Thinking Rationally About Coronavirus COVID-19 (guest post by Alex Broadbent)
The following is a guest post* by Alex Broadbent, Dean of Faculty of the Humanities, Professor of Philosophy, and Director of Institute for the Future of Knowledge at the University of Johannesburg. He is the author of many works, including Philosophy of Medicine and Philosophy of Epidemiology, and co-editor of a forthcoming volume on the philosophy of public health.
Thinking Rationally About Coronavirus COVID-19
by Alex Broadbent
This article is meant to assist in thinking rationally about the current outbreak of COVID-19, especially through clarifying conceptual matters and structuring a cost-benefit analysis. This requires a factual summary, which is provided in good faith. However, the article should NOT be treated as an authoritative guide to any factual matter relating to COVID-19 or any other disease, nor to any medical or other technical or professional matter. The author has no formal medical or epidemiological training.
Coronaviruses are ubiquitous (we find them wherever we look), and cause 5-10% of respiratory infections globally, rising to about one third during epidemics. There is no specific treatment for any coronaviruses beyond supportive care. As these figures imply, infections are typically not life-threatening. We have all had them; they are one of several viral infections aside from influenza that often inaccurately get called “flu”. However, some strains are more dangerous, such as Severe Acute Respiratory Syndrome (SARS, also originating in China) and Middle East Respiratory Syndrome (MERS), and can lead to pneumonia and contribute to serious pulmonary conditions.1,2
In December 2019 a clustering of pneumonia in Wuhan, a city of 11m people in Hubei Province (pop. 56m), China, led to the identification of a novel coronavirus, which the World Health Organisation labelled COVID-19 in February 2020. Early patients had visited a live animal and fish market, and the two closest relatives of COVID-19 are found in bats, suggesting zoonotic origin (in common with many coronaviruses and, ultimately, the majority of our infectious diseases, which mostly originate in agricultural and domestic animals3). However, the outbreak spread, showing that community transmission (human to human) was occurring.
Here is a summary of the latest WHO situation report.4
|Region||Confirmed cases||Of which, new||Deaths||Of which, new|
|Rest of world||28,674||3,949||686||202|
New cases in China are low, and decreasing; the epidemic peaked between late January and early February. The table above shows that outside China, 14% of all cases and a whopping 29% of deaths have been reported in the last 24 hours at time of writing. (A new report came out as I was making final revisions; yesterday’s report showed only 71 new deaths outside China. As I write, then, things appear to be moving fast. However, all this information becomes dated very quickly, of course; by the time you read this the picture may be very different again.
Mechanism of community transmission is not known, but is thought to be by respiratory droplets through sneezing, coughing, hand contact, and surfaces (where it probably survives around 24-48 hours), with airborne transmission also a possibility.1,5 The clinical course of COVID-19 is not well studied yet,6 and understanding crucial facts, such as case-fatality rate, requires a conceptual understanding of rates, which is one of the topics discussed below.
What explains the current panic, and is it justified? In what follows, I will first explain the panic in terms of two factors, being ignorance, and the importance of various rates, which is whose epidemiological importance is not always understood. Second, I will set out some considerations that would structure a rational cost-benefit analysis of contemplated public health interventions. A checklist of “viewpoints” for assessing a contemplated course of action are offered. Not all of these considerations are palatable, but they must underlie any rational justification of public health action or inaction. Finally, I will outline a challenge for many public health interventions, applying in this case to the problem of panic-buying of various items.
Explaining the reaction: ignorance and rates
Since we all die eventually, the threat posed by disease to human life does not just depend on how certain it is that a sick person will die. It also depends on the rate at which a disease kills people, which in turn depends on how infectious it is and how long non-survivors last. A rate is a measure of a change in one quantity in units of another7, and rates are essential to understanding both rational and intuitive reactions to disease. Cholera may have killed a half victims in a 19th century slum while tuberculosis may have killed nearly 100%, but cholera was feared more. Rightly so, because it spread very fast, and killed in a matter of days rather than years7. You were much more likely to die of cholera than TB. TB was a sniper; cholera, a machine gun.
If a disease spreads fast, it may overwhelm a health system. This happened in China. If one ignores the rate at which people are infected and die, the figures mean little. An estimate from the Chinese Center for Disease Control and Prevention reported that 81% of cases it identified were mild, (no or mild pneumonia), 14% severe, and 5% critical.8 How dangerous is this picture?
It matters how fast things happened. 80,000 people in a single city were infected by COVID-19 in just a few weeks. The epidemic probably peaked around late January or early February, 4-6 weeks after the first reported case on 31 December. That placed a massive burden on the local health system, which was overwhelmed.
Were this to happen elsewhere, other health systems would likewise be overwhelmed; perhaps more so, as healthcare resources became scarce globally. A larger proportion of the seriously ill would die, and there would be a knock-on effect on mortality among others served by health systems.
The same report suggests a case-fatality rate of 2.3%. A recent retrospective hospital study of 1,099 participants found 1.4%.6 Variations show that case-fatality depends on more than just intrinsic properties of the disease, and among the other factors, is the ability of a health system to provide proper care. The faster the disease claims lives, the more pressure it places on health systems, and the move lives the disease is likely to claim. This is part of the explanation for the current level of concern.
Comparisons with seasonal influenza are sometimes attempted. These must be interpreted carefully. Seasonal flu infects 1 billion people and kills 9.3m.5 Why, then, are people so worried about COVID-19? The answer lies in the speed with which COVID-19 took hold in Wuhan, and overwhelmed the health system. While seasonal flu has a case fatality rate <1%, the case-fatality rate of the Spanish Flu pandemic in 1918-20 was in the region of 2-3%, and this is also the region of estimates for COVID-19 in its current settings. If the circumstances in Wuhan were replicated elsewhere, COVID-19 became widespread, and 2-3% case fatality rate were extrapolated, this would be as globally significant, as was the Spanish flu, which killed more people than the Great War that preceded and created the conditions for it. This explains the concern.
Justifying the reaction: cost-benefit analysis
Behind the scenes, governments will need to undertake cost-benefit analyses of public health measures. A week-long global curfew might prevent spread but would also have significant economic costs. How should costs and benefits be balanced? Here are seven relevant considerations.
• Effectiveness. It seems obvious, but effectiveness of measures needs to be established. One study suggests that quarantining a cruise ship led to many more cases than if the ship had been allowed to dock.9 If the ship had docked on 3 February, 79 infected persons (out of 3,700 persons on board) would have disembarked. Instead the ship was quarantined, and by the time of eventual disembarkation on 20 February, 619 were infected, despite effective onboard quarantining. The infection rate on the ship was four times that in Wuhan at the epicentre. Public health measures are sometimes not intuitive, and relevant experts—infectious disease epidemiologists—need to be consulted.
• Speed of response. Because rate of infection is so important, as explained above, measures taken sooner will be more effective. If COVID-19 infects 1 billion people (like seasonal flu) over 3 months instead of 12, like flu, it will be considerably more devastating. This needs to be considered in estimating the benefit any effective public health measure. If schools are to be closed, better sooner rather than latter; and the closures will thus be shorter. I am not advocating closure of schools, merely pointing out that when is a factor as well as whether.
• Economic impact on human welfare, including health. Restrictions on travel and trade are having an economic impact. Any intervention that increases human welfare by preventing COVID-19 but damages it through an economic downturn must be assessed in light of both contributions to welfare. It is particularly important to understand that there is a direct effect of economic downturn on health (and not just on welfare in a more general sense). Wealth is a determinant of health, and a global economic downturn will have its own case-fatality rate that must be considered.
Nuancing of public health measures is also important in this context. While many people will suffer as a consequence of measures that restrict their ability to work and trade, manual labourers of all kinds are particularly affected, as they cannot work from home. A cost-benefit analysis needs to consider this, especially taking into account differential local risks, perhaps differentiating between local and rural settings.
• Upsides of economic downturn. An economic downturn may have positive effects, both on human welfare, and on non-human goods. For example, restrictions on illegal wild animal product markets in China may reduce poaching in Africa. This yields a social good, by reducing the intrusion of organised crime into vulnerable communities, and the corruption of organs of the state.
Such an effect would also yield a benefit for the environment. Whether you value environmental goods in their own right (for their beauty, intrinsic worth, whatever) of for their present or future contributions to human welfare (i.e. not just for “sustainability” reasons or the enjoyment of your grandchildren), then an economic downturn is, in this respect, good news: reduced carbon emissions, less waste, etc.
Can social or environmental benefits be offset against the economic cost of a public health measure? Where they concern human welfare, it seems reasonable to think that they can. Banning travel may reduce tourism, and damage a local economy, but if it makes rhino poaching harder, maybe that is some small consolation.
The harder case is where the benefit is non-human: can we also point to the decreased threat of rhino extinction, in similar small consolation? If you believe the environment has intrinsic worth then you must, I think, weigh non-human benefits to the environment in favour of public health measures that damage the economy. You should attach less importance to the human suffering caused by damaging the economy than to the suffering caused by COVID-19.
Alternatively, you could decide that the only value of a beautiful lake or glorious bear lies in its utility to us, either present or future generations. You must, however, engage in some sort of choice when contemplating the level of economic downturn you are willing to contemplate as a consequence of public health measures against COVID-19. This is one of those awkward cases where we are forced to weigh human against non-human goods, or else reject the latter.
• Identification of losers and winners, and weighing of their rights. Older people and those with pulmonary conditions are at greater risk of developing serious illness from COVID-19 infection. The protection of these groups must be weighed against the good of the whole. Closing schools, for example, will surely reduce rates of infection; but it may harm the children’s education, to which they arguably have an important right, for the probable benefit of their grandparents. Such trade-offs are generally considered permissible, as in taxation to pay for a public service, but not unrestrictedly so. I am not commenting on any particular measure here; school closure is merely an example and may in fact benefit children significantly. I am merely trying to structure the cost-benefit analysis by pointing out that those who suffer most from the measures may not always be those who benefit the most, and their respective rights and interests must be weighed.
• Assessment of quantity of life. Delicate though it may be, the public health response needs to take seriously the extent to which those who die are sick and elderly. This is not because their lives matter less, intrinsically. It is because they have less life left. We may imagine that if COVID-19 had not killed someone this week, they would have died of some other ailment next week, or even on the same day. Such imaginary cases are the subject of study by philosophers who think about causation, but in this case they have an evaluative import. Life is a quantity rather than a binary variable, and disease reduces its quantity, both for individuals and populations. The quantity of this reduction is less for people who are going to die sooner: the sick and elderly, in other words. If the cost of a disease to a population is the quantity of life that it takes, then this can only be measured in time, because all life is a quantity of time.
Some of us have more life left than others. Thus the cost of COVID-19 to population health is not simply a function of how many it kills, but whom. This is surely not the only factor in assessing the cost of a disease or benefit of a public health measure, but, given that life comes in a finite quantity, the quantitative effect must be taken into account in assessing the benefit offered by a public health measure against the cost of the disease.
• Egalitarian considerations. Preventive measures may benefit and harm people differentially, with the less well-off being more badly affected. While this is true for many diseases too, it is possible and perhaps probable that COVID-19 would not differentiate as much. Where cities are richer than rural areas, public health measures will benefit the rich, because risk of infection is higher in cities. Office workers and factory workers may have differential risks of infection, but a travel restriction will affect factory workers less because many office workers can, while most factory workers cannot, work from home. The wealthy are generally able to manage hard times better: they have access to savings, credit, or support networks.
We can seek to ensure that benefits are fairly distributed, and this form of egalitarianism is very attractive where public health is concerned. However, if one believes that inequality is wrong in and of itself, one needs to factor it into the cost-benefit analysis. If any measure that increases inequality is to that extent negatively valued, then one must weigh the inequality a measure causes against its effectiveness. This, for me, is highly counter-intuitive; I cannot imagine any situation in which we would prefer a measure that yields more deaths, regardless of whether the dying group is more or less representative of the population. However, those who do consider inequality a wrong in itself must factor in the point that COVID-19 may be more socially just than the global economy.
To summarize, I have offered seven “viewpoints” from which a contemplated public health measure may be assessed, in order to arrive at a rational decision:
- Speed of response
- Economic impact on human welfare, including health
- Upside of economic downturn, both human and non-human
- Identification of losers and winners, and weighing of their rights
- Assessment of quantity of life
- Egalitarian considerations
These points are made in a philosophical context, and some might cause outrage if openly utilised in a policy context. However, behind the scenes, these factors must structure a rational decision about immediate action in circumstances of uncertainty. Philosophy of epidemiology stands to human health as does philosophy of biology to human origins: illuminating, but not always to our liking.
Further considerations: panic buying and the prevention paradox
Masks, latex gloves, and other useful items are being purchased by anxious individuals and institutions. This is leading to global shortages of medical supplies that are needed for many medical purposes. Many public health measures require individual cooperation, and many of these share a feature that makes them difficult to implement: the average individual stands to lose more from participating than not.
Geoffrey Rose called this the prevention paradox: “a preventive measure that brings large benefits to the community offers little to each participating individual”.10 In the 1940s, 600 children needed to be vaccinated against diphtheria to save one life. But it was worth it; diptheria was eradicated; but motivating individuals to vaccinate was (and remains) a challenge.
A similar problem also applies to panic buying, in reverse. Refraining from purchasing supplies offers you little as an individual; in fact, it may even cost you, especially if others do not follow suit. But the global benefits are huge.
We put ourselves, and notably our children, before the mass of humanity on a daily basis. It is hard to condemn all such behaviour; people have tried to imagine living otherwise, but none have succeeded. Still, public health, like democratic process and war, sometimes requires unwinding this behaviour just a little. The messaging must not be simply that the masks don’t work; this is not credible, when so many doctors pictured in the media sport one. Nor must the message be moralising: we all behave selfishly, all the time. The message must directly appeal to a sense public good: “wartime spirit”, or something of the sort. Propaganda machines are capable of promoting compassion and cooperation, just as they can promote competition. Maybe the cultivation of such attitudes, not currently very prominent, would be a positive outcome of the outbreak of COVID-19.
- World Health Organisation. WHO Coronavirus. WHO website. https://www.who.int/health-topics/coronavirus. Published 2020. Accessed March 8, 2020.
- UpToDate. Coronavirus disease 2019 (COVID-19). UpToDate.com. https://www.uptodate.com/contents/coronavirus-disease-2019-covid-19. Published 2020.
- Broadbent A. Philosophy of Medicine. New York: Oxford University Press; 2019. https://global.oup.com/academic/product/philosophy-of-medicine-9780190612139. (p.4)
- World Health Organisation. Coronavirus Disease 2019 (COVID-19) Situation Report – 49.; 2020. https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200309-sitrep-49-covid-19.pdf?sfvrsn=70dabe61_2.
- Lockerd Maragakis L. Coronavirus Disease 2019 vs. the Flu. Johns Hopkins Medicine. https://www.hopkinsmedicine.org/health/conditions-and-diseases/coronavirus/coronavirus-disease-2019-vs-the-flu. Published 2020. Accessed March 10, 2020.
- Guan W-J, Ni Z-Y, Hu Y, et al. Clinical Characteristics of Coronavirus Disease 2019 in China. N Engl J Med. 2020:1-13. doi:10.1056/NEJMoa2002032
- Morabia A. History of Epidemiologic Methods and Concepts. Basel: Birkhauser Verlag; 2004. (p.9, p.17)
- Wu Z, McGoogan JM. Characteristics of and Important Lessons From the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72 314 Cases From the Chinese Center for Disease Control and Prevention. JAMA. February 2020. doi:10.1001/jama.2020.2648
- Rocklöv J, Sjödin H, Wilder-Smith A. COVID-19 outbreak on the Diamond Princess cruise ship: estimating the epidemic potential and effectiveness of public health countermeasures. J Travel Med. February 2020. doi:10.1093/jtm/taaa030
- Rose G. The Strategy of Preventive Medicine. Oxford: Oxford University Press; 1992. (p.12)
Very informative but also very offensively speciesist. The only nonhuman values identified as possibly relevant are the alleged intrinsic values of species, the environment, etc. The wellbeing of individual nonhuman sentient animals is ignored, and, ironically, abusive treatment of nonhuman animals is responsible for outbreaks of MERS, SARS, and the Coronavirus.Report
No, I mention intrinsic worth of environment and specifically include an animal as an example, the rhino..Report
“The harder case is where the benefit is non-human: can we also point to the decreased threat of rhino extinction, in similar small consolation? If you believe the environment has intrinsic worth then you must, I think, weigh non-human benefits to the environment in favour of public health measures that damage the economy. You should attach less importance to the human suffering caused by damaging the economy than to the suffering caused by COVID-19.”
“Rhino extinction”, rather than the wellbeing of individual rhinos seemed to be the primary concern. If that was not your intention, then I apologize and I hope you do oppose the way humans actually give little weight to the wellbeing of individual nonhuman animals in most cost benefit analyses.Report
I don’t have a firm view on whether you should have mentioned it but Kaildraper is absolutely correct that you did not mention animal wellbeing. Rhino extinction and the intrinsic value of the environment only bear accidentally on the wellbeing of individual animals. Rhinos could go extinct painlessly and living happy lives. Preserving the environment may lead to the continuation of very low welfare among many wild animals. The relations between animal wellbeing and environmental protection are real but they are far from being linear and simple. This is of course to say nothing, as Kaildraper points out, of the likely origins of those kinds of pathogens.Report
You seem to have misinterpreted source number five. You cite the source as saying that the flu infects 1 billion and kills 9.3 million globally on a yearly basis. The source in fact says that the flu infects 1 billion globally, and * infects * 9.3 million in the US, on a yearly basis. The source then says that the flu * kills * up to 646,000 globally, and up to 61,000 in the US, on a yearly basis.Report
Yes I misread it – will seek a correctionReport
Actually, I liked the article at first blush except for the billions/millions of deaths part of the flu season as those numbers seemed to be out of line. So I was not surprised by the correction.
I’ve been following https://www.worldometers.info/coronavirus
closely and this site is a wealth of specific data rationally analyzed and presented with multiple updates every day. You comments seem fairly consistent with this site.Report
Thanks, Dr Broadbent, for a thoughtful and thorough article on the effects of the COVID-19 epidemic on humans. I for one am glad you did not expand it to discuss the effect on non-human sentient beings or on the cosmos, since i am 84 years old and don’t have the time to read everything 🙂Report
Appreciated… was not expecting animal sentience to be the primary point of contention!Report
Well it’s not. One person made a comment, which you did not really address. Many more made different comments on totally unrelated topics. No one says it is, much less should be, the primary point of contention. It’s a great article.
@Dr John Redwine French: then don’t read the comments!Report
It’s not just the elderly and people with respiratory problems that are at risk here. People who are on immune suppressants for a variety of reasons are also at risk. Pregnant women are at risk. We still don’t know whether the virus specifically impacts fetal development. We do know that fevers are always dangerous for fetal development. An important difference from the flu is that people who are at risk can get the flu shot, including the elderly and pregnant women in order to lower their risk. They can’t do that with this virus.Report
Answered in this recent Lancet article (below). 9 pregnant women with lab-confirmed covid-19-related pneumonia. All babies healthy and clear:
Volume 395, Issue 10226, 7–13 March 2020, Pages 809-815
Clinical characteristics and intrauterine vertical transmission potential of COVID-19 infection in nine pregnant women: a retrospective review of medical records.Report
The 9 women you mention were all late term. My comment still stands.Report
Also, a sample of 9 doesn’t allow us to make a very informed judgment about risks.Report
I think the really unusual thing here is that children do NOT seem to be at increased risk, which is unusual, compared to flu. But of course there is lots of ignorance about exactly who is at higher risk. I didn’t mean to imply that there aren’t others at risk; I’m just indicating the main ones, following what I’ve found on WHO elsewhere. It doesn’t bear on the arguments, anyway, except that children are unusually appearing not to be at risk, which I do point out has some implications.Report
I contend that this comment is something of a straw man. The risk profile of Covid19 is well evidenced. If everyone was as vulnerable as the under fifties with no co-morbidities, then there would be no argument for any action to stop the spread. Children especially are at no significant risk from covid19. The closure of schools, the forced increase of screen time, the closures of playgrounds, sports facilities, libraries, the enforcement of play prevention in public spaces, these things are serious negative impacts on children’s lives and life chances. Children are bearing the brunt of this action without their consent and without their interests being taken into account. The argument that school closures and the rest of the attacks on children will slow the spread of this disease and thereby possibly save lives in the short term is very weak and relies on faulty assumptions. But however effective school closures are at preventing fatalities, the moral point is that children are being harmed in order to improve the life chances of other people. I can’t think of any other situation in which this would be considered acceptable.Report
Just *how* overwhelmed was the healthcare system in Wuhan? Because something I don’t really understand is, if 15% of cases are “requiring” treatment, and the system was “overwhelmed”, shouldn’t the death rate have been much higher than 2.3%? Like, approaching 5-15% during the period when almost no one could get treatment, which is what I assumed “overwhelmed” would mean, depending on the % who “require” treatment who’d have died without? I’ve read and read and read expert stuff on this, but none of it has helped me understand this disparity. Does “requiring” treatment mean something much less than “might well/probably will/almost certainly will die without it”? Is it that actually only intensive care capacity was overwhelmed and most serious cases just need some oxygen? If the latter, could the death rate actually shoot up in the direction of 10% or so if a hospital system ends up with 10 times the number of cases Wuhan had?Report
Or wait, was 2.3% the rate for a much longer time period, rather than just for when Wuhan was overwhelmed?Report
“The limited availability of beds in Wuhan raised their mortality rate from 0.16% to 4.9%“
Ah. That makes sense. I assume that a decent proportion of those who needed oxygen still got it and so that isn’t an absolute cap?Report
I think it’s pretty much an absolute cap. Also from the article:
“When asked why Wuhan was so much higher than the national level, the Chinese official replied that it was for lack of resources, citing as an example that there were only 110 critical care beds in the three designated hospitals where most of the cases were sent.”
Given that Wuhan has 11 million people, I can’t imagine that the number of people who got medical care makes a significant difference to the stats.Report
Very good if you are right. I wonder why the WHO is insisting 10-20% “required” weeks of hospital treatment then. In the current context, you don’t really *require* it, if it just makes the experience less miserably painful but you’d have lived anyway (though obviously you quite rationally really *want* it in those circumstances.)Report
I think that’s right. But here’s a possibility: they had only 110 “critical care” beds; for all I know, there are lots of kinds of beds: critical care, intensive care, serious care, etc. So it could be that if we had 0 beds the fatality rate would be higher. But I’ll have to look into this. (FWIW, Iran is probably doing less for this than any other country, and last I checked, their CFR was 9%! But their data is also considerably less reliable.)Report
Presumably for some people, some of the medical treatment doesn’t just prevent death or prevent temporary pain, but might prevent long-term injury as well. I am not a medical professional, but I would not be surprised if lack of oxygen could cause long-term brain damage or damage to other organs, that would impact life long after the event, and could have been mitigated with oxygenation during the acute period of infection, but that doesn’t actually kill the person during their period of infection.Report
Agreeing with Kenny…. It is very strange to think the role of a hospital is only there to prevent death for a *particular* individual. That isn’t how epidemiology works. Medical decision making isn’t done by applying some backward-looking counterfactual close possible world philosopher fantasy.
Let’s say severe dehydration can lead to severe medical complications or even death in 20% of cases. Even if in your particular case, you won’t have those complications, it still makes sense to seek hospitalisation. The issue is 1) often we don’t know beforehand everyone who would develop these complications, and 2) if most people *don’t* seek hospitalisation, more people in the aggregate will die / have severe complications. Hospitals treat individuals, but protocols and medical decisions are made with epidemiological considerations. It is the protocols that determine whether hospitalisation is required.
Here is another example: Not all babies that are breach would die in childbirth. However, requiring a cesarean delivery for all breach babies decreases the amount of bad outcomes in the aggregate. That is why many hospitals and doctors would say that a cesarean is medically required in those cases. Again we don’t know beforehand who *needs* a cesarean in the philosopher sense, before it’s too late.Report
I think this is all right – ICU wards aren’t that huge and don’t sit around with loads of spare ventilators. It isn’t hard for something to overwhelm a hospital if the treatment required is intensive care, as opposed to say cholera where it’s about hydration and hygiene. That’s my assumption anyway.Report
Thanks, for a thoughtful article. In some cases, we will be careful. For example, this article is about practice good hygiene. Each person should follow these rules.