Hello everyone,
I have a dataset consisting of two groups of patients (control and intervention) with different diagnoses that have experienced traumatic events graded with 1,2, etc., on certain dates (I transformed date variables into genuine Stata date variables) and received care for these events (emergency, hospital, outpatient). I am trying to calculate:
- number and type (by grading) of trauma events per patient by diagnosis and group.
- mean number of healthcare and emergency visits per patient by diagnosis and group.
- mean costs per patient by diagnosis and group.
Below is an example of my dataset.

id diagnosis group trauma1 date_trauma1 trauma2 date_trauma2 healthcare cost
1 stroke control 1 01jan2000 1 02mar2000 emergency 100
1 stroke control 2 10jan2000 2 23mar2000 hospital 200
2 cancer control 1 01feb2000 1 03apr2000 outpatient 50
2 cancer control 1 20mar2000 2 04may2000 outpatient 50
3 accident intervention Normal 02mar2000 3 01sep2000 hospital 200
3 accident intervention 2 01jun2000 3 15sep2000 hospital 200
3 accident intervention 2 15aug2000 2 02oct2000 hospital 200
4 stroke intervention 2 01jul2000 2 04sep2000 emergency 100
4 stroke intervention Normal 20jul2000 NULL 10oct2000 outpatient 50
4 stroke intervention 2 03aug2000 1 11nov2000 hospital 200
id group trauma1 date_trauma1 trauma2 date_trauma2 healthcare cost
1 control 1 01jan2000 1 02mar2000 emergency 100
1 control 2 10jan2000 2 23mar2000 hospital 200
2 control 1 01feb2000 1 03apr2000 outpatient 50
2 control 1 20mar2000 2 04may2000 outpatient 50
3 intervetion Normal 02mar2000 3 01sep2000 hospital 200
3 intervention 2 01jun2000 3 15sep2000 hospital 200
3 intervention 2 15aug2000 2 02oct2000 hospital 200
4 intervention 2 01jul2000 2 04sep2000 emergency 100
4 intervention Normal 20jul2000 NULL 10oct2000 outpatient 50
4 intervention 2 03aug2000 1 11nov2000 hospital 200