Issue with interpreting Patient Mix

[gsharma007]

Understanding:

1. The good patient will have less variability around mean and will always be around target yield i.e. responding well to the therapy.
2. The average patient will have slightly more variability than the good patient and responding decently to the therapy.
3. The bad patient will have a lot of variabilities and is likely to be away from the Ybar.

Possibilities:
a) We assign time from the yield curve and call it Ybar. Then based on the type of patient and defined variability, calculate the final yield value around Ybar.
b) We assign time from the yield curve and call is Y_measured because it may not be Ybar. Then based on the type of patient and defined variability, calculate the final yield value around Y_measured.

Note: Only thing is to make sure is, anyhow we should have Y_measured more than the Ybar in 90% of the cases.