Forum Discussion

themistoklis's avatar
themistoklis
Icon for Community Champion rankCommunity Champion
6 years ago

Calculate Reproduction Number (R0) for COVID-19 per country

Hello All,

 

Im trying to calculate the R0 index for Covid-19 cases using data from Johns Hopkins University but im facing a few difficulties. The link to the datasets is this.

 

Has anyone calculated the R0 index on PowerBI or has a sample of the formulas that can be used for the calculation?

 

https://en.wikipedia.org/wiki/Basic_reproduction_number

 

Regards

Themis

 

 

General Comment
Need Help
Show and Tell

3 Replies

  • Greg_Deckler's avatar
    Greg_Deckler
    Icon for Community Champion rankCommunity Champion
    6 years ago

    themistoklis R0 is calculated as:

     

    [Average of infection producing contacts per unit time] * [Average infectious period]

     

    I don't see how to get either of those from the data that you pointed to unless I am missing it. Which means you will need to use an estimation method and you will need to make some assumptions. If I follow everything correctly, you will first need to calculate K

     

    K = ln(2) / Td - You will need to make an assumption about Td. 

    For Td - https://en.wikipedia.org/wiki/Doubling_time

     

    You can then calculate R0 = e^(K * [Average infectious period])

     

    You will have to make an assumption about the average infectious period or get it from empirical data that others have created through testing. This is the simple model. 

     

    So, a formula for R0 would look like the following:

     

    R0 = 
  VAR __Td = 5 // some number
  VAR __K = LN(2)/ __Td
  VAR __AverageInfectiousPeriod = 10 //some number
RETURN
  EXP(__K * __AverageInfectiousPeriod)

     

    • themistoklis's avatar
      themistoklis
      Icon for Community Champion rankCommunity Champion
      6 years ago

      Hello Greg_Deckler ,

       

      Thank you for the quick reply and for the information you sent to me.

      Actually, i havent provided enough information about the "Reproduction Number" and the link from wikipedia was a bit misleading.

       

      For the calculation of this number there are various models. I am implementing the one called "SIR". A few links that would be helpful are the following:

      https://nextjournal.com/berwa/sir-model-for-covid-19-estimating-R_o

      https://onlinelibrary.wiley.com/doi/10.1111/anae.15151

      https://web.stanford.edu/~jhj1/teachingdocs/Jones-on-R0.pdf

       

      The SIR model classifies the population into 3 groups:

      S(t) (susceptible group), I(t) (infected group), and R(t) (Recovered)

       

      For susceptible group I make the assumption that "At the outset of an epidemic, nearly everyone (except the index case) is susceptible".

      For infected I make the assumption that they include both "recovered" and "deceased" individuals.

       

      These groups do exist in the datasets and im working on the formula so as to calculate this index. 

      Creating the formula using DAX is where i actually have difficulties.

       

      Regards

      Themis

       

       

      • Greg_Deckler's avatar
        Greg_Deckler
        Icon for Community Champion rankCommunity Champion
        6 years ago

        I'll have to take a look at the links you sent and then reassess the data in that context.