Test & Trace and the mysterious 16%
Since the second wave really started to surge, the Test & Trace figures have slipped from the headlines.
But the data continues to be collected and are available weekly, up to Jan 6th currently, and because I smelt a rat, I had a quick look at performance.
Of particular interest is table 15, which is entitled “People identified as recent close contacts that were reached and told to self-isolate by whether they were household contacts, contacts not managed by local HPTs [Health Protection Teams].
This, for clarity, is the data for the contact tracing done by the private call centres, as opposed to the more complex tracing work done by the HPTs (Public Health England) when, at the post-test result triage stage, it is decided whether the positive case relates to some kind of institution e.g. care home in which case the HPT coordinates tracing with that institution and not directly with individuals.
For the most recent week in the table (to January 6th), this table sows (cells BL5 & BL6) that 612,377 (83.7%) of the 731,733 people counted as traced contacts of the person testing positive were in the same household as the person who had tested positive.
This is not surprising, but what is noteworthy is the note that accompanies the table:
From 18 November 2020 there has been a change in how household contacts under 18 are contacted. They are no longer contacted individually by contact tracers, but instead the case is asked to inform these individuals that they need to self-isolate. The case provides basic information for the contacts and they are marked as complete. From 27 November 2020 this has been extended to cover adults in the same household also.
What this means in effect is that the call centres dealing with ‘routine’ tracing work are in fact actively tracing about 16% of those who are subsequently added to the data as having been traced and contacted by those call centres. 
This should, I think raise some questions. The central one is whether the private firms contracted to run the call centres are still receiving the same fee for contact work that has now been effectively delegated to people who have tested positive for Coronavirus (and might therefore not actually be very well).
I think we can guess what the answer is. As I set out when the legislative change were introduced in September, there must be at least some possibility that those changes were suggested by the providers in the first place. 
The other question, perhaps of similar importance, is what contract tracing call centres staff are doing now that much of their actual contact tracing work has been removed.
And perhaps the answer to that might also be found in the weekly tables. For if we take information for the week up to January 6th fro table 10 and table 18 we find the following>
In that week, according to table 10 (cell BL5), 336,129 people tested positive for coronavirus and had their case referred to the general (non-Health Protection Team) for tracing purposes. In the same week, according to table 18 (cell BL4) 672, 787 people identified as recent close contacts were told to self-isolate.
This works out at pretty well exactly two people told to self-isolate per person test positive and referred to the non-specialist tracing service; while there will be some slippage across weeks on account on the time taken between referral and contact tracing, such that the 672, 787 told to self-isolate on account on their contact with a positive case cannot be matched exactly to the 336, 129 positives, this still gives us a pretty good indicator of how far tracing is reaching.
So is two added people told to self-isolate per positive case about the right number you’d expect? Well, it’s hard to say exactly, but in New South Wales, where they’ve done contact tracing pretty well so far, Professor Rain MacIntyre says this about gold standard contact tracing:
One case may have 10–25 contacts, so if there are 100 cases in a day, you need to trace 1,000–2,500 contacts within 24 hours.
The data from the tables, then, suggest some kind of fix my be available. We just need to join the dots.
If, as we know is happening (I have private messages to this effect) a lot of tier 3 call centre staff (see page 5) are actually now in devalued “bullshit jobs”, with a lot of calls they make going to people who have already been isolating for some time, then what we should perhaps be doing is looking to change the whole tiering model, and upskill lots of tier 3 staff into tier 2 roles, in order to: a) increase the number of contacts per positive case by careful questioning; and
b) bring in additional functions to the test and trace service for the same money, notably offering real support for self-isolation by co-ordinating physical resources e.g. booking hotel rooms and taking the burden off worried people by ensuring they get self-isolation payments.
Action b) would, I should add, be entirely in line with the December NHS Test & Trace business development plan, signed off by Dido Harding, which makes this commitment as part of a wider objective of improving support for self-isolation:
NHS Test and Trace will provide analysis and insights to help tailor optimal packages (p.14)
As yet it is difficult to see any progress on this, so perhaps Dido Harding needs a bit of help with how the resources might be deployed to make it happen.
So where from here?
Clearly, the Test & Trace firms now making comfortable margins by not running a good service are not likely to want to play ball in any substantive improvements, so any pressure that can be brought to bear will be needed.
But this meas getting down to the detail. I’ve looked at three tables, and while I may not have got all my analysis right, it seems clear from the data that there are questions to be asked. I’ve done that in a few minutes with publicly available data.
At the moment though, Labour is content to project its own managerialist style on to the whole issue, arguing blandly that Test & Trace should be “fixed”, but offering no insight into how it might be done.
Quite simply, it needs to start doing what its leader was elected for.
That’s to get forensic, on this and other Tory incompetence and corruption.
 There will be a few cases where the person who has had the positive test provides details of other adults in a household but then refuses to committing to ask them to self-isolate, such that the burden falls back on the call centre, at least officially, but I suspect these numbers are very small.
 In that earlier piece, my focus was on the wording in the secondary legislation which opened the way for the private firms to employ less skilled and experienced staff, because that wording gave a wider range of people the legal power to instruct people to self-isolate.
In this case, however, what is important is that the legislation (section 2, para 2 c) i) shifted, in terms of children’s self-isolation, the legal responsibility “to secure, so far as reasonably practicable, that the child self-isolates for the period specified” from the Test & Trace service and on to the adults in the house. What is murkier is the later change reflected in the note in the table, is how getting other adults in the household also become the responsibility of the person testing positive, as this does not seem to be in rhe legislation. Again, this raises a question.
 Professor MacIntyre also says that Victoria State probably did not do as well as New South Wales because it did not have enough well-trained contact tracers. The comparison with the private firms’ minimum wage and minimum training strategy is pretty obvious.