Longitudinal administrative health data - flag claimants who received the same treatment accross all observations

[Robert Nesdole]

Greetings,

I have a longitudinal administrative health dataset consisting of over 400,000 observations. Patients are represented by an alpha numeric identification number and have between 1 and 7000+ administrative claims. I would like to identify all patient identification numbers that received the same treatment across all observations. I have generated a variable that identifies the treatment of interest as 1 for each observation, but I have not been able to figure out how to identify Patients who received the treatment across all observation.

Here is an example dataset to help illustrate my problem:

ID TRT

a1 1

a1 0

a1 1

a2 1

a3 1

a3 1

a4 1

a4 1

a4 1

a5 0

a5 0

In the example provided ID# a2-a4 received the same treatment.

Thanks in advance.

Robert

[Sarah Edgington]

If I'm reading your question right the idea is that you want to identify individuals who have all 1s for your TRT variable. You can use by to look within your ID groups to determine the total observations with the treatment and compare this to total number of observations. Try something like this:

bysort ID: egen tot_trt=sum(TRT)
bysort ID: gen all_same=(tot_trt==_N)

[Nick Cox]

See also https://www.stata.com/support/faqs/d...ions-in-group/ from which

bysort ID (TRT) : gen allone = TRT[1] & TRT[_N]