Statbel, processed by Statistics Flanders
Deaths: number of deaths observed in the reference or observation period. As a rule, the Statbel statistics for Belgium only include deaths in the residential population, i.e. residents who are registered in the population registers of the municipality where they have their main residence (excluding persons registered in the waiting register).
Final figures on the number of deaths are only drawn up after the registration period has been closed. In practice, this is towards the end of March following the observation year (e.g. the end of March 2020 for mortality in 2019) because late declarations of deaths must be taken into account. This is followed by a control period in which the registered data are examined more closely and corrected if necessary (in the event of contradictions, impossibilities, etc.).
The monthly figures included are provisional figures. Provisional figures lack the waiting time and subsequent checks, until further notice.
Official statistics institutes provide data on demographic events (i.e. the number of deaths) based on 2 determinations for age (according to the 'double classification'):
- age according to the last birthday = number of past birthdays. Referred to as 'exact age' by Statbel; as ‘age completed (age at last birthday)’ by Eurostat – also recommended).
- age according to the year of birth = the number of whole years between the reference year and the year of birth. That corresponds to the age that anyone of the year of birth would reach at the end of the calendar year - if they survive. Referred to as 'age in elapsed years' by Statbel; as 'age reached during the year (= reference year minus year of birth of the deceased person)' by Eurostat.
Premature mortality: refers to the number of deaths before the age of 75 (according to last birthday).
Mortality rate: ratio between observed number of deaths and the population at risk - by age and gender. As a rule, the population at risk is the arithmetic mean of the population at the beginning and end of the reference or observation period. In certain cases (such as to determine 'projective probabilities' for the purpose of the population outlook) only the population at risk at the start of the reference period is taken into account.
Crude mortality rate: ratio between observed number of deaths in the course of the calendar year and the average population of the year, usually expressed per 1,000 inhabitants.
Average population of the year: Usually the arithmetic mean of the population at the start and at the end of the reference year is taken into account. Sometimes (as with Statbel) reference is made to the population halfway through the calendar year (on 1 July).
Excess mortality and mortality deficit
This is a relative concept where the observed number of deaths in the indicated observation period is compared with the expected number based on past observations. The expected value is a function of the precise delineation of the comparison period (e.g. average of the monthly mortality in the previous 5 years). A correction has also been made here that takes into account population growth and associated changes in the age and gender structure. To this end, the standardized mortality ratio (SMR), or the ratio between the observed and expected number of deaths, is determined. The expected values are based on the mortality probabilities in broad age groups (0-24, 25-44, 45-64, 65-74, 75-84, 85 and older) per gender (M/F), based on the observations in the selected reference period.
If SMR > 1, this (in principle) indicates excess mortality; if SMR < 1, this indicates a mortality deficit compared to the expected value (according to the chosen standard).
The outcome for SMR can be tested for statistical significance, allowing to decide whether or not that outcome can be attributed to chance. Statistical significance of SMR is determined here by checking whether or not the 2-sided 95% confidence interval around the SMR result includes the (zero) value of 1: if the lower limit of the interval is above 1, then there is statistically significant excess mortality; if the upper limit of the interval is below 1, there is a statistically significant mortality deficit (at α=0.05).
Here, 'excess mortality' is determined for the mortality (or number of deaths) per month in the course of 2020 and 2021. For 2021, these are 'provisional data' because the regular waiting time for closing the registration was not completed nor the additional checks on the data have been made.
Remarks on quality
For the statistics on the number of deaths, Statbel relies on the deaths registered in the National Register (of natural persons) during the year. Registrations are closed in March of the new year following the observation or reference year; this is followed by a control period (for double counting, inconsistencies, etc.) before publication (in May/June of the new year).
Statbel only considers the deaths of persons who had their legal residence in Belgium at the time of death and who are considered part of the legally resident population (excluding persons registered in the waiting register for asylum seekers/applicants for international protection). The registers for the legally resident population in the National Register are used (since 2010) as a reference source.
The statistics on the causes of death are based on legal death certificates (death registration forms from the registry office, partially completed by a certifying doctor - in a closed envelope - and then completed by the municipal administration of the place of death). The processing thereof is entrusted by Royal Decree to the regional authorities (regions and communities) of Belgium. The Agency for Care and Health is responsible for this for the Flemish government (under the authority of authorized physicians-civil servants). The coordination of all registrations for compiling the official death statistics is the responsibility of Statbel. Therefore the various federal and regional partners involved meet regularly to better align their activities.
The federal government reports periodically on the statistics regarding deaths to various international bodies, including Eurostat and the World Health Organisation (WHO).
The statistics of the source Eurostat refer to the 'usually resident population' and not to the legally resident population. To this end, Belgium makes an estimate based on the records in the waiting register.
Agency for Care and Health: Mortality, life expectancy and causes of death (ACH)
Statistics Flanders: Demographic components