Determining cause of death in the age of covid-19

Simon Thornley

28 Feb 2021

New Zealand descends into another lockdown abyss just as I am now questioning the very foundations of the covid story.

Recently, I have pointed to evidence that covid was around in Europe before Wuhan. Since we were apparently living with covid in Europe without excess mortality or catastrophe, this seems at face value to suggest that we had lived with the virus and can do so again.

Another matter is that we have shifted how we define virus related factors, and this is causing us to believe SARS-CoV-2 is more serious than it is. Definitions are a cornerstone of epidemiology and the collection of scientific data. Death seems to be a clear-cut event in a person’s life but determining the cause of death is a surprisingly difficult process.

As Dennis De Nuto so beautifully illustrated in the Ocker flick, “The Castle”, definitions matter. Responding to the judge with “it’s the vibe, your honour” when alleging a constitutional law breach doesn’t quite cut it. While there is almost always uncertainty about causes of death and classification, this uncertainty has been stretched to breaking point in New Zealand’s covid-19 saga. The most recent death is illustrative.

We learned that an individual had been in quarantine. They were then transferred to North Shore Hospital for a serious non-covid illness and later tested positive for covid. The patient later died. A few days later, the death was considered another official NZ covid death. On the face of it, without further information, this does not indicate to me a death caused by infection with SARS-CoV-2. At no point in the article are we told that the individual had the disease that the virus is alleged to cause: a lung infection. When the reporter quizzed the Director General, Dr Ashley Bloomfield, for justification of the Ministry’s classification the response was revealing: ‘…we have been very inclusive in our approach of categorising deaths as covid-19-related…”. He went on to explain, “You might recall when we had a number of deaths last year sadly related to aged residential care. A number of those people had actually not been swabbed because of the nature of their conditions but they were categorised as probable cases because of their symptoms…”

We have learned two important things from these statements. In the first case, if you want to classify as a covid death you do not need evidence of a lung infection. In the second, when referring to the rest home deaths, you do not even need a positive PCR test. There was no further evidence related to autopsy or other clinical findings or investigations.

The next revelation came in how this decision making was justified. Bloomfield’s response was “Most countries are doing this, for example in the UK they categorise everyone who dies within 28 days of being hospitalised with covid-19 as … a COVID-19-related death.” Essentially, everyone else is doing it, so… why not?

It would be weird and totally unhelpful if this was the approach to any other disease. It’s even worse for Covid because we are now in Auckland’s fourth lockdown – justified based on saving lives from infection. It is now questionable that New Zealand’s covid-19 deaths would have lived longer without the infection. The absence of the virus clearly would be unlikely to have saved the latest death who was hospitalized with a “serious non-covid illness” and then tested positive. Not without further supporting information, which if present, has been withheld from the public domain.

Since it is also now stated that not all deaths tested positive, it is hard to be certain that the absence of the virus would have saved other ‘apparent’ covid deaths. Indeed, according to a June 2020 OIA request, five of the 21 deaths (at the time) reported tested either negative or had not been tested. This is important, since covid-19 is not a disease with specific signs and symptoms, it is unclear that the absence of the virus would have saved these negative or untested cases.

This now leaves the covid deaths who tested positive. We are simply not given enough clinical information to know whether these lives would be saved without the virus on board. What we do know is that 16/22 deaths at June 2020 occurred in rest home residents, and that 8/14 deaths until April occurred in residents of a specialized dementia unit. We also know that the age distribution of these deaths is no different from background. This is further evidence that the absence of SARS-CoV-2 would not have altered the survival of these people.

In Italy, the place that scared many of us into thinking the worst about covid-19, it has now been proved that covid deaths were systematically exaggerated. The same is likely in the United States.

The one country that bucks the trend and uses a strict definition of a covid-19 death is Singapore. It also happens to have 29 deaths from 59,925 cases at the time of writing, with a case-fatality ratio of 0.05%, less than for seasonal influenza. It may be definitions, rather than lockdown protocols, that means the virus has had little effect there. Rather ironically, lockdown enthusiasts in New Zealand have championed Singapore for its societal restrictions, but they have not championed its use of covid-related definitions.

