Thanks @sueinmn It is clear to me that two samples are needed for the initial diagnosis, but in the follow-up it is not said how many samples must test positive to declare a lack of response to the therapy. Is one enough? What I'm wondering is whether a single positive sample in the context of other negatives necessarily means you have a refractory disease or it could reflect a transient presence of the mycobacterium that has nothing to do with the initial disease... I can't find an answer to this