Computer modellers are still driving the COVID-19 fear wagon

Renewed calls for prolonged lockdowns to deal with the new SARS-CoV-2 mutations are wrong headed. It has been a year since emergency measures were declared. Yet, the policy response to the COVID-19 crisis has been and continues to be moved by fear that is in turn propelled by statistical models incapable of accounting for risk and of pondering consequences. SARS-CoV-2, the virus causing the disease identified as COVID-19, is a reality. It poses health risks to a well-defined segment of the Canadian population. And while SARS-CoV-2 can infect everyone, the models and responses largely pretend everyone can equally suffer and die from COVID-19. The logic of this pretense points toward lockdowns and heavy restrictions for all instead of carefully-designed protection of the vulnerable.

The initial reaction to this logic may have been reasonable in mid-March 2020, were it not for the fact that the panic-prone politicians discarded existing pandemic plans designed precisely to prevent panic. Their lockdown strategy, they argued, would bend the curve to protect the integrity of the medical system until it could be reenforced. Such reinforcement would help to save lives.

The system did not become more resilient and the infections did not stop after nearly two months in lockdown: the strategy was a failure. But such failure has not prevented the continuation of wrongheaded policies and restrictions. There are now misguided calls for #COVIDzero (or #zeroCOVID), which pretend to drive COVID-19 cases to zero by wiping out the virus and all its variants, if only we locked down hard again for another seven weeks. Like the previous failed strategy, this one also is driven by modellers and their flawed mathematical models.

In March 2020, the world seemed gripped by images from Italy, Iran and China, and one model stampeded policy-makers into various forms of lockdowns. It was the work that College of London theoretical physicist Neil Ferguson led. Their “Report 9: Impact of Non-pharmaceutical Interventions (NPIs) to Reduce COVID-19 Mortality and Healthcare Demand,” called for 510,000 deaths in Great Britain and 2.2 million in the US. Ottawa’s model version showed deaths would easily top 300,000 (but only 46,000 with a lockdown) in Canada, while Edmonton said 32,000 Albertans could die here and 1.6 million could be infected. World-wide, Ferguson and his team expected seven billion infections and 40 million deaths. None of that has happened.

Ferguson’s model raised troubling questions. First, Ferguson refused to publish the original source code and Imperial College refused a British Freedom of Information Act request. Writing in the Financial Post in June 2020, Peter St. Onge remarked that Ferguson’s code was unreliable and fragile, “giving different answers depending on the processing speed of the computer running the model.” Similarly, Chris von Csefalvay noted that the code was practically antique (13 years old), and it was written to model an influenza pandemic. Moreover, thousands of lines of code were “undocumented,” making impossible to take it apart and examine for errors — or to correct them. In his view, the code was “a tangled mess of undocumented steps.” Accordingly, von Csefalvay wondered how the British government assessed and validated the model. He concluded that only Ferguson’s reputation made the Imperial College model authoritative.

            Except that there was no reason to hold Ferguson’s work in high esteem. Ferguson’s “apocalyptic” predictions were gross exaggerations. An earlier model of his predicted 150,000 deaths from mad-cow disease in 2002 (the number of fatalities was 2,704). In 2005, Ferguson’s model predicted 200 million deaths from avian flu (455 persons died). Eventually, Ferguson resigned from the British Scientific Advisory Group for Emergencies (SAGE), not because his COVID-19 model was so inaccurate as to be worthless but because he was found violating the lockdown that he so vocally supported for everyone else by entertaining someone else’s wife in his London residence. After the fact, commentators wondered why anyone listened to Ferguson in the first place.

As if modellers were not discredited enough after Ferguson’s exaggerated predictions, CBC’s Laura Glowacki promoted Robert Smith? (the question mark is part of his name) in September 2020. Smith? is a mathematician at the University of Ottawa. He builds models for infectious diseases. As case numbers rose at the time, Smith? called for a “ruthless” and “draconian” return to a full lockdown “for a few months … [that] … could bring numbers down to zero new infections.” There was no mention of previous model failure. The country had already locked down hard for a couple of months, and close to 10,000 people had died, mostly in Ontario and Quebec where the vast majority were vulnerable people whom policy makers had vowed to protect and save. Alarmist modelling like Smith?’s pushed the second round of bullying restrictions. Smith? is not among the experts who see a tension between health and economy. In his opinion, the economy would be ruined without a full lockdown.

