7 Day Case Rate, Positivity Rate, Serious Case Rate and Vaccination Rate across Europe on (Friday) 22nd October 2021

2 comments

  1. **Data Sources**

    The source for the case data and positive testing data is here: https://www.worldometers.info/coronavirus/

    The source for the vaccine data is here: https://github.com/owid/covid-19-data/tree/master/public/data/vaccinations

    I’ve had requests for the spreadsheet, so here is the newest one as of 19th July: https://drive.google.com/file/d/1WqpGTxhWI-Nm3gK-xUthzq3ClZ8vl1Um/view?usp=sharing

    I won’t be keeping these versions up to date but you can see how the data is entered and calculated.

    **Version Notes**

    The ecdc stopped updating data on a daily basis as of 14th December. I figured weekly data wasn’t particularly useful and while the worldometer data is not as accurate as an official agency, at least it is daily.

    Just so you can see the extent of the variances, here is the data for the 12th, 13th and 14th of December between worldometer and ecdc: https://imgur.com/Y71Nx7o

    Some of this will be due to delays in data reporting and methodologies, others will be the source. In order to do a more detailed comparison, you need to look at the data sources on the ecdc and worldometer sites.

    I’ve added vaccination data as of 23 March 2021. This data is sourced separately and is also not always fully up to date – as such, there’s a column which includes the data date. Note that I will be reporting as per the data collected by “Our World in Data”.

    I’ve started including serious case data as of 16th June 2021 as it is relevant to whether the vaccines are working.

    **Schedule**

    I will post this chart daily (where possible) during the week, Monday to Friday but I will not update it at weekends because around half of the countries do not update their numbers over the weekend and also it’s on a computer I would not otherwise be logging onto, unless there’s something particularly notable which requires a weekend update.

    **Explaining some numbers**

    *What is the 7 day case rate?*

    Imagine a football stadium with 100,000 people in it.

    The 7 day case rate number in the graph is the number of people in that stadium who would have a coronavirus infection a week later. For example, in the UK on the 8th November the number was 239.7. That means 240 out of 100,000 people in the stadium would have an infection the following week.

    This number accounts for population size, so you can compare country-by country

    *What is the daily movement*

    The daily movement numbers are just how the 7 day case rate is moving up and down each day. The 7 day trend is the most interesting number – it’s the sum of all the ups and downs. If you think of the infection as a car then the 7 day case rate is the speed of the coronavirus infection and the 7 day trend is the acceleration (or deceleration) of the infection. If the 7 day trend is red it’s accelerating, if it’s neither red nor green it’s holding steady and if it’s green it means it’s slowing down (note – a green 7 day rate doesn’t mean the infection is shrinking, just that it’s slowing down).

    This number accounts for population size, so you can compare country-by country

    *How can I tell if the infection is shrinking*

    You can’t infer that from this data, this data only looks and new cases and the rate of change in new cases. The other side of the equation is how many people are recovering from coronavirus in each country. If you think of it like a bath – this is how much water is flowing through the taps. You cannot tell how much water is going down the plughole so you can’t tell how much water is in the bath (i.e. the infection level).

    *What is testing rate per 100,000?*

    Imagine a football stadium with 100,000 people in it.

    The testing rate is how many of those 100,000 people will have received a coronavirus test in a 7 day period.

    This number accounts for population size, so you can compare country-by country

    *What is the positive test rate*

    This is the % of tests taken which are coming back positive. The WHO says countries should aim for this to be 3% or lower and that testing not keeping up with infections if the rate is above 10%. If you don’t have enough tests, you restrict access to them to people who are more likely to be infected either by their symptoms or their risk of exposure, and that inflates your numbers.

    This number is a ratio of tests completed, so you can compare country-by country

    *What is the 3% adjusted case rate*

    Used to be the 10% rate but has been rebased around 3%. There is also a lower limit at 1% where the impact will be reduced by a factor of 5.

