Dear Sir, I now send you an essay which I have found among the papers of our deceased friend Mr. Bayes, and which, in my opinion, has great merit, and well deserves to be preserved – Letter to the Royal Society in 1764 by Richard Price
Faced with the current COVID-19 pandemic, like many other people I have been following the ins and outs of the developing scientific debate. It has to be faced that we are literally blundering about in the dark – there are so many unknowns about who gets it, when and how. This makes 'following the science' all but impossible – any scientist worth their salt is going to make it clear that there are so many imponderables involved that the trail disappears into obscurity. It's particularly interesting when one of the greatest scientific thinkers of our time, a man described by
Science in 2016 as the 'most influential brain scientist of the modern era' comes up with a new model of the coronavirus plague by reverse engineering the data from studies all over the world.
Karl Friston, a professor of neuroscience at University College London (UCL), has championed the use of Bayesian Statistics, based on the theories of an obscure English theologian and mathematician, Thomas Bayes (1702-1761. Bayes was a non-conformist minister in Tunbridge Wells, who was educated in London and later attended Edinburgh University to study theology and logic. His theorem,
An Essay Toward Solving a Problem in the Doctrine of Chances, was published after his death by a friend who went through his papers and sent it to the Royal Society in London.
Friston recognised the value of Bayes' approach and developed the process of Dynamic Causal Modelling (DCM), which has applications in all fields of science, neuroscience, behavioural, computing, medicine, psychiatry and now epidemiology. DCM applied to the pandemic works like this: based on how the virus behaves in terms of infected cases and deaths, you can create different models of how it spreads, run the models from the beginning of the outbreak and see what model fits the subsequent data best in as many different environments, countries and continents, and government interventions as possible.
Karl is a member of the alternative SAGE group who work independently of the Government, so there is no political influence. He recently made an appearance on
Lockdown TV in a 45-minute interview about the work of this group and its international links using DCM. The conclusions are stunning and I urge you to watch it – it is a fascinating piece of science and a tour de force in scientific communication and humility. The science of parsimony (Occam's razor) proposes that if there is an overarching model that can account for all the different ways the virus has spread in different environments, then this must be closer to the truth than any other theory. DCM is the ultimate tool of parsimony.
So Karl asks – why is it that Germany has such a low level of deaths, but the UK and Sweden's is so high? How come Sweden, whose outbreak closely resembles the UK, which has not had any lockdown, has the same level of deaths as the UK who have had a strict lockdown? His conclusion, using DCM, is staggering: the most effective model has 80% of the population in some way not susceptible to the illness. Quite why, he can only theorise, but this explains best the different trajectories of the outbreak in its various settings. The reasons may one day be clarified, possibly it may be Vitamin D deficiency, or the genetics of the immunology of different racial origins, or previous infections with other coronaviruses (the common cold is caused by a group of four coronaviruses which at one time were possibly originally equally lethal as COVID-19 but over time have become minor illnesses) or a combination of some or all of these along with others we cannot yet understand.
Most astounding is that he thinks that the approach of different nations may have made little difference. The model does appear to make some sense: given the lack of PPE and the level of exposure to patients and clients among young doctors and nurses and community care staff. it is amazing that more frontline workers have not contracted the virus and died. The lack of susceptibility would explain this.
The good news is that if only one in five people are susceptible, then the pandemic may be close to burning itself out. The statistics from New York about prevalence and deaths suggest that most of those susceptible will already have had it. Karl calculated on 10 May that by 12 June there would be no need to do distancing. He freely admits he was too optimistic, but the death rate is certainly dropping faster than anyone originally thought. At the time of writing this piece, there have been no deaths in Scotland's hospitals for two days – something well worth celebrating. I believe helped by a respected Government and hence good public engagement with the message.
So while we all try to digest this, I would recommend watching Karl at work in his interview. He is not just an amazing scientist but an incredible communicator. He often declares: 'I'm speaking beyond my expertise'. How many politicians have said that? He was nominated for a Nobel Prize last year and I think he will definitely get one sooner rather than later.
Finally, let me say, although the statistics point strongly to some people not being susceptible, we do not know who is or is not vulnerable. What we do know is that it can be transmitted easily between susceptible people. Given that this is a potentially fatal disease particularly for the elderly, for men, for BAME, for those with high blood pressure, diabetes and obesity, and is not a pleasant way to go, I would strongly suggest that we all continue to follow the advice of the Scottish Government, which will save the lives of a lot of susceptible older or vulnerable people who are shielding.
Many will still need to shield for some time to come and we need to have effective track and trace (thank heavens in Scotland this will be sensibly directed by the NHS) but as most of the susceptible will have already had the virus, its prevalence in society may rapidly diminish, which will vastly reduce the chance of picking it up. All this Karl clearly explains (along with his astute interviewer). Life, to a certain extent, can possibly return to normal for the majority, albeit with masks and distancing (maybe reduced to one metre as in the rest of Europe). We will have more time to work out the factors that cause and protect against the virus, and perhaps find a way to harness that knowledge and to test out effective vaccines or other treatments – while the virus may just decline in potency to become just another of those common colds that does the rounds each winter.
After listening to Karl, and with two days free of deaths, might we be reaching the beginning of the endgame for COVID-19? He might be wrong, but he is a cautious scientist. I have known him for about six years and I am cautiously optimistic now.
Karl's Bayesian methods applied to human beliefs and behaviour have resonated with me for some years – it can explain how adverse childhood experiences can have an enduring effect on distress and recovery. If your experiences have led you to expect that everything is your fault, that you are, in some way, a flawed person, then this is the script that will dominate your life – sabotaging your confidence and positive feelings. This fits with the outcomes of our
own research (at Edinburgh University and the Foundation for Positive Mental Health) on memory structures, and has led to our development of a programme that can abolish distress by integrating the bad event into a detoxified personal narrative.
'Positive Mental Training' is an evidence-based audio-based body/mind programme found in the 'Feeling Good: Positive Mindset' app, based on a Swedish Olympic coaching programme. It is approved by NHS Digital for recovery from anxiety, depression and burnout, and is a service provision in NHS Lothian and Borders since 2018. The Scottish Government Wellbeing Fund is supporting our charity, The Foundation for Positive Mental Health, to make this app available free across all Scotland during the COVID-19 crisis to support key workers, and we are extending this offer to the general public. Click here for more information. Please note that this is not a replacement for medical treatment.
Dr Alastair Dobbin is an Honorary Fellow at the College of Medicine and Veterinary Medicine, Edinburgh University. He is a Director of the Foundation for Positive Mental Health
