Every day, CBS receives municipal records with details of deceased local residents. Up until recently, weekly mortality figures were published on a fortnightly basis. These figures become available with a delay: they provide complete death counts up to two weeks previously. Ms Traag explains: ‘This is because it takes a little time before each municipal death record is entered into the local registers, released to CBS and then processed. As a result, we do not have the most up-to-date information available immediately. However, we’ve been compiling this particular statistic for such a long time that we do have a fairly good idea of the time it takes before the municipal data arrive. For each consecutive week, we are able to review how much percent was received at any time of the week. This allows us to produce reliable estimated death counts over the previous week, before all the data come in. We have now adopted this approach so we can present last week’s figures more in advance.’
‘Under normal circumstances, we will have about 80 percent of the data available from one week previously,’ Ms Stoeldraijer explains. ‘It’s slightly more at the moment: about 85 percent. Since we know this, we can estimate the total number of deaths over the past week. We’re also able to tell what part of the data we normally receive, broken down by several different background characteristics. We can thus produce estimates of the total death count as well as deaths by particular categories, such as by province, by sex and by age. The latter is split into three age groups: 0–64 years, 65–79 years, and 80 years and over. We now publish these counts over the past week every Friday. These are robust estimates. In addition, we publish the death counts over two weeks previously, including counts per municipality. CBS often receives requests for daily estimates; however, the percentage of available data and the daily death counts tend to fluctuate too much.’
The figures published by CBS show how much excess mortality is being measured throughout the coronavirus crisis: ‘First, we calculated the average weekly mortality for the first ten weeks of 2020,’ Ms Traag explains. ‘There were no coronavirus or other epidemic outbreaks in this period. The average more or less represents mortality under normal circumstances as we could expect without the coronavirus outbreak. The difference between this average and the actual weekly death count at the moment is what we refer to as excess mortality.’ Can all the excess mortality be attributed to coronavirus? ‘We are not completely sure, but the data do point in that direction. We are observing how the excess mortality corresponds with geographical patterns in coronavirus contamination, and the demographic characteristics also run parallel.’
Partnership with RIVM
As it appears, discrepancies exist between figures from CBS and figures from the National Institute for Public Health and the Environment (RIVM). Why is that? Traag: ‘CBS is working closely with RIVM. They publish their mortality monitor weekly, also using CBS figures. Furthermore, our news releases are being compiled in collaboration with RIVM. The only difference is that they use a different publication week, but the underlying death counts are the same. In addition, RIVM publishes data which they obtain from the municipal health services (GGDs). These are the deaths that have been confirmed as COVID-19 deaths. The number deviates from the number of excess deaths because CBS reports the total number of deaths, including of those who were not tested but who may have died from COVID-19.’
Throughout the coronavirus crisis, Statistics Netherlands publishes the most recently available figures on mortality every Friday.