In longitudinal data collection you are more likely to have multiple data sets hopefully with a variable to contain the period identification.
data test1;
input id $ fd cl hs hl ;
year=2002;
datalines;
1 1 1 0 .
2 1 0 1 1
3 1 . 0 0
5 1 1 1 0
;
run;
data test2;
input id $ fd cl hs hl ;
year=2003;
datalines;
1 1 1 0 1
2 1 0 0 .
3 0 0 1 0
5 0 0 1 0
;
run;
data test3;
input id $ fd cl hs hl ;
year=2005;
datalines;
1 0 1 1 1
2 1 1 1 1
3 0 0 1 0
4 1 1 1 0
;
run;
(yes I intentionally don't have id 5 in the third set. It happens with longitudinal studies - if your data doesn't have that thank the stars)
which you combine with append or set:
data all;
set test1 test2 test3;
run;
Which lends itself (actually in the last data step) to creating one or more variables that have values based on fd cl hs hl. If you variable are 0/1 with 1 =yes then sum (fd,cl,hs,hl); gives the number of "yes" per respondent.
then summarize that data.
It isn't clear from your example if "each year" means the same client in every year, or if any client in the year had a response.
Depending on which what your combinations of each year and a client or any client or what ever, then summarize with Proc means/summary with year or id (or both) as class variables.
