A simple question, but I have three variables in my dataset: ID, ICDCode, and a Visit Date. There are multiple occurrences of each ICDCode per person(ID). How do I obtain a total, distinct count of icdcodes for the entire dataset, without counting a certain ICDCode twice for an individual? For example, I want to say that there are 100 cases of Heart Disease in my dataset (without counting heart disease 10 times for the same person).
Below is code I have tried:
proc freq data= cases;
table ICDCode;
run;
proc sql;
select ICDCode,
count(*) as Frequency
from cases
group by ID;
quit;
How about simply: (Given that 429.9 represent heart disease)
data cases;
input ID ICDCode;
datalines;
1 429.9
1 429.9
1 2
1 3
3 429.9
3 429.9
3 3
2 1
2 2
;
proc sql;
select count(distinct ID) as n
from cases
where ICDCode = 429.9;
run;