Jump to content
  • Welcome to DOSBODS


    DOSBODS is free of any advertising.

    Ads are annoying, and - increasingly - advertising companies limit free speech online. DOSBODS Forums are completely free to use. Please create a free account to be able to access all the features of the DOSBODS community. It only takes 20 seconds!


Some covid SAQ NAFE and GP


Recommended Posts

On 11/09/2021 at 10:13, dgul said:

I've gone through the data  from other countries and got a range of 'the most plausible' times for the superspreader theory (essentially a 90% probability) for a new superspreader to be formed.  Min 1.5 weeks or so, max about 3.5 weeks from the peak of the first superspreader's cases.  I still think this looks short, but that's what the stats show.

So we might see that cases start to turn from about here (this is earlier than I'd supposed on the original post).  If it is going to turn it should have done so by about 23rd Sept.  If it doesn't turn by about then it means either that the most recent superspreader 'didn't find another superspreader to infect' or that it is all bollocks.

Well, we've reached 23rd without NZ turning over -- perhaps it is all bollocks... But I'll leave it to the end of the month just to give them Kiwi superspreaders a chance.

Link to comment
Share on other sites

  • 3 weeks later...
On 08/09/2021 at 22:46, dgul said:

I am fascinated by the Hope-Simpson model of viral transmission (super-spreader) and what we're seeing in New Zealand. 

New Zealand is particularly interesting because they've had very few cases to date, and thus will have very little in the way of natural immunity, but they've also vaccinated a fair proportion of the population.

This is the current situation:


[It is worth noting that AFAIKT there are only two sets of home-grown infections in the above graph -- one in early 2020 the other last month.  All the noise in between appears to be imported cases that were quarantined]

The recent wave appears to have dissipated to near zero now.  It has been very short and sharp, with a total duration of about 15 days and a peak of about 60-70 cases per day.  The current rate of about 20 per day could be the limited number of normal community transmission infections, which will fizzle out as community transmission of covid has (according to the superspreader model) an R0 of about 0.2 (I think, according to rough calculations).  

I wonder if this is what a single super-spreader event looks like; a single person become highly contagious (for whatever reason) and spreads the disease (by whatever mechanism) for around a fortnight and then 'sorts out their infection' (by whatever means).  As the majority of people aren't superspreaders the people they spread covid to mostly don't go on to spread the disease to others, and so if there is no new superspreader the disease fizzles out.

Looking at the graph above, it looks to me like the spread of disease in early 2020 was comprised of two peaks close together, and about twice the duration of the recent peak -- was this two superspreaders?

So, what does this mean?  Well, the big difference between this year and last year is that NZ now has lots of vaccinated folk.  The question is, what does vaccination do to the risks of becoming superspreaders?

There are two main theories about where superspreaders get their disease:

  • Latent infection.  In this hypothesis a person remains infected but at a low level for several months until something comes along to suppress their innate immune system.  Hope-Simpson suggested vitD levels, and this certainly matches what we see with seasonality.  However, in the case of New Zealand they've not had cases recently to create a stock of latent-infected.  So, if this hypothesis is true we should see the next covid wave in NZ in about 3-4 months (as this appears to be the normal inter-wave period).
  • Immune system suppression.  In this hypothesis a person will only become a superspreader if they have a certain quality suppressed in their immune system.  It seems easy to say 'innate immune system' here, but it could be to do with specific (perhaps enhancing) antibody types that are generated only occasionally following infection.  If this hypothesis is true we should see the next covid wave in NZ as soon as those infected have a chance to turn into superspreaders -- and if it is to do with antibodies this should be in about 2-3 weeks.
  • There is another point, though.  We saw from the early 2020 wave that the sorts of infected numbers that occurred then wasn't enough to create new superspreaders (although note seasonality -- they were coming out of summer and so had robust vitD levels)  -- so the presumption is that superspreaders comprise less than about 0.2% of the population -- this seems tiny, but does sort-of account for some of the patterns we see (but not others, which really need them to be at 1%-5%).  So, if this is still the case then the next wave (either in 3-4 months or 2-3 weeks) should be at a similar scale to the current wave.  However, if vaccination increases susceptibility to becoming a superspreader, then we should see the next covid wave magnified by this increase in susceptibility.
  • Alternatively, if vaccination reduces susceptibility to turn into a superspreader, then we shouldn't see any wave as directly following this latest one (although, of course, a brand new covid wave could form).

