Logic node and trackers

I am modelling a microsim markov for breast cancer stages.  Once the breast cancer reaches a certain stage (local) they have treatment.

When patients are treated in the local state they can opt for : 1) breast conservative surgery, 2) BCS and radiotherapy or 3) mastectomy. From this state they then enter well with history of breast cancer. I can use a tracker for each arm to keep track of what each patient opted for I assume? (this is my first question).

Subsequently, in the well state they are then at risk of a recurrence - this is dependent on the treatment they had (i.e. probability of recurrence for mastectomy < BCS and RT< BCS). I want to use a logic node in the well state to determine the probability that each patient will develop a recurrence which will vary depending on the treatment they had. (my second question is how to do this or is this possible?)

I assume it would be something along the lines of:
1) using a tracker for each treatment (counting 1 if they had a BCS/ RT/Mast) in the local arm

2) using a logic node in the well state and changing the probability of a recurrence using the if command e.g. if(BCS=1;0.3) etc but I don't know the code to do this using three treatments e,g.( if BCS=1, if RT=1, if Mast=1; 0.2; 0.3; 0.4) and then # in no recurrence to make up the remaining probability.

Does this make sense? Would appreciate advice on the code. Cheers



1 comment
  • Official comment

    I believe you can do what you need without too many complications.

    With regard to recording the treatment received, I would recommend a single tracker with different values for each of the three treatment options. Perhaps something like this.

    t_treatment = 1 (for mastectomy)
    t_treatment = 2 (for BCS and radiation)
    t_treatment = 3 (for BCS alone)

    Then you can check a single tracker to know which treatment was applied.

    With regard to the recurrence, I don't think you want a Logic Node. I think a standard chance node with a probability for recurrence would work just fine with a formula variable for the probability of recurrence.

    Parameter Variables:
    pRecur_PostMastectomy = ?
    pRecur_PostBCSRadiation = ?
    pRecur_PostBCS = ?


    pRecurrence = Choose(t_treatment; pRecur_PostMastectomy; pRecur_PostBCSRadiation; pRecur_PostBCS)

    Then reference pRecurrence in the model. It will use the appropriate parameter based on the prior treatment.

Please sign in to leave a comment.

Didn't find what you were looking for?

New post