Kappa Analysis where percent negative agreement will always be zero

[BTAinRVA]

I'm not real familiar with Kappa analysis. We are comparing the agreement between two databases on cancer diagnosis and we have some patients in both and some in one but not the other. However there will never be the case that a patient will be in niether (at least to our knowledge). Is it appropriate to use Kappa analysis in this case?

Thanks,

Brian

[Ksharp]

If your contingency table was a square table, 
then you can do Kappa analysis. Like:

    F  M
F 1   2
M 3  4