So, that's here's my guess for NZ's future.  I think that the immune system suppression coupled with 'dodgy antibodies' idea looks sound, and that vaccination increases superspreader numbers.  If this is the case we'll see a sudden and dramatic rise in cases in NZ at around the end of Sept, with the peak of the wave at a multiple of this recent peak (only 60-70 per day) and duration of about 6-8 weeks (as seen with other covid waves).

I'd note that the latest wave was pretty much only Auckland (as it 'was only one person'), but that NZ's biocontrol measures aren't very effective (it being an island is what's effective) and the next wave will see outbreaks in every city.

I'd be content to be proven wrong -- the superspreader model has lots of parameters and I'd be happy to populate some of them even if it goes against my current thinking -- and, of course, the superspreader model could be incorrect (but it does explain 'what we see' better than the 'measles' model of disease progression, as used in the Imperial covid models).  But that's my current thinking -- over to you, NZ superspreaders.

Okay.  Time for an update.

This is the graph of the cases for NZ as of today:


So, here we are.  The latest infection figures are similar in scale to the peak in Apr '20 and seen in late August.

The turn appears to have come in at the end of Sept.

I suggested that this is due to a single superspreader being present in Auckland in August.  The update data isn't proof that this is the case, but it certainly is difficult to explain using the SIR (susceptible - infected - recovered) model beloved of the modellers advising government.

But anyway, we now have some data from a 'covid naive' country about how it spreads.  It appears that we have a mini-wave which lasts about 2 weeks and has a few hundred infected.  Then, there appears to be a period with 'lower cases' for about two weeks, and then a new wave starts.  I predict that this second wave will have multiple-fold cases compared with the first -- however, the impact of their rather extreme lockdown should dampen things somewhat, so the scale of the second wave isn't easy to predict (I do think lockdowns work, but only when all non-essential activities are stopped -- which likely has a larger impact overall than the covid would if they only did a targeted lockdown).

My interpretation from the data so far.  There is an initial wave that then goes away, and then a new secondary wave comes along.  You do actually see 'similar' elsewhere -- most recently in Victoria, Australia, where they had a mini wave peaking around 20th July and then their current (major) wave started around 20th August.

There are two explanations that I'll explore further, superspreader and delayed-infection.

Superspreader theory. 

  • Here there are a single superspreader who infects most people.  Each superspreader infects a few hundred people and remains contagious for about 2-3 weeks.  In the August '21 wave there was a single superspreader.  In the Spring '20 wave there were two superspreaders, with peaks about 2 weeks apart.  The superspreader R-number will be in the low hundreds.
  • After the superspreader 'stops being infectious' there are a load of people that are now infected.  But they don't spread so much. This leads to a period of lower and slowly declining infections.  It looks like the R-number of these 'normal-infected' is under 1, which explains why the cases don't explode in the post-superspreader period (this is important -- SIR can't explain this).
  • Sometime in the superspreader event one or more superspreaders are infected.  But they don't become contagious immediately, but instead their infection 'simmers' somehow.  After about 4 weeks (2-6 weeks) they become a fully-fledged superspreader, and then start spreading to others. 
  • The big question is how likely an individual is to become a superspreader.  The data from the April '20, where a few hundred were infected, suggests that it is rather rare, and of the order of 0.1% or so.  We're now seeing a secondary wave after fewer infections, which might just be a fluke but might also suggest that the chance of becoming a superspreader has increased for some unknown reason (cough -- vaccines).
  • We'll know more as this secondary wave progresses -- at this point it might be another mini-wave peaking at about 80 cases per day (single superspreader) or explode higher (the original superspreader created multiple superspreaders).  
  • The one aspect of all this is that the growth of the current secondary wave appears to be a bit slower than I thought would happen.  I'm still thinking about this, but it could just be due to natural variability in the infectiousness of each superspreader -- ie, the first superspreader to arrive was a bit rubbish at superspreading, but potentially others will now join them and the wave will grow.