In November, CTV Infectious Disease Specialist Abdu Sharkawy expressed similar alarms. “We need the hammer, and that hammer needs to be applied with conviction. It needs to be applied with some assertiveness, and we need to apply the support that’s necessary from an economic point of view to the people that would suffer if that hammer is laid down,” he said partially conceding to economic harm. Earlier, with some awareness of greater harm, he said: “You can call it authoritarian, you can call it dictatorial. The fact of the matter is, there’s no more room right now for a balanced approach. It’s simply too late.” Medical experts calling for the confinement of entire populations is a new tyranny of “expert opinion” passing for scientific advice. No matter how one slices it, the forced confining of entire populations is not a medical measure.

What is worse, achieving zero infections by locking people down is impossible. If that was not clear in March 2020, it is clear now. The virus cannot be made to disappear at will, and no amount of hiding will eliminate it. But model builders keep driving up the fantasy. In Alberta, for instance, there are dreams of creating a zero-infection zone, in the same way the province is rat-free. Last October, CBC found Malgorzata Gasperowicz, an assistant researcher in the Faculty of Nursing at the University of Calgary. She studied biotechnology at Gdansk and has a doctorate in biology from Freiburg.

From a series of tweets based on her personal calculations, Robert Brown of CBC News Calgary gave her a platform for her alarmist prognostication of rising cases, warning of disaster if nothing was done “right now.”  Gasperowicz’ motivation appears in a pinned tweet from July 2020 (@GosiaGasperoPhD). It announces “we can achieve COVID-19 elimination in Canada.”  She advocated turning Canada into a new New Zealand.  Her October 29, 2020 tweet caught the attention of people looking for scary materials and warnings of impending disaster: “It’s too late for soft measures. We need strong decisive measures + $$ support for businesses and people ASAP, in order to substantially [sic] *reduce the scale* of the upcoming disaster. It takes 3-4 weeks from the shutdown date till the peak in cases and hospitalizations[,]” it read.  Note the alignment of language with Smith? and Sharkawy. Note how the tweet implies that there will be disaster regardless, but only strong medicine can reduce its scale.

Gasperowicz pointed out that the number of cases in Alberta was doubling every 16.9 days. Nothing was said about what hospitalisations or ICU cases would be.  Nothing was said about the rate of hospitalizations being a fraction of what it was in the Spring. No extrapolations were offered, except to mention that there would be more “upticks,” as Brown called it. It was all about cases. Modelling for actual illness, hospitalization rates and ICU interventions may have proven far too complex.  

While Gasperowicz predicted 2400 cases for December 5, there were 1765 cases at the peak of the case curve on December 8. The predictions were off by 27 percent, but it did not stop Calgary Herald’s Jason Herring from qualifying her projections on  December 12 with impressive precision.” Gasperowicz described the accelerating rate of cases each 2.5 weeks in an ominous-sounding calculus category no calculus professor is likely to teach: “über-exponential.” Predictions for thereafter were even worse, and they were adorned with a catchy slogan: “If we shut down on Nov 15, we will reach 3000+ daily new cases before numbers start to decrease. Either we control the virus, or the virus controls us.” Alberta did not shut down on November 15th and the predicted onslaught for early December 2020 never materialised. Robert Brown did not once ask questions about the origin of the data Gasperowicz crunched, the methodology she used, the assumptions built into the calculations, why the model stopped at calculating case numbers, or any shortcomings the calculations might have. Her “results” were all taken as Gospel. If all one wanted was to drive up fear, there was no need for additional information.

With no mention of the significant error spread in her October 2020 calculations, in January 2021 Gasperowicz tweeted new warnings about the new SARS-CoV-19 strands, which she finds “terrifying.” She particularly worries about B117, the British strand, claiming it is “60 percent” more virulent. Elsewhere, she claimed its virulence is 30-50 per cent higher, and constitutes a “super-danger.” Presenting freshly raised fears of the mutations, her model predicted that B117 will spread in Alberta above 2,000 daily cases by the third week in April, 2021, unless Jason Kenney implements another full shut down for 7 weeks. Seven weeks!  

