Richard Lehman’s covid-19 reviews—23 March 2020

In this weekly round-up, Richard Lehman looks at a personal selection of articles of relevance to clinicians dealing with covid-19

Covid contemplated

Nobody is an expert on covid-19. It has not been around long enough. The idea that I could be a thought leader in this field should inspire you with feelings of mirth and dread. And this is exactly what I hope to do. 

Thomas Pueyo runs a billion-dollar marketing company and was also new to the study of coronaviruses until a few weeks ago. But if you want a magnificent overview of the pandemic which is also a closely argued call to action, the place to go is his The Hammer and the Dance.

Behave yourselves

Quarantine used to mean staying isolated for forty days, often on a spice boat off the harbour of Venice. People are not very good at this on dry land. An exception may be academics, enthralled at the idea of being able to write papers in self-isolation, and it is amazing how quickly they can work under such conditions (see the silver lining section below). “How to improve adherence with quarantine is a super-rapid review of something the authors describe as a “vexed issue.” Fourteen studies out of 3163 papers made it into the analysis, and adherence to quarantine ranged from 0 up to 92.8%. So there are truly useless ways, and fairly useful ways. The latter don’t just involve the presence of uniformed officers: they also involve trust, good communication, and shared understanding. 

The duration of quarantine following covid-19 is also a vexed issue. Studies show that viral shedding can continue for up to at least 3 weeks following the onset of infection, but most guidelines give a Get Out of Jail card to convalescents a week sooner. “Is a 14-day quarantine period optimal for effectively controlling coronavirus disease 2019 (covid-19)?” asks an article from China and (not surprisingly) answers no. In the course of the discussion, the authors also discuss length of incubation before symptoms, and interestingly conclude that it is probably longest (median 9 days) in those who acquire the virus from shared meals.

Early and near

Revolving tables laden with Chinese delicacies do not account for most global cases of covid-19. I suspect children are commoner vectors in the West, however seldom we eat them (see below). Italians are particularly prone to snatch babies and children from their rightful owners and slaver them with kisses. I’m not drawing hasty conclusions about the virological perils of bambinophily: such things are for the epidemiologists to decide. What we do know is that the greatest risk comes from close contact with anybody’s airborne secretions and that these are most infective if they are in the early stages of the illness.

Kids with covid

Children may well have been key vectors early in the Chinese epidemic. In Wuhan, in January 2020, influenza was much commoner in hospitalised children than covid-19, but it is likely that a lot of the latter was missed. Dependable data are only available for 3 of Wuhan’s 400 or so hospitals, but by extrapolation the authors suggest that there may have been over 1000 children admitted with covid-19 prior to the January 23rd lockdown in the city. Tens of thousands more would have stayed at home or gone to school with negligible symptoms. “This highlights the urgent need for more robust surveillance to gauge the true extent and severity of covid-19 in all ages.” Yes. The mighty journal Nature agrees. 

Testy mutterings

Over the last two weeks in Britain, everybody has been getting testy, but nobody has been getting tested. With luck this will have changed by the time you read this. It has not been for lack of a range of possible testing products, mostly based on the conversion of viral RNA into DNA and its multiplication using polymerase chain reactions. You may remember that back in early January a team had already constructed the entire SARS-CoV-2 particle, so there are lots of other bits and proteins that can be made the basis of viral detection kits. The predictive characteristics of these tests vary a lot, as shown by studies like one based on nucleocapsid and spike proteins. The idea of a “gold standard” test for this infection is likely to remain illusory. Remember this when you read about numbers of “confirmed” covid-19 cases. As a denominator to calculate the case-fatality rate, they are almost worse than useless. 

For individual clinical diagnosis, however, viral detection tests (swab based) are very much better than useless. They have a reasonable positive predictive value, though this needs to be confirmed later by serological testing. It’s becoming clear that different people make different antibodies to SARS-CoV-2 at different stages. Standardised IgG based testing became simple and scalable weeks ago, and we need lots of it. Members of the herd who have immunity are a very valuable resource, especially at the front line of covid care.