It looks increasingly as if covid-19 is a kind of chimera, largely created by our own modern fears.

This is bizarre, as Auckland moves into its next lockdown, estimated to cost our economy half a billion dollars per week, and as the queues in food banks are expected to grow, with the lives of our poorest communities most affected.

We must scrutinise not only whether our strategy hell bent on elimination of this virus is worthwhile, but whether it’s even based on reality.

What is the end game for New Zealand with covid-19?

Simon Thornley

22/2/2021

A New Zealand household case of the UK variant triggered another short lockdown in New Zealand. This will have prompted many of us to wonder “when will covid end?” The answer requires considering that we won’t eliminate the virus, and nor do we need to.

I was recently invited by a Canadian group to debate the motion that all countries should aim for “zero covid”. My opponent, health economist, Dr Stephen Duckett waxed lyrical about the virtues of the zero covid, since he was at liberty to attend the Australian Open tennis tournament.

He talked about the relative freedom of elimination compared to the UK and US which were enduring constant restrictions and high rates of infection. The podcast was freshly posted when Victoria went back into lockdown, as did the chances of Stephen seeing Novac Djokovic play live. I briefly felt self-satisfied that it wasn’t New Zealand, but within days it was. Auckland was back into level 3.

People may rightly hope that vaccines will be the answer. After two doses, the Pfizer vaccine was  claimed to reduce infection rates by 95%, with 8/21,720 cases in the vaccinated and 162/21,728 in the placebo. The rate at which these vaccines has been developed is a testament to human determination and skill.

However, we must also ask, why do most vaccines take ten years to develop and have we cut any corners? It is now clear that we simply do not know what the long-term benefits and safety profile of the vaccine will be. In South Africa, a trial of the once claimed 70% effective Oxford-AstraZeneca vaccine was rendered ineffective due to the emergence of the new variant with 19/748 in the vaccinated group infected, compared to 20/714 in the placebo.

Unanswered questions now are how effective the Pfizer vaccine is in the elderly, since in those aged 75 years or more there was only a total of five cases, with all occurring in the placebo group. While this is promising, it would be useful to have more definitive evidence about the elderly, since they are the target population for preventing fatalities from covid. There is little evidence on the ability of the vaccine to prevent their hospitalisation and death, which are surely the events we are hoping to prevent.

Since respiratory viruses frequently mutate, vaccine efficacy is unlikely to persist. We have seen this already with the Oxford vaccine, and the covid-19 end game remains unclear. As we’ve experienced, long periods without a community case are inevitably punctuated by community spread from the virus. As others have stated, it is a “tricky virus”. Indeed SARS-CoV-2 has been found in cats and dogs. The latest case indicates the limits of human understanding of transmission as we scratch our head about the source.

With all this uncertainty, unexpected good news has emerged, but you are unlikely to read about it in the newspaper. Hospital mortality in New York has dropped by a staggering 70%, comparing rates in March 2020 to August of the same year. While this suggests the introduction of new treatments, it is actually likely to be the opposite, in that less aggressive use of ventilators were likely to have improved survival in hard hit areas. Doctors in Toronto, like the public, were gripped with fear, initially quick to reach for the endotracheal tube, and have now realized the virus is not as deadly as they feared and become less trigger happy in reaching for the ventilator. The same pattern is replicated in both Italy and the UK. The finding of widespread covid-19 positive antibodies in Italy in September 2019 and a positive waste water test in Barcelona in March 2019, support the conclusion that it has been the response to the virus that has led to spikes in fatality.

What does this mean? We have all been gripped with fear from the virus. The government, as well as some doctors, have assumed that extreme caution will inevitably guide the best response. I believe this is well intentioned. However, this has led to preventable fatalities in some intensive care units, and to extreme public policy responses, that have prioritized the battle with this virus over all other health concerns.

