I want to measure the 'cost' of diagnoses (average, standard deviation, etc). However, several diagnoses are on one claim. If I go to the claim line level, there are still several diagnoses on each line. Since the claim isn't broken down by each diagnosis code that is present, what is the best way for me to try to isolate the costs of diagnoses?
I starting thinking about some regression with binary indicator variables for diagnoses and then realized I should probably ask before I go down that road. I'm sure someone has dealt with this problem before.....
I starting thinking about some regression with binary indicator variables for diagnoses and then realized I should probably ask before I go down that road. I'm sure someone has dealt with this problem before.....
Cost of individual diagnosis?
