ARCHIVE 3 Aug 2020: challenging Carl Heneghan's blog saying cases have not increased

THREAD: I dug into the numbers @carlheneghan used for his blog saying new cases had not increased & they are misleading. He took number of +ve specimens/day in England & divided by total number of "tests processed" each day for UK - which includes tests sent but NOT RETURNED 1/6

[3 Oct 2023 note: the quoted Tweet has been deleted.

, but the original Blog post is here: ]

I used same govt numbers to recreate his chart. Govt is clear that "tests processed" is NOT the same as people tested. https://coronavirus-staging.data.gov.uk/about-data#daily-and-cumulative-numbers-of-tests. In fact it is way higher & *cannot* be used to look at positivity rates (to see whether increased cases is just cos of more testing) 2/6

Instead you need to look at *people tested". This information is available from NHS Test and Trace report.

Unfortunately the data is only weekly and also a week behind - latest info is only up to 22 July. 3/6

But from 2-22 July, while numbers of "tests processed" increased by 77% for community tests, the "number of people" tested only increased by 30%. The positivity rate for hospital settings steadily declined, for community people it went down & then up 4/6

We don't yet know what role more testing played in the increase in cases seen last week. The upshot? Transmission was declining until mid July but the latest evidence to 22 July is that it's going back up. This agrees with ONS infection survey too https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/conditionsanddiseases/bulle tins/coronaviruscovid19infectionsurveypilot/latest 5/6

Stay tuned for Thursday for when the latest Test and Trace report comes out up for data up to 30 July. ALSO PLEASE DO NOT USED "DAILY TEST PROCESSED" AS AN INDICATOR FOR ANYTHING! Half of them are never returned! 6/6