Thursday, 23 April 2020

Herd immunity - it looks like we are half-way there.

From the BBC:

In the study, published in a letter to The Lancet , staff at two hospitals in Newcastle were offered tests, with results returned in two days. Local GPs and paramedics were also eligible.

The staff fell into three groups:
- those dealing directly with patients (nurses, doctors, porters)
- staff who did not see patients but might be at greater risk of hospital infection (cleaners, lab staff)
- non-clinical staff (clerical, admin, IT)

Researchers at Newcastle University and Newcastle Hospitals found no evidence of a significant difference between the three groups, with rates of infection of 15% in the first group, 16% in the second, and 18% in the third.

Making some assumptions (and using my simple S-I-R spreadsheet):
- that the number of those currently infected has been increasing at a constant rate for the past four months;
- that people are infected for, and recover* after, 2 weeks
- these figures are the same in the general population (OK, they might be lower among the slackers working from home or on furlough).

Then the number of people who've had it and recovered is in the same ballpark as the number who are currently infected, i.e. somewhere between 10% and 30%.

After another few weeks (given the weird 'ballistic' way the numbers change), 50% or 60% will have it and/or have had it, which gives us low level herd immunity. This doesn't mean we've all had it and are all immune, or that there will be no new cases. It means that the new infections rate drops markedly - there are simply fewer 'susceptibles' for the infectious to infect. So instead of each infectious person infecting more than one 'susceptible' they'll be infecting less than one, so the number of new cases will fall and the disease will fade into the background, aka 'flattening the curve'.

* Apart from the handful who have died. 118 NHS staff have died with it. If you strip out those over retirement age who should have been politely but firmly told to stay at home, this is not materially higher than deaths for working age adults generally.


Staffordshire man said...

Belgium is reporting somewhat lower levels

Mark Wadsworth said...

SM, that is most mysterious.

But Belgians were testing for people with anti-bodies and came up with 3% as at five weeks ago.

UK says 16% currently infected today.

I think the anti-body test (Belgium) is not nearly as reliable as the "are you actually infected today?" test (UK).

The only way to check this is to go back to the 16% of NHS workers known to be infected today in a few weeks' time and do the anti-body test on them. If only some of them have anti-bodies, that can mean two things - the test is crap; or people who have recovered are notimmune.

Staffordshire man said...

I'm working from home and have a garden flat so I'm one of the lucky ones.They intend to relax some restrictions on 4th May; we've been locked down about a week longer than you. It's worrisome as Belgium is basically on top of the death chart (per capita)

Mark Wadsworth said...

SM, aha, hence the interest in Belgium. Best of luck and May the fourth be with you!

Nessimmersion said...

Sweden is reporting a steeper decline than the UL / Norway in infections - benefits of no lockdown & less carriers available now?