This release is categorised as a CSO Frontier Series Output. Particular care must be taken when interpreting the statistics in this release as it may use new methods which are under development and/or data sources which may be incomplete, for example new administrative data sources.
The statistical product “Mortality Indicators and Cause of Death by Ethnicity and Deprivation, 2017 - 2022” is designed to fulfil a statistical need to provide greater insights in terms of health and mortality by these two dimensions.
The statistical product extends existing statistical publications on deaths and cause of death to incorporate ethnicity and deprivation, which in turn are broken down by gender and age group. Both mortality rates and cause of death information are included.
This product is possible through linking Death Registration data with Census 2016 data (and in turn Deprivation Index 2016 data) to obtain the additional information needed – deprivation and ethnicity.
Death rates (Mortality) are calculated by dividing the number of deaths by the population (with scaling for ease of reading, per 100,000 persons). The Census 2016 usual resident population is taken as the population base and aged forward for each reference year, adjusting for births and deaths. No adjustment is made for migration (it is assumed migration has a negligible impact on overall figures). In summary, number of deaths is considered the numerator and the population base the denominator of death rates.
There are three primary data sources, Census, Death Registration and the Department of Social Protection's Child Benefit records, linked to create the analysis dataset for this statistical product. All linking is undertaken using Protected Identifier Keys (PIKs) which serves the purpose of linking capabilities across data sources and over time while protecting the identification details of individuals. The PIK used in this product is based on pseudonymised version of the Personal Public Service Number (PPSN).
The COPA is a pseudonymised copy of the Census of Population 2016 dataset held by the CSO for analysis purposes. It contains Census information for individuals and households of which approximately 94% of records have a suitable PIK. This allows them to be linked to similarly pseudonymised administrative data sources to create new analysis. Ethnicity is collected directly in the Census 2016 while a deprivation quintile is assigned using the Trutz-Hasse score given to the small area the person is usually resident in. For each reference period (2017 onwards), the COPA is aged forward and adjusted by removing deaths and adding births to give an appropriate analysis dataset. Births are added in based on those in receipt of child benefit and their ethnicity is determined based on the ethnicity of the person in receipt of child benefit, where identified in the COPA. All linking and analysis is based on pseudonymised data with identifiable details protected at all times.
6.5% of records in Census 2016 had an unknown ethnic or cultural background as not all persons completed the Ethnicity question in Census 2016.
Deprivation score was determined in >99.9% of records.
The DRA is the analysis data file of registered deaths. It is again pseudonymised and has the same PIK applied as the COPA for linkage with COPA and other administrative data sources. Ethnicity is determined on DRA through a linking exercise with COPA. In 2017, 22.7% of death records had unknown ethnicity. This could be for a number of reasons:
Pseudonymised data based on the Department of Social Protection's Child Benefit records was used to add births over the period onto the population base. Child Benefit (previously known as Children's Allowance) is payable to the parents or guardians of children under 16 years of age, or under 18 years of age if the child is in full-time education, Youthreach training or has a disability.
Deaths are classified to the World Health Organisation (W.H.O.) International Classification of Diseases, 10th Revision (ICD-10).
The Underlying Cause of Death (UCOD) has been defined by the World Health Organisation (WHO) as
(a) the disease or injury which initiated the train of morbid events leading directly to death, or (b) the circumstances of the accident or violence which produced the fatal injury’, and is selected for routine single-cause tabulation of mortality statistics.”
Further information on ICD is available.
The international form of Medical Certificate of Cause of Death is split into two parts: Part 1 is for diseases related to the chain of events directly leading to death, and Part 2 is for other significant conditions contributing to death. Other information in the form is also used in identifying the underlying cause of death which we use for tabulation after applying WHO coding instructions for mortality.
In ICD-10 classification ‘attributed to’ is an allowable synonym of ‘Due to’ and is reserved for categories where two conditions are mentioned, and a causal sequence exists between them.
In relation to neoplasms, the terms neoplasm and tumour are used synonymously as they both refer to abnormal cell growth. This abnormal mass of tissue/cells occurs when the body’s cells grow and multiply uncontrollably, usually forming lumps of tissue called tumours. These tumours can be benign (not cancer) or malignant (cancer). Benign tumours typically are slow growing and, unlike malignant tumours won’t spread (metastasize) to secondary sites in the body. However, there are specific types of benign tumours that can over time turn cancerous.
In contrast, malignant tumours (cancerous neoplasms) are usually rapid growing and invasive, breaking off from the original tumour and then spreading to other areas of the body via the bloodstream or lymphatic system, sometimes even per continuitatem to other organs or bones such as the pelvis or spine.
Neoplasms in ICD-10 are grouped into the categories: “Malignant Neoplasms”, ”In situ neoplasms”, “Benign neoplasms” and “Neoplasms of uncertain or unknown behaviour”. Further information on neoplasms is available.
Rounding of Deaths (to the nearest five) and Persons (to the nearest 10) was employed as a simple standardised Statistical Disclosure Control exercise for this project to ensure confidentiality. This means that extra care needs to be taken with small numbers in the reported statistics.
An undetermined category was added to the statistics to inform users of uncertainty associated with data linkage.
Death records not joining the census (and therefore not being assigned Ethnicity or Deprivation) from the previous years were included in the undetermined category and removed from the total figures, therefore there are negative numbers in some cases for the undetermined population category.
The Deprivation Quintile associated with a given person is the Deprivation Quintile of the area that person lived in at the time of the Census. In the case of births over the period, the Deprivation Quintile of a child is taken as the Deprivation Quintile of the area that the person in receipt of Child Benefit for that child lived at the time of the Census.
When making comparisons between groups, it is important to consider potentially differing age profiles. If a particular group has a younger age profile, people will have a tendency to die less frequently.
The category for Roma was only included in Census 2022 statistics and therefore could not be accounted for in this release.
It should be noted that this release is not stating that being from a deprived area - or being a traveller - does not cause death. The release does highlight, however, a statistical relationship between deprivation status or being a traveller and mortality.
The age group 40-64 was chosen to focus on in the key findings of this release due to the fact that deaths in this group often represent what would be considered premature deaths, as well as the fact that the differential between different deprivation/ethnic groups is sometimes the most prominent for this group.
Learn about our data and confidentiality safeguards, and the steps we take to produce statistics that can be trusted by all.