Best Of

Here is the solution to your BLUF:
1. Use Group By tile in Magic ETL and the following formula: GROUP_CONCAT(DISTINCT CASE WHEN Treatment IN('Consult','Testing','Evals','Preop') THEN Name END)

You don't have a nurse identifier in your sample data so I added one called "Name"

Results:

This will show which nurses saw each patient. Let me know if this isn't what you're looking for.

For the patient count piece, you need to use COUNT(DISTINCT ) with a CASE statement:

COUNT(DISTINCT CASE WHEN Treatment IN('Consult','Testing','Evals','Preop') THEN Patient END)

For analyzing which nurses saw patients that led to surgery vs. no surgery, you could add a flag to every patient that led to surgery in Magic ETL as shown below:

Then create some kind of card in analyzer using this beastmode:
COUNT(DISTINCT CASE WHEN Had Surgery = 'true' THEN Patient END) / COUNT(DISTINCT Patient)

Result:

75% of patients John saw ended up in surgery, 67% of patients Jacob saw ended up in surgery.