@StatDave Wow. Thanks, I had no idea.   I have an outcome variable that is injuries per year. It's a count so I was going to use a negative binomial model...What SURVEY procedure should I used to determine if there is a linear relationship between year and injuries per year? Should I treat injuries per year as a continuous variable and use PROC SURVEYREG? Is there another approach?    Thanks!   ... View more

@PGStats I'm not exactly sure how to describe the change in the output by replacing "OR" w/ "+"...The value of rx_anti_dep = 12 for the observation "Effexor , Amytriptyline, Metformin, Valuim, Flomax,". It should equal 2. I'm not sure where the 12 comes from since there are more than 12 "+"s (See below).   rx_anti_dep = findw(rxsublist, "Prozac", " ", "pio") + findw(rxsublist, "prozac", " ", "pio") + findw(rxsublist, "Paxcil", " ", "pio") + findw(rxsublist, "trazadone", " ", "pio") + findw(rxsublist, "effexor", " ", "pio") + findw(rxsublist, "depression meds", " ", "pio") + findw(rxsublist, "anti-depressant med", " ", "pio") + findw(rxsublist, "anti-depressants", " ", "pio") + findw(rxsublist, "Welbutrin", " ", "pio") + findw(rxsublist, "welbutrin", " ", "pio") + findw(rxsublist, "wellbutrin", " ", "pio") + findw(rxsublist, "Wellbutrin", " ", "pio") + findw(rxsublist, "Celexa", " ", "pio") + findw(rxsublist, "celexa", " ", "pio") + findw(rxsublist, "Anti-depressants", " ", "pio") + findw(rxsublist, "antidepressants", " ", "pio") + findw(rxsublist, "Depression Pills", " ", "pio") + findw(rxsublist, "many anti depression meds", " ", "pio") + findw(rxsublist, "Lexapro", " ", "pio") + findw(rxsublist, "zoloft", " ", "pio") + findw(rxsublist, "Zoloft", " ", "pio") + findw(rxsublist, "Esciotalopram", " ", "pio") + findw(rxsublist, "venlafaxine", " ", "pio") + findw(rxsublist, "anti-depressants", " ", "pio") + findw(rxsublist, "amitriptyline", " ", "pio") + findw(rxsublist, "Amytriptyline", " ", "pio") + findw(rxsublist, "SSRIs", " ", "pio") + findw(rxsublist, "Venlafaxine", " ", "pio") + findw(rxsublist, "amyitriptaline", " ", "pio") + findw(rxsublist, "mirtazipine", " ", "pio") + findw(rxsublist, "Citalopram", " ", "pio"); ... View more

@PGStats Thanks, we are getting closer! One additional modification is needed however...and I should have made this clear at the onset. Sorry.   Some observations have values with multiple drugs in the same category. For example, "Effexor , Amytriptyline, Metformin, Valuim, Flomax,".    In my categorization scheme, Effexor and Amytriptyline are both anti-depressants. When I use the following code, a value of 1 is generated for rx_anti_dep. What I really want is a value of 1 to be generated for EACH drug, so that the value in this scenario would be 2, 1 for Effexor and 1 for Amytriptyline.   rx_anti_dep = findw(rxsublist, "Amytriptyline", " ", "pio") OR                        findw(rxsublist, "Effexor", " ", "pio");   Any suggestions? ... View more

@Shmuel Thanks for your help. Your approach was my original approach, but it seemed like the IN Operator was struggling when the value contained more than one string (i.e. Drug or quoted word/phrase). For example, if the value was "Zoloft, Xanax, Immitrex, Pain killers/Muscle Relaxers.", a value of 1 was generated for "Zoloft" (i.e. catg5, let's say), but a value of missing was generated for catg3 (i.e. Xanax) and catg6 (i.e. Immitrex), etc.. If the only value in the observation was "Zoloft" this wasn't a problem, but that is rarely the case in this dataset.   I just substituted your exact code and the results are similar, except...there are instances where it actually worked (See Observation 1767 in the attached). In the vast majority of observations however, only one value was generated for one new variable. In a small number of cases, the syntax seems to have worked correctly (i.e. Observation 1767). I don't understand why b/c there are no observable differences in the values and the syntax is the same. Hmmm.... ... View more

@PGStats That's what I thought...When I include this syntax in the DATA step, alll the values in the newvar_said equal zero, despite the value of the variable being equal to the quoted string (i.e. findw(rxsublist, "ibuprofen", ,"i")...What am I missing? Thanks!         ... View more

@PGStats  W/ your suggested syntax, what should the value of newvar_said if the expression is true?   Thanks for your help! ... View more