In the context of a clinical trial for a TB vaccine in South Africa, during contact tracing a single induced sputum had a similar level of diagnostic yield to a single gastric lavage, but with relatively low overlap.  Here.

Is it shameful to say that my eyes went a bit cross-eyed at the statistics in this paper?

First things first:  I don’t do community diagnosis in TB, and I don’t do induced sputum.  And I don’t have anything like the levels of TB seen in this study group.

The authors of this study have used a variety of combination methods to look at the way in which less than gold standard tests might pick up the totality of TB.  In short, they don’t.  If you look at table 2, it is shocking how few patients are positive for both one induced sputum and one gastric lavage.  By contrast, most who are induced sputum positive are gastric lavage negative, and visa versa.  

So, then, back to three consecutive early morning gastric lavages…