The issue is that, for any epidemiology investigation, there needs to be a “case definition”.
Beginning in January 2020, the case definition, worldwide, for COVID-19 was a positive nasopharyngeal or oropharyngeal swab for SARS CoV 2 RNA. This case definition was both provisional (as any new case definition is, pending further knowledge) and appropriate for a new disease caused by a new pathogen where human immunity was presumed to = 0%.
Since then, we have learned a lot. But our increased knowledge still does not allow accurate classification of the large swab + population who are asymptomatic, who have had symptoms for >10 days, or who have antibodies.
They are all still “cases” the first time they come to attention through testing programs.
It is intuitive to say, if they have IgG antibodies to SARS- Co-V 2, that they are both “recovered” and “not infectious” - but this, although logical, remains to be proven. For example, a patient who was positive by swab a month ago and still is is (almost surely) noninfectious, but I do not allow nurses to care for that patient with no PPE.
It’s also a problem that there are over 200 antibody test products still permitted under the March EUA, many of which detect antibodies to other coronaviruses (they are not specifc).
And further down the rabbit hole - there are people, lots of them, who have evidence of cellular immunity to SARS Co-V 2 who have never been exposed and who are swab and antibody negative. It’s attractive to believe that this is because prior CoV infections confer immunity to SARS-CoV-2 in some people, maybe in a lot of people.
Antibody testing is not ready for prime time in terms of counting cases. This won’t be understood well enough to change the provisional case definition for quite some time.
Hope that helps.
So it’s messy as hell. Which is like many things in the real world, I suppose.
I’m just trying to get a handle on real spike vs. increased testing spike? But there isn’t much of a handle, it’s all very slippery . . .