The cure for fake news: how to read about the coronavirus

I’m confused by all these new words…

Seldom have we had to get to grips so quickly with a slew of new terms and abbreviations. Had you told most people last Christmas that soon they would be WFH, they might have considered it a cryptic insult. Most probably had only a vague idea what an epidemiologist did or what a ventilator was, while PPE was (if anything) a soft Oxford degree favoured by politicians, rather than the personal protective equipment now urgently needed in hospitals. Few could have told you the difference between an epidemic (the outbreak and rapid spread of an infectious disease through a population) and a pandemic (an epidemic that occurs over a very wide geographical area, perhaps the whole world).

Even the terminology of the virus itself is challenging: the coronaviruses are a family of pathogens, this particular one (being related to that which caused the Sars outbreak of 2002-3) denoted Sars-CoV-2. Covid-19 is the respiratory syndrome it causes, much as Aids is to HIV. A vaccine confers immunity (probably just for a limited period) to the virus; an antiviral interferes with viral replication in the body to slow or arrest its bad effects. Antibiotics attack bacterial pathogens, not viruses, but may be needed with Covid-19 to treat secondary infections such as pneumonia that take advantage of a compromised immune system. Expect to hear lots in the coming weeks about serological tests, which look for proteins in the blood that signal past infection, even if the virus has long since been cleared from the body. Those tests are under development, and they will provide vital information on how widespread infection has been – and whether many (or few) have had Covid-19 without any symptoms at all.

While comparisons with wartime and the blitz spirit are deeply misleading, one analogy holds: the second world war slogan “Loose talk can cost lives” is apt here too.

What’s an important number and what’s not?

What makes the coronavirus so deadly is a combination of two numbers. The reproduction number (denoted R₀) is the average number of people an infected individual goes on to infect in the absence of any social distancing or quarantining; the mortality rate is the percentage of infected people who die.

Both depend on the biology of the virus, but neither is fixed. R₀ for the coronavirus is about 2.5 to three (compared with 1.3 for common flu), but that’s an average – it varies from person to person. Reducing person-to-person contacts cuts the effective reproduction number R – and the epidemic will decay if it drops below one. The mortality rate, meanwhile, depends on the availability of healthcare for people with the most serious symptoms: it is about 0.4% in Germany, 4.6% in the UK, and more than 10% in Iran.

The bad news is that low testing rates means it is virtually impossible to track the progress of the virus, and can give a false sense of security. The good news is that if more people have the virus than we know, the mortality rate will be lower than the “official” value. There have been speculations that the extent of infection might be much greater than thought, and that most cases don’t even show symptoms. But this idea is hard to reconcile with observations from countries with extensive testing, which seem to indicate that few people get the virus without knowing.

I heard a story on social media. How do I find out if it’s true?

When you see a WhatsApp message about “A friend who is an A&E doctor at a leading hospital … ”, don’t assume it is reliable, even if it came from your own friend. Don’t, even with the best of intentions, tell your other friends about it.

If you want to check the credibility of social media posts, organisations such as FullFact and Snopes audit messages doing the rounds.

Remember also the old quip that the plural of “anecdote” is not data. We notice anomalies and give them undue emphasis. Some highly unusual cases, among so many infections, are to be expected but this doesn’t mean they are true or examples of a common phenomenon.

The news website The Conversation has a handy guide to spotting fake news. Perhaps the most useful of its tips is: “If the story appears to claim a much higher level of certainty in its advice and arguments than other stories, this is questionable.”

But what if experts disagree?

Sorting fact from fiction has become harder in this crisis because even well informed experts aren’t always united in their advice. Disagreements are normal in science, but are rarely a matter of life and death. It has been as frustrating and unsettling to science reporters as it has to everyone else to find experts at loggerheads over basic concepts such as “herd immunity”. It is, frankly, still a source of puzzlement why it took a single modelling study from Imperial College London to persuade expert government scientific advisers of what other experts (such as those at the World Health Organization) had been saying all along: that without stringent containment measures, the UK death rate would be catastrophic.

So now even the fine details of the science are being held up to us all for scrutiny. It’s unprecedented for a national newspaper to publish a full, simplified version of a scientific paper, as the Observer did with the Imperial study. There will need to be more of this and with luck, it will spread understanding of how scientific results are carefully formulated and circumscribed with caveats, so that experts are seen as well-informed specialists doing their best with the tools and data at hand.

With the stakes this high, experts exploring speculative ideas take on a responsibility they aren’t used to. A study from epidemiologists at Oxford recently caused a stir when it was reported as claiming that as much as half of the UK population might have already been infected with the coronavirus, mostly without any symptoms. The work was quite reasonably exploring a “what if” scenario. But the headline in the Financial Times made it sound like an empirical possibility. Researchers will need to be more explicit about what their findings do and do not mean.

View image in fullscreen

What do I need to know about vaccine development?

