Hi,

I would like to know if anyone has a nice (efficient, comprehensible) way to map icd9 and cpt4 codes into numerics for use as an array index, and then of course to do the inverse mapping back to the original character codes.

Some of these codes have alpha characters in the first position, i.e. 'V991', or 'E2691', but most are simply 4 or 5 numbers stored as character, i.e '19772'.

My intented application is to read health care claims data with multiple diagnosis and procedure codes and create some large temporary arrays that hold the counts of the number of occurrences of the each of the codes. I know there are other ways to do this, but I'm really curious about how best to address the mapping and inverse mapping of codes to numerics for the array indices. Please no proc summary suggestions!

here is a scant sketch of the code I have in mind:

array code_counts{0:?} _temporary_ ; code= ?format function?(icd9); code_count(code)+1;

thanks very much! Hugh