Comparing proportions in a subgroup to proportions in the overall population

#1
I am trying to compare the proportion of patients in my subgroup with the outcome of interest to the proportion of patients in the overall population with that outcome, and test for significant difference in order to assess which subgroup variables I should be including in my regression analysis.

As an example, my outcome is the proportion of patients with cerebral hemorrhages, and one subgroup would be patients <15 years old. I'd like to compare the proportion of those <15yo with cerebral hemorrhage to the proportion of ALL patients with cerebral hemorrhage. I can't find a way to do this in STATA13 using prtest, 2-way by groups or 2 way by variables as they all split the total population.

Any help would be much appreciated!
 

hlsmith

Not a robit
#4
Perhaps you can create a new dataset with the subset concatenated with the overall dataset, then create a categorical variable (set = 1 if subset; set = 2 if overall set), then just run your stats based on categorical variable (1 or 2).