Here is a real life problem to consider:

The HICDEP "DIS_ID" field (used to store CDC class C OIs) has the code for TB but the problem is, if we export from our database we are going from something general in our database to something more specific in HICDEP. In our database, we store an AIDS defining illness under Events called "Mycobacterium tuberculosis, any site (pulmonary or extrapulmonary)" and in HICDEP they have "Mycobact. tuberculosis pulm." and "Mycobact. tuberculosis extrapulm." seperately. There is in HICDEP also "Mycobact. pulm., other" (note no species) and "Mycobact. extrapulm., other" (again no species). This doesn't really help because the location is the problem. There is no general purpose "other" code either. The problem then, is how to best export TB to HICDEP format from our database. The only solution I can think of that seems mildly acceptable is to use another optional field provided by HICDEP for clarifying disease location - "DIS_OTH". In the export we can return the same patient twice, in one row "Mycobact. tuberculosis pulm." and another row "Mycobact. tuberculosis extrapulm" with, in both cases, the "DIS_OTH" field set to "disease location not stored".

I don't think the last suggestion of double entries with the use of DIS_OTH is good.

Do we need a "Mycobact. tuberculosis unknown location" or "Mycobacterium tuberculosis, pulmonary or extrapulmonary" code?

The alternative I see is to correct the data so it matches the current HICDEP codes. I checked with the CDC-C list that the definition is actually combined into one: Mycobacterium tuberculosis, pulmonary or extrapulmonary.

What do people think?

/Jesper