by Tom Chivers
Thursday, 5
March 2020

What are the odds a sneeze will give you coronavirus?

by Tom Chivers
The UK has tested about 16,000 people, so about 0.5% of those tested were positive. Credit: Getty

I had a notable experience on the Tube the other day. It was lunchtime, not very busy. A man, standing at one end, sneezed. A woman near him immediately got up and walked away.

My immediate assumption was that she was worried about coronavirus. I may have been wrong, but let’s say I’m right. I wondered: very roughly, what are the chances of that man having had coronavirus?

(Note: this isn’t meant to be medical advice, or to tell you that it is not rational to avoid people who are sneezing. It’s just in case a ballpark estimate might be helpful.)

I’m writing just after 32 new cases were confirmed in the UK, for a total of 85. That’s likely to be a major underestimate. They have tested about 16,000 people, so about 0.5% of those tested were positive; we could scale that up, meaning 0.5% of the UK population of 70 million – about 350,000 – would have it. But that would obviously be a huge overestimate, because they’re testing the most at-risk people. So we have the bounds of the problem: no less than 85, but no more than 350,000.

I’m going to arbitrarily assume that there are about 100 times as many cases as we’ve found so far. That’s two orders of magnitude higher than the confirmed cases, but two lower than the upper bound.

That would mean 8,500 people with coronavirus in the UK. Let’s assume they sneeze as often as people with colds and flu, and healthy people never sneeze. The average adult gets 2-3 colds a year and they last a week to 10 days. Let’s take the lower estimates and say 14 days with colds a year, so 4% of population has a cold at any one time. (We’ll ignore flu because it’s less common.)

Four percent of 70 million is 2,800,000 people. So if my 8,500 number is right, then any given sneeze on the Tube has about a 0.3% chance of being coronavirus-related. And these numbers are pretty conservative: we’ve taken low estimates of cold prevalence, ignored flu, and imagined healthy people never sneeze. Even if only one in 1,000 cases of coronavirus have been detected, the absolute risk that you’re sitting next to one of the undetected ones is still low, for now.

Again: this isn’t meant to say there’s nothing to worry about. The virus seems to be spreading very fast, and these numbers will change very quickly. Plus the potential impact of catching coronavirus is much higher than that of catching a cold, so it is rational to be more worried about it. Coronavirus will probably have real and severe impacts on Britain in the near future.

But right now, if you’re on the Tube or the bus, and someone sneezes, you probably aren’t at major risk. You may still wish to move away from them, and please keep washing your hands, but at this stage, a sneeze is probably just a sneeze.

Join the discussion

To get involved in the discussion and stay up to date, join UnHerd.

It's simple, quick and free.

Sign me up