I came across calculation of probability for a decision tree model - which I do not understand. As I plan to do CEA of some health interventions I would not like to mess it up.
The used method (calculation) seem to me rather strange. Could it be mistake?
.......
The text states: "in routine practice the rate of patients found to have early stage recurrence in one year is 43 out of 100 patients."
Then they calculate the probability of early detection: "P = 1 – exp(–0.43*1) = 0.35".
.......
Why the probability ratio 0.43 (43/100) is not used? The probability of 0.35 seem to me more as if it was probability that it will not take longer than 1 year to early detect 43 patients out of 100. And it makes no sense to me for a decision tree.
If I am wrong - could anyone please explain what exactly these two numbers (0.43 and 0.35) represent? I would like to figure out why it is necessary to make such an adjustments to calculate the probability (as I do not see the difference among "probability" in case of incidence proportion data and "probability" in case of survival rate - in which case I did not notice such adjustments).
Thank you.
