Trial Data Are Dirty

Program 1.3 Subsetting a Data Set for Patients with an Adverse Event 

consider what would happen if the SAS data set aes looks like this  

The SAS code you wrote would eliminate the observation for subjectid=102. This is because the aeyn field is not populated for that row and is therefore eliminated by the WHERE clause in SAS. This is a classic parent-child data problem in clinical trial data, where the parent question is left unanswered but the child response is given. A way to handle this problem would be either to include the aetext field in the WHERE clause or to add a warning to the SAS log. The code in Program 1.4 does both. 

Anywhere you have conditional logic is another place for defensive programming techniques. When there is conditional logic, there should be a catch-all follow-up statement. Assume you have SAS code such as the following. 

---