Skip to content

Instantly share code, notes, and snippets.

@tomiam8
Last active July 3, 2021 03:49
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save tomiam8/a5a71d5d321e4889c3ab12e863790f97 to your computer and use it in GitHub Desktop.
Save tomiam8/a5a71d5d321e4889c3ab12e863790f97 to your computer and use it in GitHub Desktop.
Comments on Queensland's CMO AZ comments, and Australian Covid response more generally

Personally your comments on Australia showed a lack of understanding of the situation, and erroneously applied a US perspective. Sorry for the overly long comment, but TL;DR the Queensland CMO comments on AstraZeneca aren't that ridiculous, people should pay more attention to why Australia did well; I don't know the answer but would like to find out, I agree it's more complicated than just 'made better decisions'.

First, some ‘quick’ context: Australia's government is federated, with several state governments (e.g. Queensland (QLD), NSW, Victoria) and the federal (aka Commonwealth) government with separate duties and constitutionally given powers. Personally, state’s powers are generally much stronger in practice than the US states.

During Covid the states have:

  1. Made lockdown orders
  2. Closed borders with adjacent states (at times)
  3. Run the Quarantine Hotel system for international arrivals (despite Quarantine a commonwealth responsibility)
  4. Required all public venues to record the name, time and contact details of everyone who visited the venue for contact tracing. This then swapped to mandatory electronic recording (ie you scan a QR code on your phone, enter details into the venue's google form, or the state's form system), and then because venue's were messing up managing the check-in's themselves, only the state's form system was allowed (Yes this is pretty authoritarian, but not enough space to talk about it now. Comment if you want me to explain more.).
  5. Other stuff but this is long enough as is

Quick Covid in Aus timeline: the state's instituted and enforced lockdown orders which somewhat accidentally eliminated Covid (personally, why elimination happened in Australia and not elsewhere is an interesting question that’s too often ignored). The Commonwealth banned all overseas arrivals (including citizens), except for those who went through a two week quarantine at a hotel, with limited spaces and a queue. Once Covid was eliminated, state's gradually loosened restrictions, we had a few periods of trial and error and now the general process is: very few / no lockdown restrictions for a few months, Covid escapes quarantine, contact tracers test and isolate everyone in contact, and either it goes away in a week or has spread and the state goes into lockdown for a month+ neighbouring states close their borders while Covid is eliminated in that state again, and then restrictions are eased for a few months. Repeat. The general goal seems to be to continue these measures until enough people are vaccinated to reach herd immunity, reopen borders and have Covid spread as normal diseases do – without too much drama.

Note Australia normally has no or very small local Covid transmission. This means a lot of intuition developed for elsewhere is wrong in an Australian context. For example, without Covid actively spreading, delaying vaccinating people does not literally kill people, it merely costs billions and billions of dollars (and general human misery) when a state inevitably has to go into lockdown again that could have been avoided had people been vaccinated.

Also, every time a new variant becomes more transmissible the difficulty of the entire process - the chance covid escapes quarantine, the chance contact tracers catch everyone infected, the harshness of the required lockdown to get R < 1, the length of the lockdown - worsens. Two weeks ago the Delta variant escaped quarantine in Sydney and spread, causing the first lockdown for Sydney since (you can see case numbers at https://covidlive.com.au/report/daily-source-overseas/nsw - look at the local column, overseas means cases discovered in hotel quarantine). The experience has given us a lot of information about how the Delta variant spreads – it feels quite noticeably different with a faster time from infection → transmission and spreading in cases it didn’t before ( https://www.smh.com.au/world/europe/who-says-jury-out-on-fleeting-contact-strain-of-delta-virus-in-australia-20210626-p584g0.html WHO says CCTV evidence is whatever, for normal WHO reasons.).

