Any person not entering ICU is a huge saving.
So... 1.5k positives/day have been found recently (moving average). Out of 100,000 tests (I think). Suppose that one person is stopped from getting infected for every two positive cases found (because everyone should isolate when symptomatic anyway). And that 10% of infected persons get a hospital stay, and 50% of those who went to hospital went to ICU. And the average stay in ICU is... 8 days, plus 2 days before & 2 days after in hospital not ICU beds.
And a day in hospital not in ICU is half cost of an ICU bed day.
So 100k tests = £1 million spent (maybe)
1500 confirmed positives -> 750 saved new positives.
10% of 750 = 75
38 saved from going to ICU.
£1 million /38 = £26,315.
Does a day in ICU cost >= £2631.5?
I'm finding 2010 values of like £1400 for a critical care bed, so £2000 for 2020 bed days in critical care might be about right, not sure.
I'm ignoring the 38 cases who only went to hospital but not to ICU, of course.
Are real health economists working out actual savings with better numbers?
Seems like the tests need to cost < £10 each to make testing as control strategy cost-effective.