to further clarify the phenomenon. The finding is nevertheless of significance to inform public and for policy making.