Finally, while the state's managed lockdowns, the Commonwealth government handled (and promptly screwed up) vaccine procurement. They initially ordered 10 million Pfizer, 4 million AZ to be imported, 50 million AZ to be locally manufactured, no Moderna and 51 million Novavax which remains unapproved. Yes, this was stupid and the Commonwealth should have ordered far more vaccine. While a medium-sized political drama, they probably won't lose elections because of it (as they should). Earlier this year Australia's vaccine council ATAGI recommended AZ not be given to anyone under 50, and later increased that to 60. They also specified that because patient consent is paramount, under 50s/60s could choose to receive AZ if they wished and gave informed consent, but this was buried deep in the recommendation and not really reported. The Commonwealth government increased vaccine orders to get 30 million more Pfizer and 25 million Moderna to arrive from September through to December. Why there was a single dose of Pfizer/Moderna available anywhere in the world to be ordered IN APRIL 2021 for delivery IN 2021 is a mystery, but I guess at least the Commonwealth isn't the only incompetent procuror of vaccines (what a relief no private firms can buy vaccines, at least we'll all be equitably delayed).

During the recent outbreak, the PM (head of Commonwealth government) drew attention to the ‘under 60s ineligible for AZ can still consent to it’ in an announcement doctors won't be liable for blood clots to those who consent to AZ. Suddenly everyone realized they could get AZ if they wanted! ATGAI / medical establishment / state premiers while not doing anything to stop under 60s from getting AZ, made loud announcements that the powers-that-be are recommending Pfizer. This led to the comment by the Queensland CMO that was ridiculed in the post.

So now, context over and onto that comment: QLD CMO: "No! I'm sorry if I haven't made that clear. No, I do not want under-40s to get AstraZeneca. [...] I don't want an 18-year old Queenslander dying from a clotting illness who, if they got Covid probably wouldn't die". Thezvi responds: "The idea that AZ poses a greater risk than getting Covid makes zero statistical or mathematical sense, and framing the issue that way indicates that blame avoidance and Sacrifices to the Gods of responsibility are what matters here rather than any attempt to do math on a physical world model. "

Let's do the math! EDIT: I did the math wrong. I accidentally compared the numbers of getting a blood clot with dying from covid, not dying from AZ with dying from Covid. Comparing TTS cases to hospitalisations prevented in https://www.health.gov.au/resources/publications/covid-19-vaccination-weighing-up-the-potential-benefits-against-risk-of-harm-from-covid-19-vaccine-astrazeneca is probably the best comparison, ATAGI says there's a 3% death rate from TTS. According to https://www.abc.net.au/news/2021-06-29/covid-astrazeneca-vaccine-australia-new-no-fault-indemnity-gps/100251376 under 50's who receive AZ have a risk of 3.1 in 100,000 or 0.003% of dying from a blood clot. According to https://www.acsh.org/news/2020/11/18/covid-infection-fatality-rates-sex-and-age-15163 for an 18 year old male – such as what the CMO talked about – they have a 0.003% chance of dying from a Covid infection. I’m unsure of the uncertainties, but an 18 year old Male has about the same chance of dying from a Covid infection as from an AZ blood clot. While higher age groups do become more likely to die from Covid than AZ (for a 39 year old male, around 19 times), Covid is not spreading significantly in Australia at the moment. The recommendation that under 60s don't get AZ was contingent on this fact / the assumption that Australia would continue to eradicate Covid locally.So the numbers are at least comparable: the idea that - For young (<25) Australians, the idea 'AZ poses a greater risk than getting Covid makes zero statistical or mathematical sense' is not true. In the Australian context, the CMO's comments are ballpark sensible. Finally - individuals can still choose to get AstraZeneca if they wish, given the numbers are ballpark comparable and there is a large clump of young people desperate to get a vaccine, telling them they really need to pause, look at the numbers and consider what makes sense. For young (<25) Australians, the idea 'AZ poses a greater risk than getting Covid makes zero statistical or mathematical sense' is not true, they are ball park comparable, and in the Australian context of very little covid transmission, can become correct. So - in the Australian context, the CMO's comments are ballpark sensible. Finally - individuals can still choose to get AstraZeneca if they wish, given the numbers are ballpark comparable and there is a large clump of young people desperate to get a vaccine, telling them they really need to pause, look at the numbers and consider what makes sense is reasonable.