Delayed infection:

  • It is possible that what we're seeing around the place now is that some people (cough -- vaccinated) have a delayed response to covid infection.  The usual disease progression is infection, then a few days of no symptoms, mild symptomatic disease for about 7 days, then either recovery or a shift into serious-covid.  It is possible that there is a sub-set of the population who upon infection manage to control the disease for a few weeks without symptoms appearing, and then all of a sudden the disease becomes symptomatic.  
  • I find it more difficult to match this theory with the data that we see, but it can't be excluded.
  • The problem is the recovery from the first peak -- I can't believe that all potential infectees have been infected to allow the wave to dissipate, leaving only the delayed infection folk to keep the disease alive.  This is doubly true for New Zealand, where there is likely to be a much lower level of natural immunity to the disease and where vaccination isn't complete yet.

Anyway, we'll see where things go from here.  

FWIW, my prediction is that the case load will increase exponentially until it gets to about 1,500 - 2,500 per day in about a month from here, and then it'll magically go away (this would be due to the supply of potential superspreaders becoming exhausted for some reason, rather than the population having reached herd immunity).  

[One thing that surprises me is that the NZ leaders seem to have given up -- they're now saying that they think that it has now got out of control and that they might as well relax lockdown in time for Christmas and reopen the borders.    This is weird, because it has been this bad before and their (extreme) lockdown managed to quash covid back to near zero.  Why give up now?  IMO this is an indication that there is an understanding by the various authorities around the world that something has changed and that now each covid wave will be more difficult to suppress.  I'd suggest that they think things are about to go badly wrong, but that this is inevitable and they'd like to be able to blame something other than their own misinformed decisions.  But the timing is weird -- it might make sense for Australia to 'give up' as covid is clearly now endemic in NSW, ACT and Victoria, but it doesn't make sense for NZ to give up, unless they've got an ulterior motive...   I'd note that it is possible that 'the thing that has changed' is everyone is now vaccinated, but I'd accept that it might just be Delta.]


  • Agree 2
  • Informative 4
  • Cheers 2
Link to comment
Share on other sites

3 hours ago, dgul said:


FWIW, my prediction is that the case load will increase exponentially until it gets to about 1,500 - 2,500 per day in about a month from here, and then it'll magically go away (this would be due to the supply of potential superspreaders becoming exhausted for some reason, rather than the population having reached herd immunity).  



These predicted numbers are 'low'.  That's mainly a population effect -- the UK equivalent would be about 20k to 35k per day.  This sound like 'what we're at now', but it much lower than the >60k per day that we were seeing in the peak last winter, even though NZ has very low native immunity to covid.  This is because the model factors in:

  • Similar countries seem to have about this level of cases per wave.  This is likely to do with the population structure (smaller towns with less inter-mixing).
  • The protective effects of extreme lockdown (it does work -- the problem is the pain is greater than the slight gain).
  • They vaccinated 'late' and thus there are relatively more in the population with some residual vaccine protection.
  • Lower testing (this is probably the most important factor, but it is illusory -- lack of testing doesn't make the asymptomatic-infected go away).
  • They're in spring and moving into summer (also an important factor, but this one is real -- summer suppresses upper respiratory tract infections, for whatever reason).
Edited by dgul
Never apologise, never explain.
  • Informative 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Create New...