Right, and they may have a legit reason for that, such as quality control. What I question is how many Cts count as a "case".

Lemme explain here. Ct figure is simply the number of times you have to do melt-anneal-extend cycle to generate enough DNA to reach a certain pre-determined level (threshold). That figure is backwards-related to the amount of virus in your sample: the more virus you have, the fewer Cts you'll need.

Based on what we know about PCR, it works reliably when you do 20-25 cycles. It works a lot less reliably if you do 30. Beyond 30, it becomes a mess, but sometimes you have to do it to hunt a very rare target; then you go out of the way with your QC/QA, which, I bet, they don't do. I can make some assumptions and show my work, if needed. If you are past 35 cycles in a clinical assay, it's a catch-22: either you are running a perfect assay and picking up medically insignificant amounts of viral DNA, or you are running a sloppy assay and picking up false positives.

So, my question is, how many (or how few) Cts count as a case? In other words, where is that mark on the X axis, beyond which you begin to reject positives as false? If it is 20-25, then OK, but that tells you that about half of the cases reported should be disregarded as false positives. If it is beyond 30, this is where things get unreliable.

And I guarantee you, if one wanted to manipulate the number of cases, the total number of cycles would remain the same. It is the number of cycles to threshold (or the threshold value itself) that would be manipulated: if you lower the Ct value cutoff that counts as a positive (from 25 to 20, for example), you'll detect fewer "cases" in the same population. Same effect is produced by increasing threshold setting on the machine.