Of course, Australia can’t have closed borders forever. Thezvi says’: "It would be one thing if they had plenty of mRNA vaccines to use instead. They don’t. " Australia currently - due to above mentioned Commonwealth screw ups - doesn't have enough mRNA, but from October should get 2.13 million mRNA doses per week. So again the question is not ‘get AZ or Covid’ but ‘get AZ now or wait 4-5 months for Pfizer during which you are very likely to not get Covid’. The cost of waiting for Pfizer is substantial (billions of dollars, families separated by closed borders, continued lockdowns) but is not measured in lives like it is in other countries.

"It would be a somewhat different thing if you couldn’t mix and match shots. You can." I agree that this mixing and matching changes things, but the numbers I quoted above for AZ blood clot fatalities were for the first dose.

"That’s all before concerns about population-level effects. " Not entirely sure what these population-level effects would be, so I can't really comment. But personally, as a 20 year old male in Sydney, who is currently unlikely to be exposed to Covid before I am vaccinated, I am planning on waiting for Pfizer due to it's better protection (I will, when I travel to the US to see family or for work (I have US citizenship) / when Australian borders open, be exposed to covid) and, yes, due to the blood clots (I initially dismissed the blood clots as ridiculous over-cautious appeasing of anti-vaxxers, but having looked at the numbers.... 0.0031% > 0.003% times the low chance of catching covid). If it becomes likely that Delta, or Covid generally, will not be contained and will become widespread in Australia, I plan on getting AZ immediately (you can currently get AZ walk-up as far as I know).

And finally: "I think we can lay to rest the hypothesis that Australia did better than other countries because it was more sane and has wiser systems for making decisions. Australia did better for other reasons, including being an island, that led to a different equilibrium. Now that we are in the vaccination phase of the pandemic, Australia is utterly failing. " Taking one comment on one issue by the CMO for one state in Australia, to justify this claim is ridiculous. I have never seen a substantial engagement with the question 'Why did Australia do better' but I think that would be an interesting question to ask, and I don’t see TheZvi or anyone else asking it. Even accepting the premise that it’s not for governance reasons, it is still worth investigating. Those reasons means Australia has a per capita death rate of 3.64 per 100,000, while the US has 184.23 per 100,000 (note it's far easier to have an accurate death rate when there's only 910 deaths, than 600,000, so the US is likely even higher). From my perspective the Australian governance has been a question of the state governments being mostly successful (and responsible for the success so far), and the Commonwealth government mostly failing (and responsible for the failed vaccine rollout). Just as how we should investigate and address all possible origins of Covid, surely we should be interested in all the possible ways countries that managed Covid well might have achieved their results, including Australia as an ideal comparison country to most OECD countries.

Personally, thezvi’s statement is as ridiculous as saying "we can lay to rest the hypothesis that America did worse than other countries because it was more insane and had worse systems for making decisions. America did poorly for other reasons, including being a highly populated dense country, that led to a different equilibrium. Now that we are in the vaccination phase of the pandemic, America is triumphantly succeeding." The American government messed up horrifically - but did also get a lot of vaccines and distributed them well. Australia did have flaws, did luck into certain factors, and did monumentally mess up vaccines - but also worked incredibly hard and made great sacrifices to save a lot of lives. Finally, given countries blocked the export of vaccines to Australia (reportedly around 3.1million) because we weren't deemed a country 'in need' - by which they mean didn't negliglently homicide its own citizens - it's annoying to watch the continued disinterest and dismissal, beyond the occasional interest piece by international media of ‘Australian city locked down over ’ (when it was through trial and error that we learnt you need a hard and fast lockdown) or some 'Australia is an island so who cares' comment.

I honestly aren't sure why there was such a difference between Australia and elsewhere, and I want to find out. I don't know why no one else seems to, but hopefully I've changed someone's mind to care too. I would be happy to write more on this if it's likely to be useful - so if you found this useful/interesting, please let me know!

