KendrickOrg / kendrick

Domain-Specific Modeling for Epidemiology
MIT License
49 stars 14 forks source link

Being able to do incidence plot of Kendrick simulation #124

Open SergeStinckwich opened 4 years ago

SergeStinckwich commented 4 years ago

At the moment, we can only do prevalence plot when doing a simulation (how much number of a condition at a given time). How we can have also incidence plot (rate of occurences of new cases of a disease or condition). What are the consequences on storage of results of simulation?

Some ideas can be taken from here: https://f1000researchdata.s3.amazonaws.com/manuscripts/19688/fdd786b1-10af-4bab-811b-8a34ca38979b_18002_-_zhian_kamvar.pdf R incidence package: https://www.repidemicsconsortium.org/incidence/

YvanGuifo commented 4 years ago

This link seems very explicit when formulating the incidence and gives details on the differences between this notion and the prevalence https://www.cdc.gov/csels/dsepd/ss1978/lesson3/section2.html#:~:text=The%20denominator%20is%20the%20sum,is%20at%20risk%20of%20disease.

YvanGuifo commented 4 years ago

Since we generally do not know the number of new cases, I suggest that it would be favorable to consider that the incidence is more or less constant during the duration of the disease. So according to wikipedia https://en.wikipedia.org/wiki/Incidence_(epidemiology) Prevalence is approximately the product of the incidence of illness and the mean duration of illness incidence = prévalence/durée