Michael norton – “you might conclude that if you do not test you will not find?” and “May be they just don not want to know?”
The difference more likely reflects how well targeted the tests are.
– – – – – – – –
Some people have been claiming that high case numbers are purely the result high rates of testing, and thus do not tell us the prevalence of infection. But if this were true, the curve of positive test results would be a similar shape to the curve of number of tests performed. Those two curves can be compared:
https://coronavirus.data.gov.uk/details/cases
https://coronavirus.data.gov.uk/details/testing
The two curves bear no resemblance to each other, so something else must be responsible for the shape of the positive tests curve.
Until the vaccination programme, the hospital admission curve, and the two different deaths curves, all followed the shape of the positive tests curve, lagging by about one week and two weeks respectively. So the positive tests data was predicting hospital admissions a week later, and deaths two weeks later. That is very strong evidence that the tests are reflecting the prevalence of infection.
Remember that no one gets sent to hospital just because their test comes back positive; they’re merely asked to self-isolate. People attend hospital or get sent there based on their symptoms. One of the deaths curves is based on positive tests, but the other is based on death certificate data, ie. clinicians’ assessment of cause of death.
– – – – – – – –
However, the more idea the testers have of which people to test, the more likely each test will be positive – ie. they don’t use so many tests on randomly checking people who turn out to be negative.
A perfect trace-and-test system should find all positives without checking anyone who hadn’t been exposed to infection. The entire population of New Zealand could be checked without finding one positive case, because there is no infection within New Zealand.