@lizwarr wrote: Indeed, I think I may have over-simplified the problem. It's data of hundreds of patients followed for 9 years with all kinds of chronic disease. I do have an excel file with codes of each year, but it's just that there are so many disease, I could't imagine unifying the codes manually... In this Excel file is there any logical way to determine if two codes from two different years represent the same disease? If so then it wouldn't be too hard to create such a mapping table. We could actually simply generate a SAS Format which maps all code variations to the latest version. Above approach is of course only possible if there is a 1:1 relationship between the code variations over the years. Things will get much harder if codes change because of changing categorisations/higher differentiation of diseases like in year 1 a decease has code A but in year two it gets either code A1 or code A2 because coding becomes more differentiated. So to what would you now map code A from year 1? Code A1 or code A2? ...not a decease but just thinking how coding for a person within the autism spectrum might have changed over the years. ... View more