Hello everyone, I would like to use sql to get distinct observations for each patient that corresponds to their earliest fill date. I saw the attached code on the forum but it does not give the output as distinct observations. Thank you   /**what i have***/ patient ID    fill_date patient 1      02/23/2001 patient 1      02/25/2001 patient 1      03/14/2003 patient 2       01/01/2006 patient 2      10/14/2001 patient 2       08/25/2008 patient 3       07/14/2004   /**what i want**/ patient ID    fill_date patient 1      02/23/2001 patient 2      10/14/2001 patient 3       07/14/2004 proc sql; create table want as select * from have group by PATIENT_ID, FILL_DATE having FILL_DATE = MIN (FILL_DATE) ; quit; ... View more

Hello everyone,   I want to find continuous eligibility of certain patients with insurance. Each patient appear multiple times with different prescription dates (variable = fill_date). As a first step, i want to find the first date they filled the prescription (index date) and 180 days before and after that date. The date format is MMDDYY10. and I would like to convert to numeric so that i can subtract and add 180 from this. I have attached the code. I have to find the first prescription date for each patient. I would appreciate if you please help me verify my code and how I can convert the date   /**what i have***/ patient ID    fill date patient 1      02/23/2001 patient 1      02/25/2001 patient 1      03/14/2003 patient 2       01/01/2006 patient 2      10/14/2001 patient 2       08/25/2008 patient 3       07/14/2004   /**what i want**/ patient ID    fill date           fill_date in days   start date (+180)   end date (-180) patient 1      02/23/2001         01234                      1054                    1414 patient 2      10/14/2001          05555                      5375                    5735 patient 3      07/14/2004          33333                    33153                    33513   Thank you PROC SORT DATA = XX; BY PATIENT_ID FILL_DATE DESCENDING FILL_DATE; RUN; /** IDENTIFYING FIRST.FILL DATE**/ DATA INDEX_MONTH; DO UNTIL (LAST.PATIENT_ID); SET XX; BY PATIENT_ID FILL_DATE; IF FIRST.PATIENT_ID THEN INDEX_DT = FILL_DATE; END; RUN; /**check**/ PROC PRINT DATA = MYHUMANA.DXMED (OBS = 100); VAR PATIENT_ID FILL_DATE GENERIC_DRUG_NAME NDC; RUN;   ... View more

Hello everyone, I have this code that works in changing the length of a variable. I am doing this beacuse i want to combine datasets to one, so the variables should have same length.   i am going to stack 9 files (yr1, yr2, ...yr9) and i want them all to have the same character length for 3 types of variables (dupersid, duig, condidx). I can only do for each variable at a time. Is there a way I can circumvent this so that all the length of the 3 variables will adjusted in one data step? Thank you /*********************************************************************** - change the 8-character dupersid to 10-character dupersid by adding the panel number to DUPERSIDs of panel 22 year 2017 and panels 1-21 in the PLVE file - there is no year variable in the PLVE file - implement this change by using the LENGTH and STRIP functions as folows *************************************************************************/ data vsfile ; length dupersid $10; set new.h36u20 (rename=(dupersid=t_dupersid)); if length(strip(t_dupersid))=8 then dupersid=cats(put(panel,z2.), t_dupersid); else dupersid = t_dupersid; drop t_dupersid; run; /***this is what i try to do but does not work**/ data XX; length dupersid $10 duid $7 condidx $13; set meps.YRM2011 (rename=(dupersid=t_dupersid)); if length(strip(t_dupersid))=8 then dupersid=cats(put(panel,z2.), t_dupersid); else dupersid = t_dupersid; drop t_dupersid; (rename=(duid=t_duid)); if length(strip(t_duid))=8 then duid=cats(put(panel,z2.), t_duid); else duid = t_duid; drop t_duid; (rename=(condidx=t_condidx)); if length(strip(t_condidx))=8 then condidx=cats(put(panel,z2.), t_condidx); else condidx = t_condidx; drop t_condidx; run; Log 127 data XX; 128 length dupersid $10 duid $7 condidx $13; 129 set meps.YRM2011 (rename=(dupersid=t_dupersid)); 130 if length(strip(t_dupersid))=8 then 131 dupersid=cats(put(panel,z2.), t_dupersid); 132 else dupersid = t_dupersid; 133 drop t_dupersid; 134 135 (rename=(duid=t_duid)); - 180 ERROR 180-322: Statement is not valid or it is used out of proper order. 136 if length(strip(t_duid))=8 then 137 duid=cats(put(panel,z2.), t_duid); 138 else duid = t_duid; 139 drop t_duid; 140 141 (rename=(condidx=t_condidx)); - 180 ERROR 180-322: Statement is not valid or it is used out of proper order. 142 if length(strip(t_condidx))=8 then 143 condidx=cats(put(panel,z2.), t_condidx); 144 else condidx = t_condidx; 145 drop t_condidx; 146 run;   ... View more