Our Government assures us that they are being led by the science in their strategy against the COVID-19 pandemic. But what is the science and are we actually being misled?
Whilst at medical school I attended the anti-war rally in London, along with a million others, to protest against the UK’s build up to the Iraq war. We were told by our Government, based on intelligence, that Saddam Hussein had Weapons of Mass Destruction and we believed it. It eventually turned out to be wrong but it had cost the lives of hundreds of thousands of Iraqi people and thousands of allied servicemen and women.
I worry that a similar situation is playing out today with prediction models dictating policy more than reality is. Models err more on the side of hypothesis than genuine scientific observation.
Models from Imperial in the UK, that the Government used to justify the lockdown, based their initial predictions of the COVID-19 (caused by SARS CoV-2) peak on small datasets from China and Italy1. Because the natural rise and fall (the curve) of the virus was unknown, the group used a technique called the “Counterfactual” to predict how the virus would ‘curve’ had no intervention occurred.
The modelling was very complex. Complexity, however, doesn’t necessarily equate to superiority; often these models are complex because the data they have to work with is small and the assumptions such as infection rate, and modality of spread are inconclusive2. As a clinician and academic, I have both utilised and designed models, respectively, to help find the best treatment strategy for patients. When designing a model it is important to have a dataset to formulate the model and then a separate (external) dataset to validate the model. The latter was missing here.
We are led to believe that the current reduction in predicted deaths is due to the “flattening of the curve” from lockdown measures. “Flattening” refers to reducing the expected peak from those counterfactual models. However, this is misleading as the counterfactual is a hypothetical curve that we cannot validate with direct observations. It is akin to saying that if we had not invaded Iraq, many more lives would have been lost. There’s no control group.
We could compare our population-calibrated outcomes, in due course, to Sweden’s who have, sensibly, refrained from enforced draconian measures, as a control group. Currently, the deaths per million of the population in Sweden is far lower than the UK whilst Sweden has not witnessed the significant negative impacts on finance or employment in their country. Furthermore, if Sweden achieves herd immunity3, it will unlikely have subsequent peaks in the virus unlike the UK which may have further peaks once lockdown measures are released.
I understand completely the need to reduce the burden on an overstretched NHS but I think that this was largely achieved by cancelling operations and reducing non-COVID admissions and not necessarily from shelter-in-place. Shelter-in-place strategies have no proven effectiveness in the literature for diffusely infected nations as opposed to quarantine which is highly effective in reducing the spread of infectious diseases in localised outbreaks4.
I believe that we should have followed the Government’s original plan of shielding the vulnerable and allowing schools and businesses to stay open. Systemic reviews have shown closing schools have little impact on the spread of certain coronaviruses5.
Governments are over-relying on the World Health Organisation (WHO) to provide the correct advice. It was the WHO that originally relayed that there was no human-to-human transmission of SAR COV-2 in China, which meant nations such as the UK and USA may not have been prepared for the outbreak in time6. The WHO have also stated that evidence of antibodies in patients infected with SARS COV-2 may not necessarily convey immunity7. Such an implication is far reaching as the basis of vaccines relies entirely on the body building an immune response to the viral antigens.
Although vaccines are important, successful respiratory virus vaccines in particular are notoriously difficult to manufacture. As the majority of infected cases do not need hospitalisation and the vast majority survive, the UK would do better, whilst continuing social isolation protocols for mildly infected cases, to concentrate their efforts in finding drugs that reduce the complications and mortality from SARS COV-2 in severe cases.
Many clinical trials are underway to evaluate such drugs to decrease disease progression. Antivirals such as Remdesivir8 offer hope and the increase in pharmaceutical research both at a government and private level will undoubtably yield more therapeutics to reduce the viral load and manage COVID-19 complications (respiratory and multi-organ failure). This together with widespread testing (particularly serological antibody testing) will allow more confidence in lifting lockdown measures.
As a society, we need to accept that there is nothing without risk. The danger in becoming so risk averse as to shut everything down to save lives, is that we fail to live our lives. Our evolutionary history has taught us that, as a population, we will learn to adapt to COVID-19 naturally or through scientific intuition rather than fear. I certainly don’t want my future generations to live in a box for the rest of their lives and miss the outside world I enjoyed.
The beauty of Science is that it is neither dogmatic nor permanent. It can be tested, dissected, rebuilt and above all, challenged. Models are an invaluable tool that shouldn’t be used in isolation dogmatically but instead adapt to reality, and be part of a larger breath of knowledge and expertise to determine how best to we can wake up from this nightmare.
Dr P Ariyaratnam BM BSc (Hons) MRCS MDNIHR Clinical Lecturer & Speciality Registrar in Cardiothoracic Surgery