Your method of normalization seems too strong to me. Suppose there are two groups of people getting tested: 1. those that have good reason to believe they have Covid (symptoms or extended contact with known case) and 2. those people who have no good reason to believe they have Covid (no symptoms, no known extended contact). Your method seems to assume all tests are of group 2 variety, and therefore more tests -> more +'s. However, for those in group 1, the causality can go the opposite... it's because there are more symptomatic people that we have more tests, so that is truly a signal of more cases. So, to me, to normalize properly we would need to know the split between groups 1 & 2. My anecdotal evidence would say there is a ton of group 2 vs. group 1 (and thus your method is pretty good), but I don't know if we have data for that.
it's a worthwhile point. i considered that issue a ways back and looked for ways to tease out the selection issue. what seemed to emerge was that above a certain sample rate (that's very low vs current levels, maybe 250k vs 1-1.5 million now) you can find a bit of evidence, but not at higher rates. it seems to saturate and vanish as an issue. so we saw it in the positivity figures early on (likely because most testing was among symptomatic) but it more or less vanished by april 2020. the testing numbers here have been so high for so long that selection bias looks to have washed out pretty quickly.
Doesn't this basically just bolster the longstanding guidance that PCR and Antibody tests are primarily to be used as supplements to clinical presentation and assessments of symptomatic patients, rather than blanket testing of the general public? Even with patients who do in fact have the virus in their system, the timing of the test has the potential to provide false results.
i think those are sort of adjacent issues, but broadly, yes. there is a reason that every vaccine trial required symptomatic confirmation (or symptomatic presentation to even test at all) in addition to a PCR positive as a standard for a "case." not doing so in the broader public, especially when asymptomatic spread is so rare has never made much sense epidemiologically. over testing and poor use of tests that are not really fit for purpose (and extremely oversensitive) has led to outlandish overcounting and to high levels of false and false clinical positives. this has been diagnostic malpractice.
Scariants smariants. Here's something to ponder. Astra-Zeneca vaccine was trialed in UK, South Africa, Brazil. The new variants were discovered for the first time in UK, South Africa and Brazil, and nowhere else. Alarm bells ringing? Calm down, correlation does not prove causality (nobody has bothered about that dictum very much during this panicdemic). Here's my hypothesis: To accompany the trials, genomic sequencing was massively increased in the trial countries, so much so that UK did the most genomic sequencing of anywhere worldwide. I haven't confirmed this, but in my opinion, the scariant hysteria has been caused by "Seek, and ye shall find". Very similar to the PCR fiasco fueling the "cases" hysteria. Final thought: would it have been better for humankind NOT to have had the half-understood technology with which to frighten itself?
Your method of normalization seems too strong to me. Suppose there are two groups of people getting tested: 1. those that have good reason to believe they have Covid (symptoms or extended contact with known case) and 2. those people who have no good reason to believe they have Covid (no symptoms, no known extended contact). Your method seems to assume all tests are of group 2 variety, and therefore more tests -> more +'s. However, for those in group 1, the causality can go the opposite... it's because there are more symptomatic people that we have more tests, so that is truly a signal of more cases. So, to me, to normalize properly we would need to know the split between groups 1 & 2. My anecdotal evidence would say there is a ton of group 2 vs. group 1 (and thus your method is pretty good), but I don't know if we have data for that.
it's a worthwhile point. i considered that issue a ways back and looked for ways to tease out the selection issue. what seemed to emerge was that above a certain sample rate (that's very low vs current levels, maybe 250k vs 1-1.5 million now) you can find a bit of evidence, but not at higher rates. it seems to saturate and vanish as an issue. so we saw it in the positivity figures early on (likely because most testing was among symptomatic) but it more or less vanished by april 2020. the testing numbers here have been so high for so long that selection bias looks to have washed out pretty quickly.
Doesn't this basically just bolster the longstanding guidance that PCR and Antibody tests are primarily to be used as supplements to clinical presentation and assessments of symptomatic patients, rather than blanket testing of the general public? Even with patients who do in fact have the virus in their system, the timing of the test has the potential to provide false results.
i think those are sort of adjacent issues, but broadly, yes. there is a reason that every vaccine trial required symptomatic confirmation (or symptomatic presentation to even test at all) in addition to a PCR positive as a standard for a "case." not doing so in the broader public, especially when asymptomatic spread is so rare has never made much sense epidemiologically. over testing and poor use of tests that are not really fit for purpose (and extremely oversensitive) has led to outlandish overcounting and to high levels of false and false clinical positives. this has been diagnostic malpractice.
Scariants smariants. Here's something to ponder. Astra-Zeneca vaccine was trialed in UK, South Africa, Brazil. The new variants were discovered for the first time in UK, South Africa and Brazil, and nowhere else. Alarm bells ringing? Calm down, correlation does not prove causality (nobody has bothered about that dictum very much during this panicdemic). Here's my hypothesis: To accompany the trials, genomic sequencing was massively increased in the trial countries, so much so that UK did the most genomic sequencing of anywhere worldwide. I haven't confirmed this, but in my opinion, the scariant hysteria has been caused by "Seek, and ye shall find". Very similar to the PCR fiasco fueling the "cases" hysteria. Final thought: would it have been better for humankind NOT to have had the half-understood technology with which to frighten itself?