How to estimate the infection rate of Covid 19 using confirmed cases and dates data

[Gunjika]

Dear SASers,

I urgently need your help... I have the following data:

data covid.covidppr;
input numdays confirm_cumul;
datalines;
1 1
2 3
3 3
4 3
5 3
6 34
7 82
8 414
9 1401
10 4989
11 13435
12 29800
13 59251
;

How can I use this information to compute the infection rate of Covid 19? The data is non-linear and fits polynomial and logarithmic regression trendlines very well. If infection rate can be estimated through regression curve fitting, please guide me. Can someone please let me understand the computation and the SAS code to do so? Thanks in advance.

[Reeza]

Whats the definition of an infection rate?

[PGStats]

Do you mean Infection Rates as defined in this paper?

https://healthcentricadvisors.org/wp-content/uploads/2017/03/Cal_Inf_Rates.pdf 

[Gunjika]

Dear PG,

Thanx for your reply. The thing is that the "infection rate" mentioned in this paper is "Incidence Rate".

[Gunjika]

Dear Reeza and PG,

I want something like what they have done in this paper, under section 3.3 "Estimation of infection rate".

Dear PG, in the context of Covid-19, the term "infection rate" is used interchangeably with the term "transmission rate".

[Reeza]

And you can see from either of these definitions you need population numbers or at risk numbers and we don't have that information.

[Ksharp]

@Rick_SAS wrote a blog about Covid 19  culumutive freq.

It is like Logistic Curve , could try  e^x/(1+e^x)   curve and PROC NLIN .