RE: Sp predictions on a mega rns26 Jul 2020 09:17
I’m going to repost in here as we’re talking about S&S... Our published sensitivity might be lower than anticipated. Not because the test will be poor at what it does, but that Condor might come up with people that are beyond the most useful period when the test should be deployed. We need them to come up with folks within a week of infection and test them very regularly to understand how it’ll perform in the real world. If they drag out a random bunch of samples, including those beyond the first week, then we could end up looking like Abbott Now.
Going back to the great graph from the Rockefeller fella we need to be demonstrating that we’re extremely good at finding positive cases in that sweet spot. I hope they publish the Ct figures and then communicate this well, something like 75% sensitivity overall, but 98% in the peak infectious period. After all, that’s what matters. Sir Al is a very good communicator, so I would hope this is being considered. Only presenting blunt figures might pose a risk of misinterpretation. Again, we have the luxury of Sona blazing a trail so we can learn from any mistakes.