Thread by @ADMBriggs

Week 3 of NHS/PHE Test & Trace data released.

Thread going through the headline figures with some interpretation 👇

It's a bit messy but please bear with.

gov.uk/government/pub…

From 11-17 June, 6,129 people tested positive through pillar 1 (testing in PHE labs/NHS hospitals for those in high risk settings like care homes & hospitals) and pillar 2 (testing for the general public).

Down from 8,478 in wk 1 and 6,498 in wk 2.

Of the 6,129 cases, 6,923 were transferred to T&T. More than total cases because reportedly includes some cases from wk2.

But this is confusing as it's 794 more cases and the difference between new cases and those reached in wk 2 is only 549 cases.

Perhaps also wk1 hangover?

Of those cases transferred to T&T, 4,869 were reached and asked about contacts - that's 70%.

This is lower than 73% in wk 1 and 75% in wk2, although may be slightly revised upwards in the next release as data filter through.

Of those 4,869 who were reached, 3,633 (75%) gave details of one or more contact.

This still sounds low to me, but up from 60% in wk1 and 65% in wk2. Either due to better data recording or more people willing to provide contact details.

And among those non-complex cases dealt with in tiers 2 and 3 of T&T, the percentage of cases and contacts reached within 24hrs is lower than previous weeks.

In terms contacts reached - T&T report that 82% of contacts reached overall, quite a bit down from 91% in each previous two weeks.

But remember that these % include complex cases (outbreaks & those involving high risk settings like care homes and hospitals - tier 1) AND non complex cases (tiers 2 & 3).
Yet those not reached ONLY relates to non-complex cases-meaning 100% complex contacts reached by default

And overall contacts per case is now 6.2 (compared with 8.9 in wk 1 and 10.2 in wk 2).

BUT if complex - those reached by PHE health protection teams - it's 40 per case (similar to last week).
If non-complex - those reached is static at 1.3 per case.

Note anomaly in report saying non-complex 11,107 contacts, which with 5,794 of these reached means 5,312 not reached. 240 fewer than in the data table. These may be tier 1 contacts not reached but it's unclear?

HT @_johnbye

https://twitter.com/_johnbye/status/1276143773744074752?s=20

So we have fewer contacts per case overall and a lower overall percentage being follow up.

BUT

Same number of contacts reached per non-complex case, same % non-complex cases not followed up, and same contacts reached per complex case.

How?

This is because of sharp drop in number and proportion of cases that are complex - those managed in tier 1 by PHE health protection teams with LA support.

1,460 in wk 1 -> 859 in wk 2 -> 479 in wk 3.

That's 25% of cases in wk 1, 19% in wk 2, and just 10% in wk 3.

Non-complex case numbers are staying around the same

4,421 in wk 1 -> 3,616 in wk 2-> 4,390 in wk 3

So why is all this happening? Working in tier 1 does feel like it's getting easier - I had thought this was because of operational reasons but it seems like the numbers of referrals are also dropping. This MAY represent fewer outbreaks/care home cases etc. i.e. GOOD NEWS.

The numbers of contacts reached per non-complex case are reassuringly consistent at around 1.3. Not sure if this is low or reasonable?
Note total contacts per non-complex case rising slightly week on week from 2.3 to 2.5 to 2.6.

Also, I've focused here on disaggregating the close contacts by complex/non-complex. But less so on contacts providing contact details.

The 25% of cases not providing any information on contacts could be from both complex and non-complex cases - health care worker who always wore PPE, for example, or where numbers of potential contacts weren't provided by a hospital trust. Or someone living alone etc.

And the 25% of cases better than previous weeks, but still seems high.

Summary points.
Cases being dealt with in tier 1 (complex/outbreaks) - declining.
Tier 2 and 3 activity (non-complex) - roughly the same.
Wk 3 data likely to be getting more reliable than previous 2 weeks (see previous thread).

https://twitter.com/ADMBriggs/status/1273720108750385158?s=20

Still far more contacts followed up by health protection teams (tier 1), than tiers 2 and 3.

That doesn't necessarily translate to different workload however. Remember that tier 3 has to phone every contact. In tier 1 we just phone a care home and may instantly get 30 contacts.

Only 70% of cases transferred are reached, down from 75%, this is more concerning.
And also concerning that the percentage of contacts reached by tier 3 is not rising.
But more cases that are reached are providing details of at least one contact, which is encouraging.

I would probably use these wk 3 data as baseline for inferring policy outcomes. The T&T system def feels like it's bedding in, with likely fewer issues arising from changes to data entry etc.

Finally, still some data concerns - tweet 3 of this thread about additional cases added to wk3 data being more than the number not reaching wk 2.

And no reporting % contacts reach separately for non-complex cases. This feels disingenuous.

And all of this is still far fewer than the estimated numbers of COVID +ve people in the community...

ons.gov.uk/peoplepopulati…

That's it.

Again - PLEASE PLEASE PLEASE isolate if you're asked to do so.

There is still COVID-19 in the community.

And it's still a really serious illness.

For info (have updated contacts per case since initial thread)
@felly500 @trishgreenhalgh @deb_cohen @martinmckee
@carolecadwalla @mgtmccartney @ADPHUK @Jeanelleuk @TimHarford @hughalderwick @HPIAndyCowper @petermbenglish @ShaunLintern @HSJEditor

and @profchrisham