Excess deaths or excess mortality is defined as the number of deaths above the expected number of deaths. The expected number of deaths is derived using routine-mortality data prior to the timeframe for which excess mortality is estimated. [2–4]
To analyze excess mortality among children in Europe, we utilize data on the weekly reported number of deaths among children aged 0 to 14 from 35 European countries from 2015, up to latest date available in 2022. The data were obtained from EuroStat and Human Mortality Database’s Short Term Mortality Fluctuations open dataset [5, 6]. We begin with examining the annual number of children deaths up to latest date available in 2022 (Figure 1). For example, latest data in Austria is ISO-week 39 in 2022, which ends on October 2nd and we thus examine deaths in Austria in 2015-2022 up to week 39 in each year.
A simple visual inspection reveals that children deaths in 2021-2022 are significantly higher than previous years only in Luxembourg. In all the rest of the countries shown, deaths are at similar level or even lower than in previous years. Summing up data in all countries up to latest common available time (Figure 1, last panel marked “Europe”), children deaths in 2022 amounted to 13,950. This is lower than the number of deaths for the same time period in all previous years (e.g., 16,300 in 2016), except for 2021 (13,750).
In order to more formally analyze excess mortality as well as its uncertainty, we require a model to estimate expected deaths. We thus estimated for each country the model used by the World Mortality Dataset [3], fit to death counts in ages 0 to 14. The model predicts expected number of deaths for each week-year in 2020-2022 based on the observed number of deaths for each week-year in 2015-2019, and it accounts both for possible intra-year seasonality and annual trend in deaths in each country. The model also allows derivation of the uncertainty (standard error) in excess mortality estimates that is due to the uncertainty in the model parameters as well as that arising from the forecast.
The model results are shown in Figure 2, with vertical bars denoting the confidence intervals. The figure shows that excess mortality estimates are low and, in many instances, negative.
Out of 35 countries, there is statistically significant excess deaths among children in 2022 in only 2 countries (Luxembourg, Sweden), and only in 3 there is statistically significant deficit deaths (Belgium, Malta, Serbia). In all remaining countries, the estimated excess/deficit is not statistically different than 0. Summing across all of the countries up to July 31st 2022 (Figure 2, last panel), Europe as a whole has experienced 330 less children deaths than expected, though this deficit mortality is not statistically significant, with a standard error of 220 and a confidence interval ranging from -770 to 104.