Intensive care doctors had to learn that being cautious did not necessarily save lives, and instead led to harm. We need to learn the same about lockdowns and vaccines instead of naturally acquired immunity.

The concerns of livelihoods from small businesses, fiscal responsibility and mounting government debt have fallen into the background. Civil liberties and our way of life have faced the greatest restrictions since World War 2. Like doctors, our public policy makers need to dial back the fear, and contain the unintended consequences created by unnecessary, myopic, and destructive goals, such as the pursuit of national elimination of SARS-CoV-2.

Michael Baker’s mysterious data

The architect of NZ’s elimination strategy for Covid19, Michael Baker, criticised former “Bachelor” star Naz Khanjani for implying her mild experience of Covid19 was the typical experience.

He said her comments were “dangerous” misinformation, and a “fallacy”. Yet a reality TV star show had it right – her experience is very, very, typical.

So we have some questions for Prof. Baker from the strange statistics he gave to the media.

  • If his 1% mortality figure is correct, then why is the WHO publishing statistics indicating that the median is 5 times lower?

https://www.who.int/bulletin/online_first/BLT.20.265892.pdf

Is he saying the WHO and Ioannidis is wrong?

The 1% figure provided by Baker led Prof. Rod Jackson to predict 60,000 deaths in Sweden.

https://www.nzherald.co.nz/nz/rod-jackson-has-sweden-made-a-fatal-mistake-with-covid-19-coronavirus/RUR7CV376CXFC4Q2M7J7YAYW6M/

Currently, there are about 9,300 deaths in Sweden, again indicating that the figure is grossly inflated.

  • If Baker is right and the virus is 20x more deadly than the flu, how can it be that Denmark, Estonia, Finland, Germany, Malta, Norway, and Northern Ireland have had no increase in overall mortality, despite widespread exposure to the virus?

https://www.euromomo.eu/graphs-and-maps/

Also, if the virus is 20x more deadly than the flu, then why are the observed deaths in New Zealand occurring with an age profile that is the same as natural death occurring in past years? Surely, if the virus were so deadly, it would shorten lives, as the 1918 flu epidemic did?

https://www.covidplanb.co.nz/our-posts/is-new-zealands-covid-19-story-past-its-use-by-date/

  • Does Baker consider that part of the chaos happening overseas is due to the policies enacted, and not the virus itself?

https://collateralglobal.org/

Debate of 2020: Thornley vs Baker on Covid19 response

Finally, the debate New Zealand should have had in March 2020. Simon Thornley and Michael Baker discuss using elimination and lockdowns against Covid19.

We correct NZHerald story about NZ Journal of Primary Health Care

Our letter to the NZHerald regarding the recent story on our NZ Journal of Primary Health Care (https://www.publish.csiro.au/hc/Fulltext/HC20132).
Vaccine caution
We recently wrote a scientific article in a leading medical journal which featured prominently in a Herald news report.
Our article was not “rebuked” by the scientific adviser of the Ministry of Health. As part of the usual scholarly review process, the editor of the journal asked the ministry for comment. When one group raises questions about the work done by another, the latter is always given an opportunity to respond.
The article and the response are available for any reader.
The article seriously mischaracterises our views about Covid-19 vaccines. Our letter urges caution about the speed of the rollout of the Covid-19 vaccine, since historic vaccines for respiratory viruses, such as swine flu, have been associated with adverse effects. The associate editor of the British Medical Journal, Peter Doshi, has raised questions about the efficacy of current vaccines. None of them yet have evidence of success in reducing severe infection (hospital admission, ICU, or death) or interrupting transmission (person-to-person spread). At least, the trials could not test for these, given the compressed time-frame.
Developing and distributing these vaccines to seven billion people of the world is a non-trivial task.
This does not make the vaccines useless but does raise legitimate questions about basing our border policy on the effectiveness and wide availability of vaccines.
Finally, we caution against the types of ad hominem attacks reflected in this article. This is not the way to undertake either good science or good policy.
Simon Thornley, Ananish Chaudhuri, Gerhard Sundborn, Grant Schofield, Auckland.