All drugs have to pass stringent tests before they are approved for use. In general they are first tested on animals, and then enter “clinical trials” on humans, of which there are three stages. Phase one, conducted on small numbers of healthy volunteers, checks for safety, looking for toxicity or serious side-effects. Phase two then tests whether the drug actually works in humans with the condition targeted. Finally, phase three trials are conducted on large numbers of people – typically hundreds or thousands – to see how it will work in the clinic and make sure there are no other complications or failures before making the drug generally available.

Vaccines will have to go through all of these phases, and could, like most candidate drugs, fail at any one of them. While drug-regulation agencies are considering ways to streamline and fast-track the process because of the urgency of the situation, this can only go so far. A coronavirus vaccine would be given to millions of healthy people – ultimately, perhaps to most of the world’s population – and so it’s more essential than ever that its safety is checked thoroughly.

What’s more, even when a vaccine gets approved, manufacturing it at such a large scale will be a tremendous challenge – only a few pharmaceutical companies worldwide have that capacity. And distribution will present difficulties too, especially to more remote parts of developing countries, or among overcrowded, poorer neighbourhoods in large megacities. If, as is hoped, a vaccine is ready by the middle of 2021, that speed will be remarkable and unprecedented.

What about drug treatments, including existing ones?

It is hoped that some antivirals already developed and tested for other conditions, such as Ebola (remdesivir) or Sars, could work for Covid-19, because they attack similar targets in the virus. In that case some of the testing has essentially already been done, so we would get a head start. There is also broader interest in “drug repurposing” – investigating whether the many drugs that failed for their intended use, but which passed the safety trials, could prove effective for quite different conditions.

In tests on a new drug – how safe or effective it is – it matters how big the sample is. Drugs don’t get approved until they have been tested on hundreds or thousands of people, so outcomes for just a dozen or so are highly preliminary and far from proving efficacy.

Be especially cautious of reports that rely heavily on quotes from drug companies. They are very unlikely to be lying, but institutional press offices have become accustomed to putting a rosy spin on their researchers’ work and drug companies are under commercial pressures that don’t always encourage them to give the full picture. You needn’t be cynical, just cautious.

Should I keep scanning Twitter?

Whether Twitter is a handy, up-to-the-minute source of information, or a hive of misinformation and conspiracy theories, is largely up to you. All the usual caveats about using Twitter – beware of strong opinions from pseudonymous tweeters with 15 followers, or of hot takes from media loudmouths with no expertise – are multiplied and compounded here. For the problem with the subject of coronavirus on Twitter is that it’s not just the trolls and attention-seekers who are weighing in. Plenty of serious academics, too, have taken it upon themselves to become epidemiologists and virologists overnight, criticising or praising scientific studies or government policies with all the assurance of seasoned veterans. Often what they say is not ignorant or wild, but merely narrow and lacking context – they might be good at interpreting statistics, say, while having no knowledge of viral transmission or mass behaviour. Fear seems to be causing the usual academic caution to be replaced by hubris.

View image in fullscreen

I read this great analysis on Medium…

As the site itself says, “Anyone can publish on Medium … but we don’t fact-check every story.” Make that your starting point for anything you read there.

This doesn’t mean it’s wrong. But the authoritative-looking stuff you might encounter is probably not coming from experts. One post that went viral, Coronavirus: The Hammer and the Dance, which used an array of charts and statistics to argue for strong containment measures to avoid massive fatalities, was by Tomas Pueyo, a Silicon Valley consultant who works on educational online content and has no epidemiological training. His post was promoted by a content marketing firm and was shared by the likes of Margaret Atwood and Steven Pinker. Whether or not it was accurate in all its details, it sent out a needed wake-up call about the dangers. But other unqualified “corona influencers” have spread dangerous messages, for example about hoaxes and bogus treatments. Medium has now announced that it will give “careful scrutiny to coronavirus-related content to help stem misinformation that could be detrimental to public safety”, and will remove any that it deems harmful.

How do I spot a coronavirus conspiracy theory?

First ask: does this theory confirm the suspicion I’ve had all along? About, for instance, 5G? Because if so, double your scepticism and then double it again. We aren’t so much duped as flattered into buying these theories, and the first defence against confirmation bias is to recognise that no one has natural immunity to it.

The harder ones to avoid are political, because they both speak more directly to our prejudices and are tougher to disprove. Is it “obvious” that, say, the government is just trying to keep its big-business donors happy at the expense of our own safety? We’d do well to assume in the first instance that leaders and policymakers are not Machiavellian strategists but error-prone muddlers like the rest of us.

Sorting fact from fiction is made harder by the media habits acquired during Brexit – the tendency of broadcast journalists and other media reporters to tweet unattributed and unconfirmed rumours. This grave global crisis should prompt some journalists to take a hard look at the ways they have become accustomed to working.