I am trying to separate out a column as it currently has multiple effects listed in the individual column like so: I use the scan function and create different column names for the various symptoms. From there, I would like to arrange them alphabetically but have the missing values be last (i.e. have the Symptom1 column be acidosis for all those who have that symptom) so that I can use the proc freq function to determine how many people total had each individual symptom. I tried using call sortc but this had all the missing values first (1 of the clinical effects had 23 different symptoms, but many only had 1 or 2 listed) and then would start the symptoms in alphabetical order so that the "acidosis" was in different numbers instead of always being symptom1 like I would like it to be. Any suggestions?
... View more