The problems with R as a single covid19 KPI

Roadmap reliant on covid19 KPI of R

Like a good leader, Boris Johnson set out his covid19 KPI (key performance indicator) on May 10th of R. Since then there have been few updates on where England is on R. While there is some confusion of what R actually means. Indeed, mid-week the biggest media news item was the standard R of the UK not unexpectedly facing a big Recession!

Good managers not only choose KPI’s that are easy to understand, but also ones that are easy for people to measure.

What is R?

R is the rate of Reproduction of the virus in the population. It is used by those working with anything viral to predict the trend and length of an epidemic. If the number is less than 1 then the epidemic is on the downward trend. How long that trend will take is seen by how low the number. If 100,000 people are infected with an R of 0.5 it will take 16 months to halt the spread of the virus. At 0.9 it takes 86 months.

This is where the issues begin. In his address, Boris stated that R was somewhere between the two. Why? Because R is actually quite difficult to calculate and depends on the available data. While R could therefore a good expert metric, it’s not a good covid19 KPI for general use because it needs both statements of confidence in the figure presented and good input data.

Consider Infection

Key data required is the number of people infected (I). To achieve this requires reliable infection testing in significant numbers. Hence the 100,000 test per day target by the end of April. Again, a great KPI when there is no attempt to manipulate figures to achieve a target, not the progress it’s actually showing.

Put 100,000 tests into the context of the England population of 56M. This would mean 0.18% of the population is tested. At this rate it would take 560 days for the whole population to be tested once. While everybody does not need to be tested to predict R; the current number is too low to state a reliable I for the whole population. Or to identify variations such as location.

An old KPI was Death

For weeks, the news led on the numbers of deaths reported today. Latterly this has included all deaths where covid-19 has been cited, which overcomes the ‘data issue’ of only those in hospital. But the KPI aim was the need to prevent the NHS from being swamped. This is what the death rate measured not as people understood the total number of deaths in the country from covid-19. (For a discussion on real death rates read )

The second issue on this KPI was its announcement as a single figure rather than ongoing trend graph which might smooth out variations created by delays in reporting deaths. Telling me 200 deaths were reported today, requires me to remember how many were reported yesterday and last week. This isn’t easy, so it’s difficult to judge performance.

Displaying data is important

From lockdown, control has been created over the number of people becoming infected. As we emerge from lockdown, opportunities increase for people to meet others and become infected. This can often be from those who are infected, but display no symptoms. The numbers infected will increase and with it therefore R and probably deaths. What is necessary to accept is a growth in R then a fall which prevents NHS swamping, but enables the Economy to restart.

The Johnson roadmap assumed a consistent downward trend without any renewed peaks in R. Why? Because subliminally it needed to demonstrate that the situation is under control. This just delays the honest discussion of no need to fear if R rises slightly for a short period.

Management & Marketing Conclusions

Choosing a small number of KPIs to manage is essential. Such KPIs need to be easy and reliable to collect and specific as to the aim they are targeting. They should be shared truthfully and effectively to provide real information. Here it’s the trend that is important over the longer term not a single number, with the ultimate aim measured in months/years not weeks.

In the current stage of the covid-19 crisis, what is still key are the trends in:

  • R (is it under control),
  • I (what proportion of the population is infected currently based on this number of tests)
  • Deaths (how severe is the impact).

As a marketeer, I can’t resist the cheesy acronym of RID for the 3 necessary covid19 KPI for how soon we might be free of the terrible consequences of this virus; as in ‘together we can get RID of covid-19’.