Read on Twitter
1/ Yesterday, Matt Hancock declined to answer a question about contract trace redundancies, and instead launched into praise for the service, incl. a claim that 98% of contacts are reached within 24 hours.
Let's just check that out, shall we, twitter? https://twitter.com/Channel4News/status/1358858214507479051
Let's just check that out, shall we, twitter? https://twitter.com/Channel4News/status/1358858214507479051
2/ Let's go to the weekly tables, available https://www.gov.uk/government/publications/nhs-test-and-trace-england-statistics-21-january-to-27-january-2021 and look at the most recently available stats (week Jan 21st to 27th)
3/ Well, it looks like Hancock is right. Here's the relevant bit of Table 17. It says 98% of those identified as recent close contacts of a +ve case were told to self-isolate within 24 hours.
4/ For clarity, this is for cases undertaken by the T&T system, and not by Health Protection Teams (HPTs), who deal cases separately with cases in institutions e.g. care homes, prisons etc. as these raise different issues.
But looks like an open & shut case. 98% is in the data.
But looks like an open & shut case. 98% is in the data.
5/ Well not quite. Because the question arises: these close contacts are reached within 24 hours of *what* exactly?
When you listen to Hancock, you probably assume that it's within 24 hours of the test coming back positive, don't you? But is this really what this table means?
When you listen to Hancock, you probably assume that it's within 24 hours of the test coming back positive, don't you? But is this really what this table means?
6/ Well, it turns out it doesn't mean that at all.
If we look at table 13, we find that only 75.7% of those testing positive were reached and asked to provide details about recent close contacts within 24 hours of the "case being transferred".
If we look at table 13, we find that only 75.7% of those testing positive were reached and asked to provide details about recent close contacts within 24 hours of the "case being transferred".
7/ So things are starting to get murkier. It looks like Hancock's 98% figure refers only to close contacts who were contacted within 24 hours of the person testing positive giving their details, *not* within 24 hours of the test coming back positive.
6/ Still, it's safe to assume that for all these 75% positive cases contacted within 24 hours and asked for their close contacts' details, the call centres will great straight on the blower and get these contacts self-isolating, right,
Well, erm, not quite.
Well, erm, not quite.
7/ Because if we turn to table 18, we find that only 63.6% of those identified as recent close contacts were told to self-isolate within 24 hours of a positive case being transferred to the contact tracing system.
8/ It's ok if you're starting to get a bit suspicious now, twitter. So am I. I mean, what does a case "transferred to the contact tracing system" even mean?
Let's have a look at the methodology for all this stuff, shall we?
Let's have a look at the methodology for all this stuff, shall we?
9/ Here it is. But something's off. While the tables give figs for a) +ve cases contacted by T&T within 24hrs of transfer; b) contacts contacted within 24hrs of T&T getting their details, there's no data on how long it takes between a test coming back +ve and its transfer to T&T
10/ This seems daft. Public health interest lies not in how fast cases are processed once in T&T system, but in how long it takes overall between a person getting tested and their contacts isolating; while there's testing data and tracing data, the bit in the middle is missing.
11/ Does this explain why there are lots of call centre tracers reporting that they end up calling people who are several days into their 10 day isolation period? Is the 'transfer' process being held up within the overall T&T system to make the figures look good?
12/ I don't know exactly, but the fact that there's this gap in the data makes me suspicious.
And you know what else makes me suspicious, especially about the way people are being made redundant at a time when only 63.% of contacts are being contacted within 24hrs of "transfer"?
And you know what else makes me suspicious, especially about the way people are being made redundant at a time when only 63.% of contacts are being contacted within 24hrs of "transfer"?
13/ It's this note - the note at the bottom of all the tables I've referred to so far.
Read it carefully.
Read it carefully.
14/ Yep, that right. It really does say that. Since late November, NHS T&T has been counting as 'contacted' people they've not actually contacted, because they get to just tell the person who's tested positive to do it for them, if the contacts are in the same household.
15/ Now, it's arguable, but I'm not even sure that's legal. Para 14 (6) of the Coronavirus Act, and the July guidance on this makes clear that anyone being required to self-isolate must have their well-being individually taken into account by the relevant public health officer
16/ But heh, that's complex, so let's leave aside for now whether the law is being broken (some of it depends on how a PHO is defined - apply for details) and focus back on the numbers.
17/ Usefully, table 16 tells us not just how many people identified as recent close contacts were reached and told to self-isolate by the method used to reach them, but also whether they were in the same household or not as the +ve case.
18/ And yes, you're reading it right. Some simple maths tells us that just 9.9% of those contacted were outside the household of the +ve case. That is, 9 in 10 people counting as contacted are in the same household. And of these, 93.2% are not actually contacted by Test & Trace.
19/ So overall, for about 87% of all contact tracing claimed by Test & Trace, there's no actual contact - just an assumption that it's going to happen, and data recorded to that effect.
20/ So maybe that's why Hancock was a bit reluctant to talk about redundancy. With this set up, there's not actually a lot of work to do, now that cases are dropping but he can't admit that.
21/ There *are* things that these staff could be doing, of course. They could be narrowing the time gap between a +ve case being contacted and the 13% of close contacts they do need to call, or - more ambitiously - they could be doing more actual contact tracing
22/ But, for reasons I set out https://bickerrecord.medium.com/self-isolation-and-the-law-41a72e765b2a, it is not in the commercial interest of the call centre firms to change the skill mix to enable more pro-active contact tracing, & nor it seems is it in their interest to tell us how long it *really* takes to transfer cases
23/ I don't know all the answers to what should be done, but what I am pretty clear about is that, when Hancock tells you it's 98% in 24 hours, he's misleading us.
END
END
