Growth: the doubling time of COVID-19 cases

In an outbreak of an infectious disease it is important to not only know the number of cases, but also the rate at which the number of cases is changing.

A helpful metric to measure the rate of change is the doubling time. This is the time it took for the number of cases to double. If the number of cases is 100 today and it was 50 three days ago then the doubling time is 3 days.

The doubling time in a disease outbreak is not constant and for the outbreak of COVID-19 it has changed in recent weeks and will continue to change with time.

What we would want to know but do not know: the doubling time of total cases

What we want to know is the total number of COVID-19 cases. 

But since not everyone is tested it is only the number of known cases (confirmed cases + suspected cases) that we can know. Since the coronavirus disease has only mild symptoms for most (see below) it is likely the case that the number of confirmed and suspected cases is smaller than the number of total cases.

What we do know: the doubling time of known cases

The WHO is publishing daily Situation Reports (here) which list the number of confirmed cases. But the WHO does not publish daily updates of their doubling time.

The volunteer editors of Wikipedia, however, have calculated the doubling time of cases of COVID-19 and update these calculations daily. They can be found here.

Because these calculations are straightforward and the Wikipedia editors rely on the latest WHO Situation Reports we trust this source. We therefore are not duplicating this work by calculating and reporting the doubling times ourselves and rely on these calculations instead.

Based on the global WHO data up to and including 9th March 2020, the doubling time for COVID-19 is as follows:

Doubling time for the global number of cases (including China): 19 days

Doubling time for the global number of cases (excluding China): 5 days

The fact that the doubling time is longer when China is included is due to the fact that the number of daily cases has declined after the lockdown in China.

Exponential growth leads to large numbers fast, but most of us do not have a good intuitive grasp of this

As we said previously, the doubling time will change and it would be wrong to make projections based on a constant doubling time. But it is important to remind ourselves of the nature of exponential growth.

If during an outbreak the number of cases is in fact doubling and this doubling time stays constant, then the outbreak is spreading exponentially. 

Under exponential growth 500 cases grow to more than 1 million cases after 11 doubling times. And after 10 more doubling times it would be 1 billion cases.

This is in no way a prediction for the number of cases we need to expect; it is simply a reminder that exponential growth leads to very large numbers very quickly even when starting from a low base.

And it is important to be reminded of the nature of exponential growth because most of us do not grasp exponential growth intuitively. Psychologists find that humans tend to think in linear growth processes (1, 2, 3, 4) even when this is not appropriately describing the reality in front of our eyes. This bias – to “linearize exponential functions when assessing them intuitively” – is referred to as ‘exponential growth bias’.

Psychological research shows that “neither special instructions about the nature of exponential growth nor daily experience with growth processes” improved the failure to grasp exponential growth processes.

License: All of Our World in Data is completely open access and all work is licensed under the Creative Commons BY license. You have the permission to use, distribute, and reproduce in any medium, provided the source and authors are credited.