Suggest Detection Rate Metric

[anthonygtellez]

Professor Gabriel Leung who lead the SARS pandemic response in Hong Kong, recently estimated that the death rate is a good indicator of the overall true infection rate.
Source: https://www.youtube.com/watch?v=Y7nZ4mw4mXw

The true infection count could be calculated as:

• lower threshold = (number of deaths) * 80
• upper threshold = (number of deaths * 100

Using this count we can calculate the delta (likely asymptomatic carriers) using the upper/lower thresholds and the currently detected cases to determine if we are detecting (testing) enough people.

• Asymptomatic Delta = ((upper threshold - detected cases) + (lower threshold - detected cases)) / 2

[ryanwoconnor]

Great stuff here Anthony. Can you find me a paper on this or something in writing as well for people to read up on? Will look into incorporating the SPL into next dashboard revision.

[anthonygtellez]

Looking for a paper this is the closest one I've found around some of the different modeling techniques they have used. https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(20)30260-9/fulltext
The death rate seems to be something of a more recent way to try and discern the true number of infections. If I can find something in writing I'll share.