ABSTRACT
Analysing infection and mortality data for COVID-19 as a function of days for 54 countries across all continents, we show that there is a simple scaling behaviour connecting these two quantities for any given nation when the data is segmented over few ranges of dates covering the most rapid spread of the pandemic and the recovery, wherever achieved. This scaling is described by two parameters, one representing a shift along the time axis and the other is a normalisation factor, providing a reliable definition of the mortality rate for each country in a given period. The number of segments for any country required in our analyses turns out to be surprisingly few with as many as 16 out of 54 countries being described by a single segment and no country requiring more than three segments. Estimates of the shift and mortality for these 54 countries in different periods show large spreads ranging over 0-16 days and 0.45-19.96%, respectively. Shift and mortality are found to be inversely correlated. Analyses of number of tests carried out for detecting COVID-19 and the number of infections detected due to such tests suggest that an effective way to increase the shift, and therefore, decrease mortality, is to increase number of tests per infection detected. This points to the need of a dynamic management of testing that should accelerate with the rise of the pandemic; it also suggests a basis for adjusting variations in the testing patterns in different geographical locations within a given country.
