Did the Government abandon a working strategy when it shut down contact tracing after the now notorious March 12th COBRA meeting that decided to move from the “Contain” to the “Delay” phase of its flu-based pandemic plan? It’s what I said at the very beginning of April:
the worrying thing is I think we might have had a working strategy and abandoned it out of panic
— Alex Harrowell (@yorksranter) April 1, 2020
Self-publicity aside, it’s perhaps the most important question in any discussion of this disaster. This Grauniad piece contains a startling data point:
Public Health England, overseeing the scheme, believes compliance will be good, pointing out that it was able to trace and isolate 95% of contacts of cases before testing and tracing was abandoned on 12 March.
The celebrated Feasibility of controlling COVID-19 outbreaks by isolation of cases and contacts paper, which modelled the impact of different levels of contact-tracing effectiveness on simulated epidemics characterised by different values for parameters such as R0 (average transmissions per case) and k (overdispersion, or clustering), appeared in The Lancet on March 5th but had been available in preprint since mid-February. The following chart illustrates the key finding.
If 95% tracing really was achieved, even an outbreak with an initial R0 of 3.5 might have been contained. (Here’s another more recent paper with similar conclusions.) Current estimates of the real value put it much lower, between 2 and 2.5 for the Diamond Princess cruise ship outbreak, with a very high proportion of transmission coming from super-spreading events. If the PHE source speaking to the Guardian actually means 95%, in the sense of the number, 95 out of 100, rather than a general impression that compliance was good, this would also explain why the UK case numbers remained stable pretty much up to the Thursday when everything fell apart.
Interestingly, there is data that tends to corroborate an extraordinarily high degree of public consent and cooperation in the early stages of the crisis. Apple’s monitoring of requests for directions through Apple Maps suggests that most of the decline in mobility in the UK happened voluntarily, before the official lockdown on the 23rd of March:
This implies the Government has perpetrated an appalling disaster out of panic and an apparent inability to take details like managing a call centre seriously. The big question is whether the public spirit demonstrated in March and April can now be recaptured, after, well, this.
I mean it with the call centre,by the way. Much media reporting at the time had “Whitehall sources” complaining about NHS 111, which was a key element of the test-at-home protocol. However only Kate Osamor MP seems to have taken this up as a political issue, because if we started caring about call centres, where would it end? This is likely to become a political issue again very soon so you may as well skate to where the puck will be.
As the author of The Churchill Factor didn’t say: We will fight them on the beaches. We will fight them on the landing grounds. And then we will make a hash of the admin, give up, soil ourselves with fear, throw our rifle behind a hedge, and scarper, leaving everyone else to their fate. After all, what’s in it for me?
And it is Johnson who’s the problem. There was a reason the Government abruptly switched from a communications strategy implementing the Krebs principles to one that emphasised a rolling cast of ministers, about the same time things got serious. I contend that No.10 decided to “get a grip” and drove the clown car off the cliff.
(Not that Shadow Health Secretary Jonathan Ashworth MP helped much by demanding in Parliament that the Government move to the delay phase. If he didn’t know the delay phase meant no more tracing, he should have done.)
Update: This new paper uses genetic sequencing to determine if very early cases in Germany and Washington State were actually related to later outbreaks, and found that they were not, and therefore, that the initial public health response was successful.