    Countries with a higher positivity rate are not testing enough of their population, so this multiplies their most recent day’s case rate by a factor of their positivity rate divided by 3% – for example on 6th April Greece’s positivity rate was 6% so their daily case rate was multiplied by 6%/3% = 2

    **This is not** an estimate of the true scale of the pandemic within the country** but is a proxy to rank countries by accounting for countries who are showing a lower positive case rate because they’re not doing enough testing.

    This number accounts for population size AND level of testing within a country, so you can compare country-by country

    *What is the 7 Day Vaccines per 100,000?*

    Imagine a football stadium with 100,000 people in it. This is the number of people in the stadium who would have received a vaccine dose of that stadium in a 7 day period.

    This number accounts for population size, so you can compare country-by country

    *What is the 1st Dose %*

    This is the % of a population who have received their first vaccine dose. It covers total population which includes children so will be lower than the number some sources cite (% of adult population)

    *What is the 2nd Dose %*

    This is the % of a population who have received all of their vaccine doses to be considered vaccinated (includes single dose vaccines like J&J)

    *What are serious cases?*

    For most countries, these are the number of patients currently being treated for COVID in ICUs. The numbers are not a perfect comparison because some countries below the line (India, Brazil) don’t have ICU capacity, so they look low, but in reality they’re just not getting the right care to their serious cases. But for the purposes of comparing against Europe, it’s a fair yardstick.

    *What are serious cases per million?*

    This shows the number of people in the country currently counted as a serious case – most countries class a case as serious as being treated in ICU, but the definition may be changed by different countries.

    This number accounts for population size, so you can compare country-by country

    *What is % of cases that are serious?*

    This is the number of serious cases divided by the 7 day case rate **from a week ago**, which is a proxy for how many cases are progressing from infection to a serious case. This is not a perfect metric because serious cases may have originated earlier than the past 7 days, but it’s a good enough proxy because every country is being calculated in the same way

    This should reflect how well the vaccination efforts are going, because in an unvaccinated population with ICU capacity, this number should run broadly the same minus any drastic changes in population dynamics – older people are more likely to end up in ICU.

    This number may be unusually high in countries with low populations (e.g. Liechtenstein) or countries who are underreporting how many active infections they have (e.g. Romania).

    *What is the comparable unvaccinated rate?*

    This functions very similar to the 3% adjusted rate but for serious cases. I’ve gone back to pre-vaccine rates and looked at what the serious case rate on average was throughout December 2020, and it was around 3%. As such this normalizes the 3% adjusted rate to what it would be if the serious case rate were 3%.

    For instance in the UK on 4th July there were 3.9 serious cases per million – without vaccines that would be the level you’d expect if the 3% adjusted case rate was 9.7 per 100,000 as opposed to the UK’s actual rate of 256 cases per 100,000. This is indicative of a country that has a high vaccination rate where case rate is less important.

    This number accounts for population size, so you can compare country-by country

    **Why sort by the comparable unvaccinated rate?**

    Look at Poland on 9th November as an example – their actual 7 day rate was not actually much higher than the UK’s (200 vs 160) and is lower than Denmark’s (221), but their positive test rate is 49% – basically 1 out of 2 tests came back positive compared to 5% for UK and 2% for Denmark. Poland only test a fraction of their population compared to the UK or Denmark (706 vs 3198 / 8658 per 100,000) so their actual rate of infection is going to be much higher – their 7 day rate is just kept artificially low by the amount of tests they’re (not) doing.

    Likewise we are moving to a world where case rate is less important and hospitalizations and deaths is a more critical metric – the comparable unvaccinated rate adjusts the score downwards for countries with low levels of serious cases, which is reflective of high levels of vaccination. As such this metric captures everything.

    This method is **not perfect** as countries account for positivity differently. For instance, some countries (e.g. UK, Italy) include lateral flow tests in overall testing rates and positivity whereas other countries (e.g. Germany) only include PCR so their rates will appear higher and testing rates will appear lower. In addition, countries may alter the way they account for serious cases. While it’s important to use this metric to help smooth out countries not doing enough testing and relaxing restrictions in the face of high levels of vaccinations, remember that it’s not perfect.

    There is no one unbiased source and you should seek to get your information from a range of sources and assess the underlying bias in the information you’re reading and the possible agenda that source may have.

Leave a Reply