I wasn't sure how to integrate this but here’s a rebuttal to the common reasons I've seen Australia’s success put down to:

  1. 'Australia is an island' While Australia is an island and this certainly helped in letting Australia achieve it's own elimination strategy, Australia's strategy scales and could have been implemented elsewhere. While Australia might be an island, the individual state's aren't and they have repeatedly contained outbreaks to a single state by closing internal borders. Those internal land border closures are hard and show regions that aren’t islands can block Covid from spreading! Australia's sixth-largest city (Gold Coast, QLD - admittedly only 680,000 people) spills across the NSW/QLD border. If you're thinking 'Neat but we could never shut the X/Y border' I can assure you I thought the same thing about the internal borders. So - I agree being an Island helpd, but it receives far too much credit.

  2. Australia has a low population density! Yes, Aus. population divided by Aus area is a small number. But Australia's population is incredibly concentrated - The three largest cities hold 50% of the population, and personally have less suburban sprawl / have slightly higher density than American cities (measuring this is complicated with city limits defined differently, but the Sydney and Melbourne train systems would be number 2 and number 3 in the US by annual ridership for a very loose indicator of city-ness).

  3. Australia just never really got Covid Melbourne's second lockdown ( https://en.wikipedia.org/wiki/COVID-19_pandemic_in_Victoria#Second_lockdown ) saw a period a bit longer than a week of around 500 cases a day in August (see graph in https://www.theguardian.com/australia-news/2020/jul/09/melbourne-shop-owners-fear-the-worst-as-second-lockdown-begins, that outbreak still represents the overwhelming majority of Australian covid cases and deaths, was caused by using half-hearted measures to avoid lockdown and taught states to use a full lockdown as soon as it's clear Covid has escaped contact tracing). This corresponds to around 700 weekly cases per million, more than most of Europe during the same period (june, August): https://en.wikipedia.org/wiki/COVID-19_pandemic_in_Europe#/media/File:Weekly_cases_of_Covid-19_in_Western_Europe_per_million.png . Unlike much of Europe that decided to end lockdown with the (relatively) ‘low’ case numbers, Melbourne drove it to elimination by enduring the longest lockdown in the world (112 days), so Australia certainly did have Covid numbers comparable to Europe, and they likely could have eliminated Covid too. Australia had already eliminated Covid before at this point, so something different about Melbourne vs Europe caused the different decision.

  4. The temperature / environment of Australia lined up more favourably. This point does have some merit - Covid started in Australian summer/autumn, giving Australia more time with the training wheels to learn, if summer reduces Covid transmission. But Covid has lasted well over a year – Australia has had Covid transmissions in both summer, winter, summer again and now winter again (🎉🎉🎉 sentences to give March 2020 me a heart attack) - so while a factor, this doesn't explain everything and doesn’t dismiss Australia’s success.

That’s all the points I could remember seeing, feel free to make more though and I’ll try to respond – some list of things that could explain the difference would be good.

Thanks for reading this far too long comment! Sorry that lot's of it came off as adoring of Australia's response – while good relative to many other places, there are real major issues to critique (particularly the thousands of standard Australians overseas). If people are interested in other stuff about Australia let me know, I'm in lockdown for at least another week so plenty of time to kill. (Ranked choice voting! The mandatory venue check in's! Sydney/Melbourne's higher public transport use! The healthcare system! Pension plans! Avoiding US-style culture war with a kind of culture-skirmish!) Finally, just as most people ignore Australia, I tend to ignore everywhere else with an exception for the US, and Europe occasionally - that's bad too! So I'd love to hear more about other people's views and thoughts on things too.

@tomiam8
Copy link
Author

tomiam8 commented Jul 3, 2021

(Cross posting my response to the response for posterity)

Thank you for your response, I'm looking forward to any future posts (either the full response or just your normal posts, they've been very informative and interesting)!

Re: statistics are for blood clots, not deaths - yep, thanks for pointing that out, can't believe I missed it; I've added a comment noting this to my original response. The original ATAGI (Aus. Technical Advisory Group on Immunisation) statement (https://www.health.gov.au/news/atagi-statement-on-revised-recommendations-on-the-use-of-covid-19-vaccine-astrazeneca-17-june-2021) itself says TTS / clots has a 3% Australian fatality rate, so all age groups have at least a few orders of magnitude different risks between AZ-clot-death and Covid-deat, some of them (eg 40-50 age group) a very large difference.

