I see your point, but I disagree when you say MALDI-TOF was 100% for all... because while it was very specific, it had the opposite problem, being sensitivity too low.
View attachment 337952
Some of which were around 60%, which is far lower than any individual APAS test.
I think holding any test up as the holy grail is foolish, that's not how the industry works. I think there is a huge usefulness in the APAS, and your information provided confirmed my thoughts.
The point I was making around APAS is depending on the parameters in the algorithm you should be able to segregate results into (for instance)
VP - extremely positive, definite growth of definitive type.
MP - mild positive - definite growth of unknown type or unknown growth of definite type
FR - for review - non-definite growth of indefinite type
VN - very negative.
APAS algorithms can then be tuned such that VP and VN can be removed. This will take some trial and error and will likely have review stages for a few years, but if you cannot see the commercial benefit to the laboratories of automating this process and removing 40-60% of plates from review (or being able to process double the plates) then I think you're too involved in the technical flaws, rather than the business opportunity.
I look at it as a business, the technology isn't perfect, nor will it replace technicians in the near future, but it can increase technician productivity.
