How to flag used observations in regression model?

[sjmoussa]

When I perform a logistic regression model using about 50 predictor variables my sample size decreases from n=18500 to n=7468 due to missing data on some of the variables. I want to create a new dataset that only includes the complete cases (n=7468). Is there an easy way to do this in SAS?

[Peter_C]

if these predictor variables are
1 : all numeric
and
2 : listed in macro variable &predictors
data new_dataset ;
set old_dataset ;
if nmiss( of &predictors ) > 0 then delete ;
run ;

[Doc_Duke]

You could output a dataset that includes the predicted risk and then subset on that value being non-missing.