We will close the breakout at 12:25pm EST

I have not gotten the break out room

oh, ok, thanks

I have not gotten yet, I am in 3

I left the group. I will try to call IT and log back in

its' a lag thing when all the cameras are on for me

I'm now surprised that the field used to think that R0 was (just) pathogen-specific.

I missed the beginning so I am not sure if my group (3) mentioned the lack of impact of the change on the vaccination rate. (shiny app). sorry if that was already discussed

Yes Oumar, I also mentioned that :)

Can you provide the citation?

Excellent! Thank you1

Thx Alessandro

I really appreciate the ebook (An Introduction to Infectious Disease Modelling) you shared. It is simple to grasp and will be extremely useful.

Of course Tafadzwa! Glad it's been helpful :)

@sean -- Module 14 of SISMID goes in depth on the application of fitting methods that Micaela presented at the end

I was jumping on and off and prob missed this, but in case it was not discussed:, when choosing to integrate death on model. do you chose crude death rate or choose a set of disease/age specific that are similar to your the disease you want to study

Module 6 also goes into more depth on model fitting

OK. thx Matthew

@Oumar this is an interesting question. And is really one of complexity. The simple non-age structured SIR models that we discussed on the first day would take just the crude death rate (I believe I used the parameter delta in my slides)

This gives a very oversimplified implicit age distribution and there are some corresponding biases in the dynamics because of it

If you make a full age structured model then you can incorporate explicit age-specific mortality rates

Which would be the preferred approach assuming that you have the data and the computational ability

I see, thank you!

Note that increasing the complexity to an age structured model to incorporate realistic mortality then means you need to represent age-structured mixing ... which means you need to either know or assume what the contact matrix is. So as you increase complexity (to account for mortality rates that you might know) you end up needing to add more parameters (the contact matrix) that are perhaps hard to know.

This is some people choose to make simpler models -- because there fewer complicated assumptions to make. Thus, while a simple model is unrealistic, it is less dependent on complicated assumptions and easier to interpret.

It's a long philosophical discussion about how complicated to make models ... one that is always fun to have, and more fun to have over a meal.