CaseID TEST1 TEST2 TEST3 TEST4 DATE1 DATE2 DATE3 DATE4 1 WB IgG       2/5/2015       2 WB IgG EIA total EIA total   3/4/2015 3/4/2015 7/13/2015   3 EIA IgM WB IgM WB IgG   3/4/2016 3/4/2016 7/3/2016   4 EIA total WB IgG WB IgG WB IgG 2/4/2015 2/4/2015 7/7/2015 12/2/2015 5 WB IgM EIA total     7/11/2015 9/11/2015     6 WB IgM WB IgG EIA total   12/12/2015 12/12/2015 3/12/2016   7 WB IgM WB IgG EIA total WB IgM 10/8/2015 10/8/2015 11/2/2015 11/2/2015     And re:output- as outlined in my first post, I am looking at different combinations of tests (used to “confirm” a case’s disease status) for each observations. If WB IgG (regardless of other tests performed on that ID) say cat=1 If EIA total/EIA IgM and is followed by WB IgM within 30 days, cat=2 If cat=1 is true AND cat=2 is true, cat=3 If neither (only EIA total, EIA IgM, WB IgM but not within 30 days), cat=4. Output I want will be a basic proc freq of “cat”.   hope this helps understand what i need   Thanks a lot! ... View more

For the analysis that I am doing, it was important for me to see records per ID (per patient), and I thought wider format would be useful. Specifically, I had to look at statistics like # of cases- "with ONLY test A and B", "Only test B", Test B and C", "Only B and C" etc., and it sounded easier for me to pull from wider table than long one. If it works, I don't have problems going back to the long format (again.  But something i probably should have clarified before- the labs are NOT true duplicates. I want to treat like duplicates if it is the same test performed within certain duration (like 3 months).   how would you duplicate using the long format? any quick hints/sample codes? dif/lag fucntions?   thank you ... View more

Thank you LinusH! an "Observation" is a patient's unique ID. the data was transposed from long to wide, not wide to long. so, in it's orginal form, the data looked like..   ID                             LOINCS_1               RESULT_DATE 123                           987                         01-SEPT-XXXX 123                           945                         07-DEC-XXXX 134                           945                         21-OCT-XXXX 186                           756                         23-JAN-XXXX 186                           654                         12-JAN-XXXX 234                           654                         14-FEB-XXXX 234                           987                         27-MAR-XXXX   data in this form was transposed to a wider format as in the earlier attachment.   Hope that clarifies..? ... View more