How do you collect COVID-19 data?
Reuters collates and checks coronavirus infection and death counts from countries and territories around the world each day. We also review that data for irregularities and correct it whenever better data becomes available.
Where does the data come from?
The figures largely come from official country, state, county and territory government and public health department websites. We also occasionally get information from news conferences, press releases and verified tweets and social media posts by government officials. In cases where the national count lags behind state figures, we may use those local figures to arrive at the total. In other cases, we may defer to global health agencies such as the World Health Organisation or the European Centre for Disease Prevention and Control for missing historical data when the primary source does not make historical data available.
Total infections and deaths for each country include both confirmed and probable infections and deaths where that data is available. If probable infections and deaths are not reported by a country, we only use confirmed infections and deaths.
Is this data a complete count of every COVID-19 infection and death in all countries?
No. The counts are based on the best available information accessible to Reuters at any given time. Countries may miss infections and deaths for any number of reasons, such as delays in processing testing data, lag in reporting, lack of testing, etc. Some countries may exclude certain categories of infections and deaths that others include (for example, probable infections) or have different criteria as to what may be counted as a COVID-19 infection or death.
Why does the data change sometimes?
Some countries change the way they’re reporting to include or exclude a certain category of infections or deaths, which can lead to a significant increase or decrease in their total and what they reported previously.
Are these counts the latest?
While most countries report new numbers daily, others may only update on weekdays or even less frequently. We use a rolling average of the infections and deaths reported over the last 7 days to even out some of that irregular reporting and get us closer to a daily figure.
Why do you use “percent of peak” to describe a country’s outbreak?
The percent of peak number featured throughout this project represents where a country is in comparison to the highest point of its own outbreak.
As a simple example, if a country was reporting about 100 COVID-19 infections each day, but is now only reporting 25 on average, that country is reporting 25% of its peak number of daily infections. That tells us that the country has come a long way from the height of its outbreak, but still has a way to go to fully contain the virus’ spread.
You can use the percent of peak to compare countries because that figure is relative to the scale of each country’s outbreak, but there are some important caveats.
Say there are two countries of equal population size, both likely initially to suffer similar outbreaks. Let’s say one of those countries acted much more aggressively to contain its outbreak early on and therefore had a much smaller initial peak. Even though this country may be reporting fewer COVID-19 infections now than the more lax country, the proactive country’s percent of peak may be higher because its initial peak (the denominator) was smaller.
It works the other way, too. Let’s say a lax country responds strongly after its initial peak and halves the number of infections it now reports. That may represent a much larger change than a 50% decrease in the more proactive country from its smaller peak and, in absolute terms, means far fewer people are currently affected by the virus in the initially lax country. It’s all down to the denominator, how big that initial peak was.
There is another important caveat: If a country changes the way it reports infections or deaths, that may mean the current number it reports isn’t truly comparable to its previous peak any more. Say a country decided to intentionally minimize the number of infections it reports after a very high initial peak. That may mean the smaller numbers don’t really reflect a meaningful improvement in the outbreak and would make that country look artificially better in comparison to others.
With those caveats, we use percent of peak to highlight trends in countries regardless of the size of their outbreak. Watching the percent change tells us when countries are approaching a new peak and, in most cases, gives a good sense of how the current situation compares to an earlier stage of the outbreak.
How are countries, territories and regions defined?
We use International Organization for Standardization definitions for countries and territories and the United Nations standard area codes for statistical use for world regions.