Serum farming

We may also need to farm these convalescents for their blood. Seriously? Satire can be hard to distinguish from reality in the time of covid, as it was in 1729 when Jonathan Swift put forward his Modest Proposal that the starving Irish should roast and eat their children. Once the words “herd immunity” had slipped from the lips of the UK’s political masters a couple of weeks ago, the world stood aghast to see what might happen next. Would Britons proceed rejoicingly towards the collapse of the NHS and a cull of the old and the ill? Or might we convert schools into serum ranches to extract antibodies from the blood of young herdlings who had all acquired and recovered from covid-19? The use of vampires was possibly suggested in private briefings. Who can know. In fact convalescent sera could be an important therapeutic or even preventive tool in the fight against the virus, argues a viewpoint article in the J Clin Investig. For those at highest risk, this sounds very welcome.

Trump card

Just because POTUS has backed a combination of hydroxychloroquine and azithromycin as a cure for covid-19 doesn’t mean that it isn’t. It is just that the tiny French exploratory trial on which his extravagant hopes are grounded scarcely even counts as hypothesis-generating. To quote from the young Max Beerbohm, “it would take a pen much less eloquent than mine to do justice to this subject.” Wait for proper trials of this and dozens of other plausible contenders. There are plenty.

Arriving antivirals

The world is afire with -ivirs. Most of them are repurposed drugs with known activity against other viruses. We have reasonable starting knowledge about their safety, so people enrolled in trials have at least some information to guide their choice. Don’t try to keep up with the trials unless you have a team to help you and you don’t need any sleep. Fortunately there seem to be many good people of that description who are doing the work for us. From France comes A brief review of antiviral drugs evaluated in registered clinical trials for COVID-19 in which the chief contenders are stem cell therapy (n=23 trials), lopinavir/ritonavir (n=15), chloroquine (n=11), umifenovir (n=9), hydroxychloroquine (n=7), plasma treatment (n=7), favipiravir (n=7), methylprednisolone (n=5), and remdesivir (n=5). Open label designs predominate, unsurprisingly. Now is also a good time for you to learn about antivirals you may never have heard about, such as arbidol.

Covid’s Metamorphoses

Clades and subclades. Fomites and vectors. The hair of the student you fancied catching the sun, and the drone of the lecturer’s voice. Those temps perdus of didactic epidemiology are coming alive again, complete with meaningful arrows and a certain linguistic splendour. These are seen to best advantage in Tomas Pueyo’s beautiful graphics chart 6 depicting what he interprets as inevitable mutation of SARS-CoV-2, which he fears will ensure its persistent virulence in the human population. Or else the virus may just fizzle and become lung junk, like other human coronaviruses. There will be a huge literature on coronavirus metamorphosis for teachers of the future to call on: examples are appearing every day. Before very long lecture theatres will be full again, and young hair will shine once more in the dusty sunlight. 


In those coming days, people of my age who are left will dip madeleine biscuits in their tea and remember the time of fear. We had it too good for too long. Even at the best of times, I don’t like prognostic scores and modelling, and I certainly don’t like them now that I am a subject. Currently a lot of people are drawn to predicting who will live or die in this pandemic, using methods like those described in Building a covid-19 Vulnerability Index. Better models are no doubt available. Maybe I will get to read follow-on studies in a few years’ time over my tea and biscuits, showing which of these fitted the data best. Or maybe I won’t.

Silver lining 

A cynic on Twitter wrote that at least the coronavirus has stopped people talking about AI. This will never happen. In trillions of years’ time, when the Universe is in the last throes of cold death, a thin network of artificial intelligence will remain embedded in the lonely scattered quarks and photons, still promising imminent genomic cures for everything. 

Happily a more basic kind of AI is showing its worth in tracking covid-19 research. CORD-19 is one of several amazing large-scale initiatives which have sprung up in the last couple of weeks. Here are some others:


Richard Lehman is professor of the Shared Understanding of Medicine at the University of Birmingham.

Competing interests: None declared

Articles from MedRxiv get special prominence in these reviews. MedRxiv is a completely free site for the rapid exchange of knowledge which was launched last year as a joint initiative by The BMJ, Cold Spring Harbor Laboratory and the Yale University Open Data Access project. All MedRxiv articles carry the caveat: This article is a preprint and has not been certified by peer review. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice. This review is intended as a quick source of information in a readable form. While every effort has been made to be accurate, the opinions are those of the author and should not be relied on without reading the full articles cited in the context of current NHS guidance.