On January 22, Gasperowicz tweeted: “#COVIDzero (aiming to eliminate all community transmission of Sars-CoV-2 as fast as possible) is the solution: 7 weeks of effort and AB can be like NZ.” And on January 27, she said: “B117 is in the community[.] Current restrictions are not enough to prevent its spread. Assuming 10 cases on Jan25 & 50% transmissibility of B117, we can have: 1000+ daily cases on Mar 23 2000+ daily cases on Mar 31[.] This model does’t [sic] include the effect of schools reopening.” We’re a month away from March 23rd, Kenney has relaxed some of the restrictions, and the number of cases is not growing. One can guess that April 23 will not likely bring “disaster,” if we go by previous predictions but Global News was sufficiently impressed to make the claim with absolute certainty that Gasperowicz’ modelled projection “shows how a seven week lockdown will drop new COVID-19 cases to zero in Alberta.”

Gasperowicz has not described what the lockdown she recommends for seven weeks looks like, but given her intention to eliminate the virus from circulation, one can assume that it includes stringent stay-at-home directives, shutting down most economic activity and government services.  She also wishes to stop all travel, and published a co-authored column in the Calgary Herald in February 2021 arguing that if Alberta can keep rats out of its borders, it could certainly keep the coronavirus out. Although the column mentioned New Zealand as a model jurisdiction that had kept virus-free, they do not mention re-infecting flare ups. New Zealand had declared itself victorious over SARS-CoV-2 twice by mid-February 2021, only to call for another one in Auckland for three days. There was no thoughtful consideration of the spin-off and collateral damage of stopping and starting time and time again every time cases pop up. By the end of February 2021, with no explanation for the change in the face of the “super danger,” Gasperowicz’s recommendation for total confinement in Alberta, reportedly, was now only 6 weeks

According to Global’s Jacqueline Wilson, Gasperowicz says “all non-essential businesses would need to close and all international and inter-provincial travellers would have to quarantine.” That most jurisdictions in Canada, including Alberta, have made a monumental mess in imposing what is “essential” for everyone was not part of the discussion.

A glance at the New Zealand case charts shows that the country has not been at zero cases but for a few consecutive days here and there.  Given that their standard reaction to reappearing cases (“outbreaks”) is locking down, chances are they will have more lockdowns. Having recurrent lockdowns does not make lockdowns a success. We have seen the same with PEI in early March, 2021. As Sweden’s ranking epidemiologist, Anders Tegnell, puts it: “fighting Covid-19 is a long-term undertaking, meaning temporary lockdowns will ultimately backfire. …once they’re lifted, infection rates will again rise.” And vaccines will not get us there by incantation either. Even with the vaccines, serious scientists do not expect the elimination of the virus. “Even if you vaccinate, you’ve still got a fairly large number of susceptible people there,” says Michael Head, a senior research fellow in global health at the University of Southampton. “So, we will still see outbreaks happen. Viruses simply aren’t rats. And when cases keep popping up, #COVIDzero is a misnomer or a deceitful expression, if zero means zero.

Alberta’s economy is plugged into the world’s and depends on its ties to the rest of the world, whether in agri-foods, tourism, energy or mining.  It could not easily close its borders, airways, highways and railways, much less for another 6-7 weeks without returning to the enormous damage to human lives and to the economy already caused by the first and second rounds of confinement. Alberta is no island, making the virus here much more difficult to contain. Not that the issue is geography. The reason PEI has had so few cases is because not many people go to or through PEI. Conversely, Manhattan has been one of the most disastrous COVID-19 areas in the world.  The difference is many people want to go to, or need to go through, Manhattan.   

Although PEI and New Zealand are hailed as lockdown successes, they demonstrate the opposite point: it is impossible to hide from the virus, let alone make it disappear. #COVIDzero is a well-intended but irresponsible fantasy posing as medical advice that, if instituted, will again bring greater social and economic harm. So, let’s say no to #COVIDzero and to the fear it inadvertently peddles with the fantasy of virus elimination.

Meanwhile, we can hope that media learn to treat #COVIDzero experts the same way they treat those claiming the virus doesn’t exist.  After all, denying the existence of the virus seems as detached from reality as it is claiming that it will disappear if we hide from it.

%d bloggers like this: