Written By:
Incognito - Date published:
5:14 pm, September 25th, 2021 - 29 comments
Categories: covid-19, health, science, vaccines -
Tags: modelling, predictions, Te Pūnaha Matatini
About those numbers. They’re obviously very wrong, aren’t they? We all know that the truth hurts but when it is this uncomfortable and painful we simply decide it stops being the truth. There are words to describe and name this kind of decision making.
It might just be me, but I do get the impression that Government’s kindness is firming up. Infringement fines for Covid-19 breaches have gone up heaps and people who want to travel down alert levels now have to show evidence of a very recent negative test. In addition, they’re sharpening the language around mandatory vaccination and this is under “active consideration”. Even if Government doesn’t take it all the way many businesses will take the hint and do the heavy lifting by mandating vaccination for their employees.
Anyway, it takes a lot of guts to put those numbers out and defend them in public. Many a professional reputation of scientists has been burned in the bonfire of public displeasure not to mention savaged by the baying media with their insatiable thirst for blood.
Other scientists and experts have been quick to condemn the numbers without paying much attention to the guiding messages from the model simulations. They should know better and at least choose their words more carefully when commenting in interviews and in public. But nothing is as exhilarating for the ego as getting a mike shoved in your face and a camera pointed at you, and we all live in TV studios nowadays with our kitten- and kids-proof semi-professional Zoom studios and setups.
All models are wrong, but some are useful
– George Box –
I don’t think it is particularly helpful to single out people, but some of the criticisms of the model simulations by Te Pūnaha Matatini are frankly at the level of arm-chair amateurs with very little understanding of the matter at hand.
At the very beginning of the pandemic last year a medium-sized team of scientific experts went public with their initial modelling, which instantly became famous or infamous, if you prefer.
Left unchecked, the virus would eventually infect 89 percent of New Zealand’s population and kill up to 80,000 people in a worst-case scenario. [my emphasis]
The key point is that NZ Government did not leave things unchecked. They went out hard, repeatedly, when it was necessary and demanded by the circumstances, and I believe New Zealand is record holder in the World Stringency Index tables.
Unfortunately, many still seem to be hung up on those scenarios and because the worst-case scenario never eventuated, anywhere in the World because other countries didn’t leave things unchecked either, they still exhibit the same gut reaction to the results of the most recent model simulations.
Even some other experts seem to be miffed. Some expert reactions can be found the Science Media Centre and others are in mainstream media. One criticism that was rather unhelpful was asserting that the numbers were unrealistically high because the modellers assume that we would “just sit on our hands and do nothing”. Nothing could be farther from the truth, as a quick read of the report will tell you straightaway, but essentially it comes down to same message: if we make certain choices and then do nothing else, which is not going to happen anyway, then numbers of people ending up in hospital with Covid-19 will steeply increase in a large outbreak resulting in overwhelming of capacity of the health system and likely leading to many many deaths.
Short-term views of highly vaccinated countries such as Singapore, Denmark, and Ireland have been used to ‘show’ that the number predictions taken over one year for the whole of NZ are not plausible. These hand waving critics completely ignore many other important factors. In fact, as with other countries such as Israel and Taiwan, a rapid re-think is required when the pandemic is clearly going out of control, again. These cases show that there is no room for simplistic ‘thinking’ and simplistic ‘answers’ AKA kneejerk twitches of a flatulating brain and I believe that our Government knows this very well indeed and it shows.
One of the big unknowns at present is the vaccine effectiveness against the Delta variant.
As the effects of the vaccine on transmission remain uncertain, we consider three levels of vaccine effectiveness: high, central, and low to illustrate a range of possibilities.
This has a profound effect on the results and it should be pointed out that it is mainly the number predictions based on the assumption of central vaccine effectiveness that are mentioned and discussed in the news media. This is not easily influenced or controlled although spacing the two vaccine shots by 6 rather than 3 weeks does give a better immune response on average. There might be other ways that are able to boost overall vaccine effectiveness, e.g., giving a second shot with a different vaccine, but these options have not yet been approved and/or recommended in NZ.
They reiterated an important message again.
Nonetheless, it is not thought to be possible to achieve widespread population immunity to the Delta variant by vaccination alone, which means that moderate public health measures [excluding stay-at-home orders, i.e. Alert Level 3 or 4] will be needed to augment the vaccine programme.
In other words, and in summary:
Nonetheless, the results suggest that a combination of high levels of vaccination within the community, a strong test-trace-isolate-quarantine system (assuming case numbers are kept sufficiently low) and moderate public health measures may be enough to attain population immunity, greatly reducing the need for strong public health measures, such as stay-at-home orders and workplace closures.
It is worth pointing out that the attitudes and sentiments of Kiwis towards Covid-19 are relatively stable, possibly trending upwards into more positive territory, which is possibly due to the recent Delta outbreak, and he overall potential uptake of the 12+ population is estimated to be 85.5% based on the latest survey taken at the end of August.
This provides a good starting point for discussion and public conversation. The numbers per se are not so important, as the authors of the study mention:
Model results are intended as a guide to the relative benefits of different levels of vaccine coverage in the population rather than exact predictions or a detailed roadmap.
Indeed, it is hard to fathom why this elicited such visceral response with many and why people stared themselves blind on the imaginary numbers (not in a mathematical sense) rather than on the key message. People seem to forget that the flu kills many hundreds of Kiwis each season despite an active and pro-active vaccination programme. Maybe because this has been endemic for such a long time it has become more acceptable, unlike deaths caused by small sudden outbreaks of measles or RSV, for example; anything out of the ordinary attracts a disproportional amount of attention, which is often magnified by people with invested and/or political interests. And people like to focus on (the) negatives more than on (the) positives.
Some tough decisions will have to be made in the near future and we’re all in this together, so we can’t honestly shift responsibility and blame (!) to Government-alone nor does vilifying experts help one iota when some of us don’t like what they’re saying. In fact, it undermines the guiding principles and message from science and breeds unhealthy scepticism, cynicism, and distrust of scientists and authorities with the general public, but perhaps that is the (subconscious?) goal all along?
The core plank of our pandemic measures remains vaccination although it is not (yet) the silver bullet and it might never be, as is the case with influenza.
The server will be getting hardware changes this evening starting at 10pm NZDT.
The site will be off line for some hours.
Excellent thoughts Incognito, thank you.
You know Kiwis can do anything….we've moved on from being virologists and epidemiologists and now we are all modellers.
Trouble is now we are modellers we have forgotten what went on last year during the outbreak then.
Then we are dealing with the effects of deliberate mis- and dis-information,
We took measures to prevent the worst from happening but then the worst didn't happen so obviously we didn't need to take those measures to prevent the worst from happening
That should be a line from Monty Python but sadly it's an argument seen daily in online blather.
Fortunately those fools are nowhere near the decision-making, and for that we can thank the majority of the NZ public who backed the decision-makers.
Personally I think the govt did leave a bit late to choose the pfizer vaccine & we started off with a really limited supply.
But I've paid careful attention to the explanations given by Hipkins & Ardern re how & why our vaccine supplies were ordered in increasing quantities, & the vaccine programme has been rolled & I've concluded they are plausible & make sense.
I think things are going pretty well on the vaccination front, & am hoping & trusting that vaccination coverage will get even better over the next four weeks.
Thoughtful summary of what has been happening. Thanks Incognito. Thankfully our Government listened to the science and considered our population make up. You are correct to say "this Government did not sit on their hands and do nothing" but to hear some folk, Judith Collins included, that is what is said or implied constantly.
The idea you can over react to a pandemic, where fast reactions and implementation of prevention is key is really silly. We don't stop bailing a boat because it is not sinking as fast as you first thought.
Very good post Incog. Covid is a clusterfuck mess that most nations have mucked up badly and here in Aotearoa we have the luxury of still having our fingers crossed.
The response is basically whackamole. First round is easy, second round a bit more difficult, then business gets involved and ankletaps scientific advice so the third round is much harder, then the fourth round where the anti vaxers start attacking reality, then the fifth round and the sixth …
Ardern and the Government have properly pointed out that we cannot hope to keep winning at whackamole. One day we will lose, like every other nation in the world except China that threw its might at the problem and Taiwan who were just exceptional.
This is why the vaccination program is so important. Put to one side the complaints about its slowness, no nation has reached safe levels of vaccination yet and our response fills my heart with hope.
To estimate how good a country's response is you have to:
1. See how well it has kept Covid out,
2. See how well adopted the vaccination roll out is going.
I am not surprised that the Government is ramping up the vaccination rhetoric. Every reluctant vaccination that we achieve that gets us closer to 90% or 95% means that the gnarly hard policy decisions do not have to be implemented.
Taiwan's whackamole stint this year was an outbreak of the alpha variant, not delta. Taiwan got complacent and took a few months to stamp the outbreak out. The other countries that used the elimination tactic last year have struggled with delta…Vietnam, Malaysia, Thailand etc. In Oz, Qld, SA,WA, the NT have all had small outbreaks of delta and have whackamoled them, but the virus never got into the sort of communities that delta got into in NSW and Victoria and South Auckland. Also, they have not had hardly any lockdowns apart from the first one last year, so less problem with compliance. Victoria also used the whackamole approach but after far too long in previous lockdowns, they have had a real problem with poor compliance, something which has not gone unnoticed across the ditch.
Shouldn't be a surprise, really, when ya think about it…:
https://m.youtube.com/watch?v=B4Asqq78nBE
On the death rate from influenza:
https://www.otago.ac.nz/otagomagazine/issue45/inbrief/otago664450.html
They estimated the number because most deaths are not recorded as 'flu. A lot of deaths are from heart attacks due to inflammation of the heart. This means that general knowledge of how deadly 'flu is is lacking.
When we say 500 people die of 'flu every year, people find it hard to believe as hardly anyone knows anyone who has died of 'flu. Influenza death also tends to be associated with poverty.
Modelling of any disease is hard and there are many assumptions. Epidemiology is looked down on by some other branches of science even, because the numbers feel wrong.
So yes, do the modelling based on the assumption of no intervention, but make it clear what that means. Nobody did any modelling based on gross negligence, like what happened in US and Brazil.
The important thing with modelling is what combination keeps the R value sufficiently far below 1. There has be a bit of safety margin built in due to the assumptions about effectiveness of measures and human behaviour.
It is quite funny in situations like this where everyone becomes an expert via Dr Google.
Just get the jab and then moan about it would help.
Perhaps don't present to the public THE extreme case "He said the modelling assumes the country would have no restrictions."- So We listen and hear 7,000 deaths p.a. and 100,000 infections. Did any involved; Hendy, PM, Dir of health, advisors not think of how that headline nos. would impact on those watching or reading a report on the 1:00pm briefing, or that no govt. would allow that to happen? Perhaps those in power thought that such a forecast would create urgency for people to be vaccinated, and did not contemplate what other reactions and to minimise these, especially with the mindset given what we have seen in the world and from our lockdowns ?
https://www.rnz.co.nz/news/covid-19/452180/modeller-stands-by-government-s-covid-19-data-release-amid-criticism
In our household we are fortunate to have someone who was required as part of their qualifications to attain l3 statistics and explain in part along with google some understanding to the "Outbreak tracking and projections: Update for data as at 1pm 7th September 2021"
https://cpb-ap-se2.wpmucdn.com/blogs.auckland.ac.nz/dist/d/75/files/2017/01/outbreak-tracking-and-projections.pdf
Found on Face book
Could not work out how to ref.
But a well worth read by Gerard Otto re modelling and Reti's response to Hendry modelling
80000 deaths equates to 16000 per 1,000,000 population in NZ. So how does this figure stack up against the real world experience with the virus?
The answer is here:
https://www.statista.com/statistics/1104709/coronavirus-deaths-worldwide-per-million-inhabitants/
As the table shows, the very worst statistic is Peru with 6125 deaths per million population. The UK was only 2026 deaths per million population.
Clearly there is no real world example that supports anything close to 16000 deaths per million population. So, it seems clear that the 80000 projection was completely wrong.
Modellers don't have the luxury of hindsight. They use the best available information and make assumptions to fill in the gaps.
The estimate for the case fatality rate was 4% at the time I think. With better diagnostics, and more effective treatment this has ben shown to be a little lower (about 0.3%). Obviously higher in Peru for some reason.
So a better figure might have been 10 000 deaths in NZ.
As for the 7000 per year from the latest modelling: The UK is currently averaging 100 – 200 deaths per day. This is about 3500 or more per year scaled to NZ. The UK probably has over 90% of the population with antibodies from vaccination or infection. So the modelling doesn't seem too unreasonable.
Just as well you have been sold that is all that the message was intended IMO to do, SCARE us, how the ignorant are fooled 🤦🏼♀️ I will give you 1 other scenario from the same report that the 7000 is sourced from.
85% vaccinated of over 12's with a high VW will result in 191 deaths and 1,720 hospitalisations over a year(page 6). That is a few less than the 7000. Funny how that wasn't part of the manipulation of the telecast or the analysis of others here. For those " You want answers?! I want the truth! You can't handle the truth!"
Review this link
And here I got the impression that those on the left would seek out the truth and not distort. I hope this will help those superior to me to explain the modelling and outputs to the rest of us
https://cpb-ap-se2.wpmucdn.com/blogs.auckland.ac.nz/dist/d/75/files/2017/01/modelling-to-support-a-future-covid-19-strategy.pdf
So, you think modellers should make the assumption that people will notice people dropping around them like flies and continue with life as if it wasn't happening?
Even without government intervention people would naturally develop strategies to try and reduce the chance of getting sick such as avoiding sick people etc.
To be useful models should factor in likely interpersonal and government reactions to the situation otherwise they will never be useful and will never approximate what will happen in the real world.
If the model was useful, it should be possible to point to at least one real world example where the scenario actually played out in a similar way as predicted. As I pointed out, there is no such example.
Hendy's modelling of the number of cases that we would get, given the public health measures we did take was, in fact, very close. Inconvenient as that reality may be, for those obsessed with bagging the Government.
As was made clear in the briefing, modelling, as you would expect, gave a range of scenarios from worst case to best, to inform descision making. That is what it is for.
Contrary to your statement, there are places where the worst case scenario has played out. There is good reason to suspect, as some of those places don’t even test for covid anymore, that it is even worse numbers than any we have seen.
If you have better stats please give the link.
Same document as the 80k figure, which you have already no doubt read.
Hendy predicted 0.0004% mortality if NZ imposed outbreak controls similar to what we now call Level 4. That post and discussion is from march 2020.
That 0.0004% equates to roughly 20 people.
So far we've lost 27. 18 months later.
Gotta say, the more you guys fixate on 80k, the more I have to go back and refresh my memory on exactly how accurate his models actually proved to be.
Rodney jones? screw that dude. What’s his track record on predicting the numbers for this outbreak?
In the UK people are dying at 1000 per week (5000 pa in per capita terms here), despite this the government has no lock downs as they head into winter, and is not going to establish vaccine passports.
Presuming that governments would act with a high death rate is not always accurate.
Well, going by some other nations, modellers should at least incorporate into their range of assumptions the possibility that a statistically significant portion of the population are
fucking moronsplague enthusiastsreluctant to change their behaviour, yes.Hendy included the range of possibilities from us being a bunch of
fucking moronsfreedom-loving individuals who'd want to avoid lockdowns or travel controls or isolation or quarantine, to us actually having a full L4 lockdown to control any outbreak so grandma might see the year out.Fortunately, we chose the latter extreme.
Yes I have read the report and I realise there are a number of scenarios in increasing levels of control.
But the worst case scenario of "no control" is completely unrealistic in the real world, and hence not useful, and can only lead to scaremongering, especially when picked up by the media and politicians with an agenda.
"No control" implies that people and governments ignore the fact that people are dropping like flies everywhere.
However, "No Control" is a completely unrealistic scenario because fear drives people into steps to minimise risk. For instance, businesses would start forbidding employees not to come to work if sick. In the SARS outbreak many (especially in Asian countries near the outbreak) started wearing masks. And people would start avoiding people who appeared sick.
So, a more useful worst case scenario would be to allow for natural type control measures such as these, rather than assume no control measures whatsoever.
Well, including a null value can help provide context to the scale of other responses.
But on the flipside, the yanks and brits spent much of 2020 closer to "no control" than any genuine attempt at infection control, so I'm not sure how "unrealistic" it actually was. Spring Break and Sturgis, for example.
It was made in March 2020 when the death rate from infection was still a guestimate (before the treatment with dexamethasone).
Modellers do not come Out with a single outcome, refer to my post above and the link to the paper that the 7000 deaths with a different mix of variables were based on, same paper has 191 deaths. there is a matrix of variables and from these tables of outcomes. Funny how no one else can link there comments to the tables or raw data, perhaps they do not know how to find these. If only they had google and 2 minutes
The link to the full study is in the first three words of the original post (OP):
It is also highlighted further down in the OP:
People ought to read before they comment in which case they wouldn’t have to use “google and 2 minutes”.
I have read both links I have referred to regarding the report, my google and 2 minutes comment referred to seeking out and locating the report and the background to the methodology, which I cannot see the post refers to.
Regarding the 50% effectiveness there is no allowance that there would be a response of a 3rd jab, which anyone with a brain in leadership would instigate.
The link right at the top of the OP links directly to the original study by Hendy et al. In that report, you’ll find the references to the methodology they used, as is good practice in scientific study reports.
The authors of the report clearly mention waning immunity and booster shots:
People ought to read before they comment.
"People ought to read before they comment." Who is that directed to ??
You did not link this that I can see.
https://cpb-ap-se2.wpmucdn.com/blogs.auckland.ac.nz/dist/d/75/files/2017/01/outbreak-tracking-and-projections.pdf