Why comparing Wales’ coronavirus numbers with the rest of the UK may be misleading

In the current global crisis we find ourselves in, the huge variety of information sources available can be overwhelming. They can also be somewhat meaningless without any context so it seems natural to combine data from various sources to build up the whole picture, and combining data from the four UK nations is fairly common.

One such example features in The Guardian, and the way the data is laid out means it’s easy to make comparisons. Looking at how much darker the map of Wales is or when you see that seven of the top 10 areas for number of cases are in Wales, you would be forgiven for thinking that Wales has been hit with a disproportionately higher number of cases. But it’s worth stopping and considering whether we’re comparing apples and oranges.

The first thing to establish in any case is whether we’re measuring the same thing. The logical assumption would be that anybody with a positive test is counted in the case figures, and the information presented by Public Health Wales suggests that this is mostly the case, with a note indicating that tests done in all settings in Wales are included in the data. The only tests not included in the numbers are those done by commercial labs with swabs obtained from drive through testing centres in England and from home testing kits.

At the time of writing Wales are reporting a cumulative total of 14,356 cases, and there are also 485 extra positive tests shown in the commercial lab data. The note explains that these tests may include repeat tests or may contain people who were already tested in Wales, so we can’t be 100% certain about the numbers, but assuming a worst-case scenario that all the positive tests are new cases, it means Wales are missing just over 3% of confirmed cases from their numbers.

Higher and lower

Now let’s turn our attention to the UK data. The UK Government’s dashboard shows that there have been a total of 284,868 confirmed cases, but a note below states that this total includes commercial tests that are not in the national totals. If we add up the national totals and compare, this gives us 95,439 cases that have not been allocated to any nations – a third of all confirmed cases floating in the ether with no way of knowing where these extra cases are.

What this means in practice is that national and regional figures within Wales are lower than the actual number of confirmed cases, and the case numbers should be on average around 3% higher. Meanwhile, the national and regional case numbers for the UK as a whole should be an average 50% higher.

This is not to say that Wales hasn’t had some bad outbreaks – we definitely have. But when statistics from four different organisations backed by national governments with their own decision making powers over testing, data collection and analysis are grouped together, it’s always worth taking a moment to consider why the numbers might be different beyond the seemingly obvious reasons.