Communication about Coronavirus
There’s been some dire headlines about the current outbreak , such as ‘Virus Panic’ from the Daily Mirror last week, featuring one woman wearing a face mask, and reports of hand gel selling out. As a main official recommendation is to wash hands, buying hand gel hardly seems a panic reaction.
In contrast, sophisticated financial traders do seem to panic easily, and we’re seeing huge falls on stock markets when they are faced with uncertainty.
For this seems a classic example of risk vs uncertainty, the known unknowns against the unknown unknowns. Seasonal flu is an established risk: in an average year it kills an estimated 6,000 people a year in the UK, but gets no publicity (so far this season flu has killed between 18,000 and 46,000 in the US). But coronavirus is new and unknown, although we do know that it is a lot nastier than seasonal flu, being perhaps twice as infectious and ten times as lethal. But these figures are not only uncertain, but highly variable — it’s a lot more infectious among passengers sharing the air-conditioning in a cruise liner, and it’s a lot more dangerous to older people.
I think the current official communications to the public seem to be fairly consistent and good, with neither premature reassurance nor fear-mongering, with advice to wash hands, don’t cough and sneeze over people, and self-isolate if feeling unwell (incidentally, good advice for every flu season). And they seem to be largely following the open and honest approach that John Krebs took when handling multiple crises in the Food Standard Agency. When faced with communicating with the public around, for example, scrapie in lamb, he recommended the following messages:
- our knowledge: what we do know
- our uncertainty: what we don’t know
- our plans for contingencies: what we are doing
- your actions: what you can do in the meantime
- flexibility: things will change and we shall get back to you
Plans for contingencies are generally based around running epidemic models under a wide range of assumptions, and focussing on what is known as a ‘Reasonable Worst Case Scenario’ (RWCS), which is communicated to authorities responsible for dealing with the outbreak. The problem is when the media reports the RWCS as if it is what is expected to happen — there were alarmist headlines when a document was leaked which mentioned 1/2 million deaths.
It’s not helped by the RWCS sometimes assuming all unknown quantities are at their most pessimistic values — which, in my opinion, is not very reasonable. If you cast your mind back to swine flu in 2009, Chief Medical Officer Liam Donaldson’s back-of-envelope RWCS was 65,000 deaths, which got a lot of coverage. The actual death toll? 360. Of course this is only one example, and we may not be so fortunate with coronavirus.
In the face of deep uncertainty, it’s completely appropriate to try and be resilient to extreme events, and completely inappropriate for the media to portray these possible extremes as likely to happen.
So let’s hope that we are actually faced with a Reasonable Best Case Scenario, in which the actions taken by the public and authorities prevent the virus achieving its potential harm.
Our new podcast RiskyTalk is now available, featuring eminent guests discussing the task of communicating about risk and evidence across a wide range of challenging areas, starting with episodes on climate change, stats in the headlines, and genetics.
