Read on Twitter
1/9 I've been receiving some great questions re how the numbers in our preprint can be interpreted. Hopefully this thread helps clarify. https://www.medrxiv.org/content/10.1101/2021.01.21.21250266v1
2/9 Backing up a little bit, this figure shows weekly deaths OF ANY CAUSE in California. Gray lines show 2016, 2017, 2018, and 2019. Red line shows 2020. The blue line shows the projected ("expected") number of deaths in CA had the pandemic not occurred.
3/9 Notice that there are much more deaths in 2020 than in prior years (red lines versus gray lines). There are also much more deaths in 2020 than compared to what we would have expected to see had the pandemic not occurred (red line versus blue line).
4/9 This is Table 1 from our preprint. The numbers on the left represent DIFFERENCES between observed and expected. Visually, you can imagine vertical distance between the red line and blue line.
5/9 You can describe numbers in that column by saying eg, that there were roughly 10k more deaths among working-age Californians than expected. Or, eg, that among essential sectors, facilities workers had the largest ABSOLUTE INCREASE in mortality (compared to expected).
6/9 Here is the table again. Now let's focus on the numbers on the right. These represent the RATIO of the observed numbers to the expected numbers. We're still comparing red line to blue line, but dividing rather than subtracting.
7/9 You can describe numbers in that column by saying, for example, that among food/ag, mortality was 39% higher than expected. Or that food/ag had the largest RELATIVE INCREASE in mortality.
8/9 As we say to our students, there's no such thing as a stupid question. Please feel free to ask here, or email me privately, if you want further clarification. Thank you all again for your interest.
Also: here is Table 3. Just a clarification here that the deaths column here does NOT have the same interpretation as the first column of numbers in Table 1. This deaths column in Table 3 is mostly just provided to give a rough sense of how big the occupations are, etc.
