*** When will we know we are making progress against #COVID19? ***

When we understand the dynamics of the number of deaths. But there are many limitations to the data and number of deaths, as I will show.
1/x
The amount of testing is ramping up rapidly. We have massive testing in the US, even though the per capita rate can be much higher in relatively smaller countries. And the testing so far is only good for active infection, as we do not have widespread antibody testing.
2/x
There is concern some countries do not report deaths accurately & therefore we will never know. Also, death rates also relate to healthcare capacity, comorbidities, smoking & age of the population. See the age distribution of Spain, Italy, and the US.
https://www.bloomberg.com/news/articles/2020-04-01/china-concealed-extent-of-virus-outbreak-u-s-intelligence-says
4/x
IMHO understanding death rates (total, per day, & growth factor will be key) will be our best bet. Much better than the number of cases alone. Hospitalizations will be complicated since there might be a bias towards "more likely to hospitalize if COVID+" (space permitting).
5/x
If you look at raw numbers of deaths per day alone you cannot fully understand. You have to adjust for patient populations. Also, remember we really do not know how this is playing out in many countries and accuracy of death cause reporting.
6/x
And a line graph of raw numbers would look most alarming for the US. And indeed this is a very serious situation. But for planning and forecasting adjusting for the population at a minimum is required.
7/x
If you adjust death rates for the population of these countries the situation is somewhat different. (Shown as cases per million)
8/x
And again the line graph for data adjusted for the population.
9/x
*** But tracking the growth rate is important ***
A growth factor over 1 means the number of deaths still growing. Below I show growth rate per country; 1.2 or greater in red, 1-1.19 in yellow and 1 or less in green (good!).
10/x
It will be important to understand how this evolves allowing forecasting. I am afraid we will not know that accurately for the US (and less so in many other parts of the world). Retrospectively we will know better once antibody testing is here. https://twitter.com/Rfonsi1/status/1245348792678797312?s=20
11/x
Data and graphs for today.
12/x
#COVID19
You can follow @Rfonsi1.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: