Hello everyone, I have a question about how to get categorical variables RR estimates in PROC GENMOD.   My outcome is a binomial variable (death), which 1= event happened, 0=event didn't happen. I listed 3 variables (x, y, z) as exposures that I am looking for to get their RR estimate in the PROC GENMOD procedure. All 3 variables are categorical, x has 5 categories, y has 2 categories, z has 5 categories. There is no missing value among those 4 variables.   I found a PROC GENMOD tutorial from this link: https://stats.idre.ucla.edu/sas/faq/how-can-i-estimate-relative-risk-in-sas-using-proc-genmod-for-common-outcomes-in-cohort-studies/   The code I used is: proc genmod data = test; class x (ref="1")/param=ref; class y (ref="1")/param=ref; class z (ref="1")/param=ref; model death (event="1") = x y z/ dist=binomial link=log; estimate 'Beta_x' x 1 -1/exp; estimate 'Beta_y' y 1 -1/exp; estimate 'Beta_z' z 1 -1/exp; run;   With the above code, I have this output. Although you can see the estimates don't show for different categories of x and z.   I searched for other solutions. Then I found this solution: https://communities.sas.com/t5/Statistical-Procedures/RR-estimates-using-proc-Genmod-for-categorical-variables/m-p/549129#M27406 link to this website: https://support.sas.com/kb/23/003.html   So I tried to use LSMEANS, the below code is I modified to get categorical variables RR estimates. proc genmod data = test; class x (ref="1")/param=ref; class y (ref="1")/param=ref; class z (ref="1")/param=ref; model death (event="1") = x y z/ dist=binomial link=log; lsmeans x z / diff exp cl; /*I only added this line comparing to the code above*/ estimate 'Beta_x' x 1 -1/exp; estimate 'Beta_y' y 1 -1/exp; estimate 'Beta_z' z 1 -1/exp; run; Although nothing new results showed up with this code, SAS generated the same results as the first screenshot I attached. Also, with the following Note and Warning: I am not sure if I used LSMEANS in the right way. Also, I wanted to ask what "1 -1" stands for in "estimate 'Beta_x 1 - 1/ exp"? I have seen some tutorial shows "1 -1" or "1" or "0 1 -1", but I don't really understand the function of it. If there are some tutorials I should know before doing this code, I would love to know about them!   I appreciate your help! ... View more

Hello everyone, I have a question about how to find/flag the second-to-last observation by the same ID.    The dataset I have like this: ID Assessment date Status  Status_new1 Status_new2 1 2012-06-19 No Yes Yes 1 2012-06-20 Yes Yes Yes 1 2012-06-21 No Yes Yes 2 2012-06-22 . No No 2 2012-06-23 No No No 2 2012-06-24 . No No 2 2012-06-25 No No No 3 2012-06-19 Unknown Yes Yes 3 2012-06-20 Yes Yes Yes 3 2012-06-21 No Yes Yes 3 2012-06-22 . Yes Yes 3 2012-06-23 No Yes Yes 4 2012-06-01 Unknown No Unknown 4 2012-06-03 . No Unknown 4 2012-06-05 No No Unknown 5 2012-06-08 No No Unknown   The observations that I would like to find/flag is where I used red color: ID Assessment date Status  Status_new1 Status_new2 count 1 2012-06-19 No Yes Yes 1 1 2012-06-20 Yes Yes Yes 2 1 2012-06-21 No Yes Yes 3 2 2012-06-22 . No No 1 2 2012-06-23 No No No 2 2 2012-06-24 . No No 3 2 2012-06-25 No No No 4 3 2012-06-19 Unknown Yes Yes 1 3 2012-06-20 Yes Yes Yes 2 3 2012-06-21 No Yes Yes 3 3 2012-06-22 . Yes Yes 4 3 2012-06-23 No Yes Yes 5 4 2012-06-01 Unknown No Unknown 1 4 2012-06-03 . No Unknown 2 4 2012-06-05 No No Unknown 3 5 2012-06-08 No No Unknown 1   I am able to create a new variable, named "count" to get a serial number by ID. But I don't know how to find the second to last observation, because the obs are not fixed. Some IDs could only have 1 ob, some IDs could 3 to 10 obs. If there is a way that can create a new variable to flag every second-to-last observation by ID in the same dataset, that would be great!   Thank you for your help! ... View more

Hello everyone, I have a question about how to define status in multiple obs by the same ID.    The dataset I have like this: ID Assessment date Status 1 2012-06-19 No 1 2012-06-20 Yes 1 2012-06-21 No 2 2012-06-22 . 2 2012-06-23 No 2 2012-06-24 . 2 2012-06-25 No 3 2012-06-19 Unknown 3 2012-06-20 Yes 3 2012-06-21 No 3 2012-06-22 . 3 2012-06-23 No 4 2012-06-01 Unknown 4 2012-06-03 . 4 2012-06-05 No   The dataset I would like to have is like this: ID Assessment date Status  Status_new1 Status_new2 1 2012-06-19 No Yes Yes 1 2012-06-20 Yes Yes Yes 1 2012-06-21 No Yes Yes 2 2012-06-22 . No No 2 2012-06-23 No No No 2 2012-06-24 . No No 2 2012-06-25 No No No 3 2012-06-19 Unknown Yes Yes 3 2012-06-20 Yes Yes Yes 3 2012-06-21 No Yes Yes 3 2012-06-22 . Yes Yes 3 2012-06-23 No Yes Yes 4 2012-06-01 Unknown No Unknown 4 2012-06-03 . No Unknown 4 2012-06-05 No No Unknown   If any obs in the same ID equal to "Yes", the new status variables (No matter is status_new1 or status_new2) should be equal to "Yes". The only difference between status_new1 and status_new2 is demonstrated in ID 4. Considering ID 4's status are not very clear, I want to create status_new2 to do sensitivity analysis to see if there is any difference when I recode them as "No" and "Unknown".    The tricky part to me is how to recode the new variables when the status included two different statuses in the same patient.    I used "where status = "Yes" " to output all the patients who had status equal to "Yes", and created their status_new1 as "Yes". But I don't think I could use the same way to do for status = "No". Because "Yes" and "No" could be mixed.    Thank you for your help! ... View more