data have; input ID $ Specimen :$10. Date :mmddyy. Status &:$20.; format date yymmdd10. ; cards; 12345 W2345678 1/5/2020 Corrective Final 23456 M9876799 3/4/2023 Corrective 84569 N45678987 5/22/2022 Corrective Final 78766 P73839303 6/9/2022 Final 35473 R37390202 4/7/2023 Corrective ;   data want (drop=cntw); set have; cntw=countw(Status); if cntw >1 then output; run; ... View more

Sorry, without a clear cut example can't provide anything.  What I would say is that what you are doing is called "binning" the data.  There are numerous papers out there and several different methods.  For example one way would be to normalise the data - i.e. instead of going across, have all the data going down, and then use something like proc rank: http://documentation.sas.com/?docsetId=proc&docsetTarget=p0le3p5ngj1zlbn1mh3tistq9t76.htm&docsetVersion=9.4&locale=en Or apply the logic yourself in a datastep. ... View more