Skip to content Go to Homepage
Insights: Blog
Proxy measures of Covid-19 infection rates
by Prof Mark K. Smith

The greatest challenge of the coronavirus pandemic is two-fold; firstly, sufficient capacity of healthcare professionals to deal with severe cases – the ‘flattening the curve’ approach if you like, whereby social distancing can alleviate spread and contain the peak, when it inevitably comes. Secondly, determining how long the lockdown has to last.

The former is reasonably predictable – the ratio of diagnosed cases to death allows governments and healthcare professionals to, if not cope perfectly, at least size the problem. The way technology can be used to help with these predictions is wonderful and special respect goes to Johns Hopkins University’s map of cases worldwide.

The latter – how long the lockdown will last – is a much greater problem and needs some good lateral thinking. I have read ideas around issuing of virus locaction tracking apps, augmented with QR codes, or Bluetooth signals in GPS ‘not-spots’. I think it doesn’t take long to dismiss this as a: too late, b: too silly, c: missing the most vulnerable – the elderly who are not app users anyway. Just a quick check on Ofcom’s ‘not-spot’ maps shows you just how many hundreds of square miles this would miss.

Now you also have a nation (many nations) in crisis where governments are trying to prop up their economies with various ‘furloughing’ and business loan ideas. Last week, British Airways put 30,000 of their people on furlough. Their expectation is that people will be able to draw down a part of their salary from the government – sadly, but understandably, this is capped at £30k per annum and 80% of that as a maximum. I’d imagine pilots might be somewhat better paid. 

So what governments desperately need to know is what’s the ratio between deaths and the mildly affected? If they know that, then the concept of herd immunity comes in. If, say, the ratio was 100 mildly affected to one death, then in the UK with (tragically) around 3,000 deaths you can then postulate that 300,000 are mildly infected. If that figure is 1000 then it’s 3 million.

The ever wonderful Wikipedia defines herd immunity as: ‘a form of indirect protection from infectious disease that occurs when a large percentage of a population has become immune to an infection, whether through previous infections or vaccination, thereby providing a measure of protection for individuals who are not immune’.

So lockdown duration is more about mathematics than anything else. Once sufficient members of the population have acquired natural immunity (vaccines are at least a year away and Covid-19 is of the same viral family as the common cold which still has no vaccine) then we can all be released back into the wild.

How do you do that? Test 65m people? Mmmmm that’s not happening in the UK. We are still not testing enough healthcare professionals, which means up to 25% are self-isolating and the majority will not have it. It is even worse in the US, where they are much further behind.

Ask people? Given unemployment is rising at a rate that has never been seen before, direct questions about ‘are you infected’ or ‘have you been infected’ may well introduce bias as people will want to say whatever they need to say to keep going financially.

Try apps? Don’t be daft. I have 200 app on my phone and use maybe 10. My 87-year-old father has a phone with the same number of apps and uses, oh, none. My mother has not got a phone.

Here’s an interesting thought – and I bet ContactEngine is not alone holding this kind of data – right from the get-go of the pandemic, our clients (telcos, logistics companies, banks and so on) needed to ask their customers if they were feeling well because they needed to meet with them to repair, install, discuss, deliver or whatever. They asked a very direct question about Covid-19 and self-isolation – but it also became more nuanced, as people were self-isolating for different reasons.

I think those replies would be less biased than if you asked people directly as a part of a survey (which sadly UK Government would probably do with another post drop – I have 6 at risk people in my family and none have yet received their letter]. How daft is that in the modern era? I mean it is literally so 18th Century.

So out of curiosity, some of our smartest mathematicians started to look at the data we held, and it told a fascinating story. From a fairly decent data set, we could show a percentage of people in a household who thought they had Covid-19. The still wonderful Wikipedia suggested that 2.1 people live per household in the UK (compared to 2.6 in the US and 4.9 in India).

When we crunched the numbers and extrapolated up, we came up with a much larger number than you might imagine. Now it is not for us to publish this, the maths need checking and checking again, we need more data (sadly more and more is coming in) and we are not the Government, but it was quite reassuring because, if we are anywhere near right, the chances of dying from this nasty blighter are lower than you might imagine and herd immunity is gathering pace rapidly so we might be able to be released back in to the wild sooner rather than later.

Now if anyone is interested in contacting me on this please do so on and we can share our working-outs.


Solutions in your sector


Enabling a differentiated experience throughout the end-to-end customer journey for the world’s largest telcos

Retail Banking

Improving lending and mortgage journeys and making day-to-day banking easier for your customers


Transforming the citizen experience across federal and state government agencies


Reducing churn with better customer engagement across sales, onboarding, retention and regulatory compliance


Improving customer and agent experience through automation of onboarding, in-life customer care and loss prevention


Ensuring right-first-time delivery and installation of complex and large item deliveries

Retail/Major Appliances

Improving customer engagement to ensure right-first-time delivery and repair through proactive conversation


Managing and improving customer conversations across complex networks of manufacturers, dealers and customers

Icon Book Demo

Isn’t it time you stop assuming? Book a demonstration with us

Thank you, we will be in touch shortly!

Error Submitting Form

Back to top