The fallacy of Covid19 ‘fact checking’

Covid Plan B was ‘fact checked’ as ‘misleading’ for publishing on Facebook our article which used the existing conventional standard of statistical interpretation to find that a Danish study on mask wearing meant there was no significant benefit to wearing a mask against Covid19.

This so-called ‘fact check’ used a non-conventional approach which would mean that any study showing no significant effect of the studied intervention would mean the intervention does work.

This is clearly astounding. It reverses decades of scientific interpretation. It defies common-sense. But that is what ‘fact checking’ has become in the Covid19 era: a means of upholding the establishment policy position (using non-scientist media staffers).

It is not a means of checking facts. It is a means of denying them.

We outlined this deeply worrying development in an article in the British Medical Journal. Danish mask study: masks, media, fact checkers, and the interpretation of scientific evidence | The BMJ

Should we abandon convention altogether? If we did, we may eventually promote ineffective treatments. As an example, electrostimulation, laser therapy, and acupuncture are not generally thought to improve smoking cessation success, yet several promising pooled effects were calculated in a meta-analysis, although the majority were not“statistically significant.”

The tone of the“fact checking”piece that apparently supports mass masking as having a“small protective effect”over a conventional interpretation  as“misleading”turns usual scientific practice on its head. Pointingto observational evidence to contradict trial results is another subversion of usual epidemiological practice. While this may seem trivial, it is a subtle distortion of results and the politicisation of evidence in the covid-19 era.

Full PDF here: bmj.m4919.full

 

NZ’s solo effort on elimination

A short piece from us published in NZ Journal of Primary Health Care.

https://www.publish.csiro.au/hc/Fulltext/HC20132

How many more lockdowns, billions of dollars and social and health harm is an acceptable price to pay before this misguided and expensive strategy is abandoned? We implore Prime Minister Jacinda Ardern, Director-General of Health Dr Ashley Bloomfield, and fellow health advisors to reflect on the points raised in this paper and to abandon elimination as a strategy and the use of lockdowns. We believe that future policy should return to the initial approach that was taken. That is to reduce transmission of COVID-19 through reasonable use of infection control, to maintain capacity in our hospitals and intensive care, while focusing public health and infection control efforts to protect the frail and elderly of our community.

No mortality difference between Sweden and Norway, but Norway result came at huge cost

An important study (preprint at time of this post) shows similar mortality rates in Sweden and Norway despite different national responses to the Covid19 virus. But critically, Sweden’s mortality outcome came at a much cheaper economic cost.
Despite an order of magnitude difference in case-fatality rates in Sweden (higher) compared to Norway, the two countries had very similar overall mortality profiles.
There was a big difference though in national costs. Norway’s more restrictive policies resulting in public spending 2.6-fold more than Sweden (Norway: 4,176 Euros per person & Sweden 1,580 per person) during the epidemic.
It also reveals that the spike in mortality in Sweden which had caused consternation, and some unfortunate glee among pro-lockdown observers, was most likely due to ‘displaced mortality’ from low mortality in earlier seasons. Norway had no overall mortality spike.

Lead vaccines: answers needed

As lead vaccines announce good results and intentions to register for fast-tracked safety authorisation in EU and the US, immunologist Byram Bridle, reminds us of questions that they will need to answer:

Dr. Byram Bridle, PhD, Associate Professor of Viral Immunology, University of Guelph, Ontario, Canada