I think ATAGI does a lot of math / logical wheel cog turning... and then mostly (although not entirely) ignores the result for political reasons? The ATAGAI statement links to a calculation of expected number of blood clots vs avoided deaths and hospitalizations for each age group at three different levels of covid spread ( https://www.health.gov.au/resources/publications/covid-19-vaccination-weighing-up-the-potential-benefits-against-risk-of-harm-from-covid-19-vaccine-astrazeneca pg 9).

For every scenario except the first, hospitalizations prevented by AZ are an order of magnitude larger than TTS cases, including the level of infection that occurred in Melbourne's second wave. So, the cost-benefit calculation seems to strongly favor AZ, unless you assume Australia can avoid any major waves of covid. So, ATAGI states '[their] advice is specific to the context that there is currently no or limited community transmission in most of Australia and would be different in other countries.' which, while at least acknowledging the assumption, ignores that the timescale to vaccinate people is much larger than the timescale of a covid wave, and you won't be able to vaccinate most people in time. Finally, given increasing the age recommendation from over 50s to over 60s occurred in the political aftermath of a 52yr old woman dying, I agree that it seems the vaccine age decisions were made for the wrong reasons (also - no calculation of lockdown / border closures ongoing harm).

Personally, that means I need to re-evaluate my decision, and it turns back into is it worth waiting for Pfizer for the increased protection or not. I think my position should be at least to get a vaccine as soon as my government allows me to mix and match, or if the chance of wide-scale covid spread becomes decent (eg if cases/day > 100). I'm still thinking on if it makes sense to get vaccinated anyway.

Re: was the full-containment decision made based on math or just because:
I think the electorate forced a full-containment strategy, with politicians forced to do that or lose elections, after Australia happened to eliminate Covid in the first wave. While the math should have been done, I don't think public opinion does math for anything, and this is interestingly different from 'governmental groups should make decisions logically'. Some evidence towards that view for how much the electorate likes containment: The WA (Western Australia state, not Washington state) Premier (aka Governor) was well known for being by far the most aggressive with closing internal state borders, and the most pro-containment government generally. In early 2021 they had a state election, and for the strong pro-containment position the government was re-elected with the largest majority in any Australian election - 89.8% of lower house seats, 70% 2 party preferred support (ie 70% prefer government compared to main opposition party). The opposition party conceded/acknowledged they would lose the election 16 days before it took place! The election was called with 0.7% of the vote counted 45 mins after polls closed. So the electorate really liked containment - not just in WA but in Australia generally - and so any self-interested politician (ie all of them) did containment. The decision to contain after Australia had eliminated it was made by the electorate, not by the politicians. You could still investigate the decision to eliminate initially however.

I think if you really wanted to argue for the 'Australia is just insanely sensible and has good decision making processes', you could start picking individual decisions that were made well... and there are quite a few of those, but also there's a lot of those everywhere, and there's also plenty of bad decisions.

So I'm not really sure where this leaves us.I think some people overseas might have had a tendency to go 'Australia is doing well. Australia makes decisions well! Australia is great country all around!' because, it's another country, you don't have time to read endlessly about all it's nuances. I think I missed that that view might exist, because it is certainly a wrong one (Australia's fine but there's plenty of bad political and covid decisions, the vaccine's just one large enough that it's easily noticeable from overseas). And that's not to say that Australia does have some institutional differences that are good (Ranked choice + compulsory voting! Except for New York...). The same way that ignoring the 'good half' of Australia's decisions because of its specific lucky conditions is bad, ignoring the specific lucky conditions is also bad.

One question in particular - how forcefully were lock-down restrictions imposed in other countries (were there fines? how often were people fined?)? It's been hard for me to get an accurate feel for that, and I suspect that could be a major cause of some differences that were compounded a lot by those lucky conditions (and doesn't reflect Australia having superior decision making processes).

Thank you for the well wishes.
I'm currently estimating a 25% chance this current Delta wave will represent the hardest part of Covid for my state, potentially Australia, by far, and a 50% chance that it still won't be eliminated in 1.5 month's time and my state will move to a lockdown similar to Melbourne's 2nd. Ie, I agree with your worry that things will turn south.
Thanks again for responding, sorry for writing so much!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment