OFFICIAL! Coronavirus now more interesting than breasts
Hopefully just temporarily… but breasts are outnumbered 3-to-2 on today’s homepage on Medium:
(However — when selecting Tags for this article, “Coronavirus” came up with 0 hits, while “Breasts” had over 300…)
[Comments on a later article by the same author are here:]
Who is Tomas Pueyo?
On a more serious note — “I’m Not An Epidemiologist But…” is the title of a thoughtful Buzzfeed article by Ryan Broderick, and a number of us should perhaps be using that phrase as our mantra, these days:
Broderick notes that “Pueyo is the first to tell you that he’s not a doctor nor a scientist. He’s the vice president of growth at Course Hero, an educational content sharing platform that allows students to collaborate on study notes.”
That’s not quite true — Broderick, not Pueyo, was the first to tell us that. True, there is some blurb at the bottom of Pueyo’s recent articles in Medium — but they are extensive articles, and it takes 20 minutes’ reading to get to that point:
“2 MSc in Engineering. Stanford MBA. Ex-Consultant. Creator of viral applications with >20M users. Currently leading a billion-dollar business @ Course Hero.”
Again, I’m not an epidemiologist, but as to “currently leading a billion-dollar business”, Pueyo is actually “vice president of growth”, according to Broderick. I suppose you could say he’s one of the leaders of the business that employs him, then, if using “leader” broadly.
In any case, the articles of which Pueyo is the author or lead author are substantial pieces of analysis (and/or synthesis). It would just have been better to state his (their) credentials at the top, to provide context when reading the articles, and perhaps to understand what may influence or bias his (their) conclusions. (By the way, I’m an accountant, but that’s not quite so relevant in my case, as I don’t have millions of people reading and promulgating my proposed solutions to a pandemic.)
It is also fine by me if the correct recipes for responding to the pandemic are best articulated not by an epidemiologist or medic, but by a businessman, or data specialist, or journalist — whoever has the clearest thinking or most plausible strategy deserves to be heard.
However, I would prefer there to be some means of rapid expert vetting of coronavirus memes or popular articles. Not necessarily to hold them to scientific peer review standards, but at least to allow them to be quickly corrected or removed, where necessary.
Pueyo’s criticism of the Imperial College paper
Pueyo refers to “a very important paper published over the weekend from the Imperial College London”, which “has been brutally criticized for core flaws: They ignore contact tracing (at the core of policies in South Korea, China or Singapore among others) or travel restrictions (critical in China), ignore the impact of big crowds…”
Firstly, Imperial didn’t “ignore the impact of big crowds” — they considered it less significant than the other “non-pharmaceutical interventions” they were modelling, and therefore did not investigate or model that area more deeply at that stage:
“Stopping mass gatherings is predicted to have relatively little impact (results not shown) because the contact-time at such events is relatively small compared to the time spent at home, in schools or workplaces and in other community locations such as bars and restaurants.”
Perhaps the Imperial team was wrong to say that — perhaps they’re completely wide of the mark, or perhaps they are understating other factors, e.g. people do need to get to and from mass gatherings, typically head to a bar or restaurant before and/or after, etc.
Pueyo on Herd Immunity
I also found the section “Herd Immunity and Virus Mutation” in Pueyo’s second article rather weak. He/they are dismissive of the the idea that people who are infected and recover are immune to the virus, because this assumes the virus doesn’t change too much. And evidence is cited which suggests that the virus has already changed somewhat.
Maybe they’re right, but I would like to hear an expert’s view on that — has this virus changed any more than one would expect? Even if it has changed somewhat, what is the expected probability of a recovered patient catching the virus (or a variant thereof) in the near future? Presumably that makes a massive difference — if there is immunity (at least for six months, a year, two years), the many workers, in hospitals and in the wider workforce, who have caught the virus but recovered, are just the people you want to have around.
From what I’ve read, experts believe that there is a high probability that those who have recovered will be immune for a period.
Lack of analysis by age
Another area I found lacking in Pueyo’s articles is an analysis by age. For example, he describes the “hammer” (or extreme preventative measures) that afflicted countries could/should apply now to minimize disaster, but generally considers these only on an across-the-board, population-wide basis.
This may be the right conclusion, but other approaches should at least be modelled — such as applying extreme restrictions to cocoon the elderly, those with poor health, and anyone who comes into contact with them. While the rest of society carries on, taking reasonable precautions, but not shutting down the whole economy.
There may be obstacles to this — lack of public support for “locking down” the elderly, civil liberties concerns, lack of means to monitor/enforce compliance, etc. But there should at least be a frank discussion among scientists and policymakers of this differentiated approach — less of a “hammer”, and more of a precision instrument.
Other ideas to consider — would there be healthy volunteers willing to expose themselves to the virus in a controlled fashion, with immediate quarantine? Such volunteers, once they recover, would then presumably be able to help those at risk, or just live a normal life, with greatly reduced risk of being infected again and passing the virus to others.
That may be a fanciful or unethical proposition, but the scope of the debate should be wide, before we settle on appropriate actions.
Links to the Pueyo articles
Despite the above concerns, the 2 Pueyo articles are excellent resources:
and:
Lack of consideration of the economic impact & its consequences
The UK press in particular have spent years joking about the “magic money tree”, or lack thereof. That’s all been forgotten about now, partly because death or impending doom is a much better headline.
But I’m sure there must be serious studies out there, looking at how recessions damage health, both directly and indirectly. Swathes of people may die when no-one has enough money to take care of themselves, or pay taxes for the government to take care of them. There are many ways of dying other than from coronavirus symptoms, and many of them are expensive to prevent.
Countries with strong social safety nets may be willing and able to spend whatever it takes — but many countries are not in that position. Many of the jobs in such countries, where life was always more precarious, depend to some extent on trade with richer countries. And those countries, consciously or unconsciously, may be heading not for a mere recession, but a depression on a scale that few living souls have seen.
Maybe a new Great Depression is inevitable at this stage, and maybe the massive “lockdowns” considered by Pueyo and by Imperial College, on an ongoing or on-off basis, are, despite the initial pain, the best ways of minimizing the long-term economic damage. But business that go bust, tend to stay bust.
The costs and benefits of different approaches to COVID-19 is (literally) the most morbid of subjects, and there is little public discussion at a more granular level. We might be preventing (at least) hundreds of thousands of deaths in the short term, but potentially ending up with an alternative disaster in the medium and longer term — how do we know which scenario is worse, if we don’t face up to the pluses and minuses?
In other words, policymakers need to integrate their consideration of economic and health implications, as the two are bound together in a circle, whether vicious or virtuous.
Another example of a coronavirus “meme”
This one, widely shared, is actually not a “bad” example, in the sense that it’s visually appealing, people quickly get the message, and the differences between the 3 scenarios are broadly correct. It is nothing like the many actual bad examples, which are misinformation, either intentionally or through stupidity.
But it’s not really a good example, either. People will have read a number of articles (including the Pueyo one) about the importance of reducing “R0” to below 1, so that the virus peters out. But here we see a figure of 0.625 (in the third row) apparently leading to a 2.5x increase in cases…
Presumably by showing “2.5 people infected” they mean “currently or previously infected” — so the number of active cases would actually have petered out and be less than 1, after 30 days. But would it have been too much trouble to include less misleading descriptions, and provide a link to the methodology/assumptions used? They did add the formula in the comments on the original tweet, but didn’t give a fuller clarification, and most people just get to see the recycled image, not the original tweet.
The importance of context
While freedom of the press is difficult to argue against, sometimes you wonder if there should be some minimum standards, especially during a public emergency…
Ideally, all reporting on coronavirus deaths should be set in the context of a “bad year for flu” — for example, “The WHO estimates that between 290,000 and 650,000 respiratory deaths globally each year are associated with seasonal influenza. Public Health England estimates that on average 17,000 people have died from the flu in England annually between 2014/15 and 2018/19.”
Few of us paid much attention to the above previously, and we went about our lives as usual, rightly or wrongly.
Or more broadly — the elderly, or those in poor health, have a heightened risk of death — people die every day, and we can’t prevent this indefinitely. Figures on actual or projected deaths due to the virus need to be compared to the “normal” death rate.
The impact of COVID-19 may be much worse than the previous death rate due to flu and the many other causes, but perhaps not 100 times as bad, which is what the media coverage, government reaction, and public perception might suggest.
Other examples — “a nurse has died”, “a young person has died”, “several doctors have died”. Few things are more tragic than deaths in the cause of saving others, so these cases will inevitably make headlines. But the reporting also needs to reiterate that 99% of young people without health conditions (or whatever the current analysis suggests) will make a full recovery within a matter of days, and will be temporarily immune (if this is what the facts suggest).
In other words — the absolute numbers may be more newsworthy, but the percentages/rates are generally what’s important.
And is the second-by-second, drip-drip-drip “league tables” of infections and deaths what we really need? Wouldn’t daily reporting of changes since the previous day give enough information, without overwhelming people?
A final point
If laymen (like me) express their views, that’s less problematic in the sense that people are unlikely to consider such views authoritative. But there are currently too many talking heads from the medical profession itself. Not all doctors, nurses and experts are equal — some heads are wiser, some have a better understanding of the bigger picture, and so on.
It would be better if such professionals, whether formally, through their industry bodies, or informally, through self-restraint and collaboration, avoided a cacophony of voices, and allowed the best-informed amongst them to take the lead in informing the public.