  1. How many of the total study subjects are being reported on? Partial results can range from being representative of the entire data set to being biased.
  2. How many study subjects had detectable immune responses and what was the magnitude?
  3. Were there antibody responses in the respiratory tract, which is where SARS-CoV-2 infects, and did these antibodies efficiently neutralize the virus?
  4. Were SARS-CoV-2-specific T cells induced? A balanced anti-viral response should include antibodies to prevent infection and T cells to kill viruses that get past the antibody barrier.
  5. Did the immune responses have a ‘Th1’ or ‘Th2’ bias? The former type of immune response is optimal against viruses, the latter is usually sub-optimal and sometimes even dangerous in the context of respiratory viral infections.
  6. Did the vaccine confer long-term immunological memory? A prophylactic vaccine may be useless without this. If immunological memory is short-lived, vaccinated individuals could become susceptible to infection before enough people are immunized to achieve ‘herd immunity’. Another term for this is ‘duration of immunity’ (i.e. how long does immunological protection last?)
  7. How did the vaccine perform in senescent animals and/or elderly humans? Those most in need of protection against COVID-19 are the elderly and immunocompromised.
  8. How was safety assessed and what were the results?
  9. Have the results been published for review by other scientists? If not, when? It is recommended to publish in open-access journals, which are available to the public. Comments merely reflect opinions unless there are validated data to back them up.
  10. Related to #6 above, what is the plan to manufacture and roll-out enough vaccine doses to achieve herd immunity in any given country? What is the realistic timeline for this? If >1 year, there is no way to know if COVID-19 vaccines will confer protection for this long because they didn’t exist one year ago. The development of every historical vaccine took >4 years, so there were years-worth of ‘duration of immunity’ data. For example, optimistic projections suggest ~1 billion doses might be possible by the end of 2021, but for two-dose regimens, that means only 500 million people could potentially be vaccinated in just over one year. With the global population at 7.8 billion people, that represents only 6% of the world’s population. The lowest estimate to achieve herd immunity is 60%.
  11. How will equitable distribution of vaccines be accomplished? For example, pre-order waiting lists seem to be dominated by developed countries; are any developing countries on these waiting lists?
  12. What is the cost of a full vaccine regimen going to be? Is it affordable for developing countries? Some epidemiologists have predicted that efficient roll-out of vaccines to developing countries will require the price to be <$6 US.
  13. Storage conditions for vaccines could impact distribution and market competitiveness. What data are available to support the claimed storage conditions? For example, the default storage temperature for RNA in research laboratories is -80o This is based on a plethora of scientific evidence that RNA is more stable at this temperature compared to -20oC.

Danish mask study result; no statistical difference from not wearing one

Only two days after the NZ government announced mandatory mask wearing rules the much awaited Danish mask study was published, and it is conclusive; masks give no statistically significant protection from Covid19.

Here’s the study: https://www.acpjournals.org/doi/10.7326/M20-6817

Results:

A total of 3030 participants were randomly assigned to the recommendation to wear masks, and 2994 were assigned to control; 4862 completed the study. Infection with SARS-CoV-2 occurred in 42 participants recommended masks (1.8%) and 53 control participants (2.1%). The between-group difference was −0.3 percentage point (95% CI, −1.2 to 0.4 percentage point; P = 0.38) (odds ratio, 0.82 [CI, 0.54 to 1.23]; P = 0.33). Multiple imputation accounting for loss to follow-up yielded similar results. Although the difference observed was not statistically significant, the 95% CIs are compatible with a 46% reduction to a 23% increase in infection.

Conclusion:

The recommendation to wear surgical masks to supplement other public health measures did not reduce the SARS-CoV-2 infection rate among wearers by more than 50% in a community with modest infection rates, some degree of social distancing, and uncommon general mask use. The data were compatible with lesser degrees of self-protection.

Below is our Euler diagram summarising the findings of the Danish mask trial.

The small difference in event proportions, with a slightly higher proportion in the control group, was not distinguishable from a chance finding (“not statistically significant” is the boffin term).

This is probably the best evidence we have up to now, which is disappointing for advocates of mask use to prevent covid-19 infection. This evidence is consistent with previous trials which found no effect in trials designed to assess the effect of masks to prevent the community transmission of influenza.