BookmarkSubscribeRSS Feed
SASPhile
Quartz | Level 8


 proc sql;
    create table want as
      select * from  have
   where  prxmatch("/1|Y|yes/i",Results_In_Death)
   union
        select * from  have
        where prxmatch("/fatal/i",OutCome)
  union
  select * from  have
  where  prxmatch("/death|dead|fatal|(subject|patient|participant).*(expire)/i",Term_Verbatim)
  union
       select * from  have
       where prxmatch("/na/i",Death_Date)=0;
 quit;
 
 
 the idea is to subset the dataset have that meet the where condition.but the results are not expected. for example results_in_death is showing values like 0,na etc.
    also outcome is showing values like disease progression etc
 how to subset the data just based on where conditions?

3 REPLIES 3
ChrisNZ
Tourmaline | Level 20

Don't you mean


proc sql;
    create table want as
      select * from  have
   where prxmatch("/1|Y|yes/i",Results_In_Death)
     and prxmatch("/fatal/i",OutCome)
     and prxmatch("/death|dead|fatal|(subject|patient|participant).*(expire)/i",Term_Verbatim)
     and prxmatch("/na/i",Death_Date)=0;
 quit;

Note that the parentheses in the third test seem oddly placed.

 

SASPhile
Quartz | Level 8

it is OR not AND

ChrisNZ
Tourmaline | Level 20

If you use OR then you will get dates not equal to NA or results_in_death not equal to yes. Why are you surprised?

Catch up on SAS Innovate 2026

Nearly 200 sessions are now available on demand with the SAS Innovate Digital Pass.

Explore Now →
How to Concatenate Values

Learn how use the CAT functions in SAS to join values from multiple variables into a single value.

Find more tutorials on the SAS Users YouTube channel.

SAS Training: Just a Click Away

 Ready to level-up your skills? Choose your own adventure.

Browse our catalog!

Discussion stats
  • 3 replies
  • 2222 views
  • 0 likes
  • 2 in conversation