Headlines like “Worldwide death toll from the coronavirus passed 100,000” or “COVID-19 could claim [Canadian] 22,000 lives” are newsworthy, but leave important questions unanswered, like:
Graphs of per capita mortality over time answer these questions, allow meaningful comparisons between countries at different stages of an infection, and provide a gauge with which to compare other risks.
The percentage of a population tested varies significantly from one country, province or state to another, skewing statistics. The accuracy of tests also vary. Death statistics lag confirmed case statistics by weeks and are also distorted by different classification methods, but less so, making them the more reliable indicators.
Pundits of epidemiology say, “humans don’t understand exponential growth“, but this isn’t entirely true. Mathematicians and engineers use logarithms to understand exponential characteristics. The above graphs have a vertical logarithmic scale. Each horizontal line is a factor of ten from the line above or below it, rendering exponential data as a straight line. The slope of the line varies with the contagiousness of the disease and roughly correlates with what epidemiologists call the R0 (R naught), the average number of infections caused by each of the infected.
The 1918 Spanish Flu is estimated to have killed 1% – 6% of world population, or 10,000 to 60,000 deaths per million people — literally and completely off the scale of the above graphs. Barring the unforeseen, like viral mutation, nothing in the data suggests COVID-19 will be that bad.
For a more personal perspective, check your mortality with Statistics Canada’s Mortality Probabilities. For example, in Canada, a male 65-year-old has a 1.137% chance of dying (or 11,370 per million), and a female 65-year-old has a 0.731% (or 7,310 per million). Dying of COVID-19 shouldn’t be a 65-year-old’s biggest fear.