Eurostat, the statistical office of the European Union

National Statistical Institute (Spain)

Social Statistic Department, Health Unit ( INE, Spain)

Data on causes of death (CoD) provide information on mortality patterns and form a major element of public health information.

CoD data refer to the underlying cause which - according to the World Health Organisation (WHO) - is "the disease or injury which initiated the train of morbid events leading directly to death, or the circumstances of the accident or violence which produced the fatal injury".

CoD data are derived from death certificates. The information provided in the medical certificate of cause of death is mapped to the International Statistical Classification of Diseases and Related Health Problems (ICD).

In case of deaths with judicial intervention the information is provided  by the forensic report.

Eurostat's CoD statistics build on standards set out by the World Health Organisation (WHO) in the International Statistical Classification of Diseases and Related Health Problems (ICD).

The regional breakdown is based on the Nomenclature of Territorial Units for Statistics (NUTS 2).

Classification and updates applied by years

Public Health.

Concepts and definitions are described in the Commission regulation (EU) No 328/2011 in articles 2 and 3.

The definition of usual residency used in Spain for Causes of Death Statistics is the residency declared on the medical death certificate. The information is provided by a relative and it cannot be checked with any Spanish administrative source for legal constraints.

Stillbirth is the death prior to the complete expulsion or extraction of a viable product of conception from its mother. The foetus is viable when the weight is at least 500gr or has a gestational age of at least 22 weeks.

The statistical units are the deceased persons and the stillborns, respectively.

Deaths that occurs in the country.

Neonatals of non-resident mothers are not considered residents.

We include  non-resident deaths in the spanish national statistic.The Statistic on Causes of death includes all deaths happening in national territory regardless of the deceased´s residency.

Spanish statistic on causes of death does not include residents dying abroad.

The measured statistical phenomenon relates to all national territory.

Time series available from 1994 onwards.

Not applicable.

The unit is number.

Data refer to the calendar year (i.e. all deaths occurring during the year).

The last reference year available is 2023

CoD data was submitted to Eurostat on the basis of a gentleman's agreement established in the framework Eurostat's Working Group on "Public Health Statistics" until data with reference year 2010.

A Regulation on Community statistics on public health and health and safety at work (EC) No 1338/2008 was signed by the European Parliament and the Council on 16 December 2008. This Regulation is the framework of the data collection on the domain.

Within the context of this framework Regulation, a Regulation on Community statistics on public health and health and safety at work, as regards statistics on causes of death (EU) No 328/2011 was signed by the European Parliament and the Council on 5 April 2011. 

CoD data according to this regulation is submitted to Eurostat since reference year 2011.

Data from Spanish Statistics Causes of Death are shared with the World Health Organisation (WHO) after a formal request. Some data are shared with OECD too.

Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

The tables published in the NSI  website consider classification variables that don´t allow the identification of a register. Five age groups are considered in the tables. From 0 to 4 the information is sometimes presented year by year but only at nacional level,so the confidentialiality is guaranteed.

 Data referring to year t are released in December year t+1 in the INE website.

The need of the information on 2020 deaths due to the pandemic situation changed the calendar for the availabilty of 2020 data. Provisional data referring to  January-May 2020 were realeased in NSI web site in December 2020 and definitive 2020 data ( whole year) were realeased in NSI web site in November 2021.

From December 2022 onwards, provisional data for the first half of year t are published in December year t in the INE website and provisional data for year t are published in June year t+1.

The release calendar is available at the beggining of the year at the INE website.

The data realise policy for the statistics on causes of death data is the same as the general data realise policy for any structural statistic. ( The publication month is announced at the beggining of the year).

The channels to inform the users:

1. Calendar published at the beggining of the year
2. Press release
3. Tables at the INEwebsite
4. Metadata at the INE website

Annual for final data. Half-yearly for provisional data.

Together with the annual publication of the data, a press release is published with the main results. The press release can be found on our website (provisional data for the first half of 2024).

Apart from the press release some infographics are pubished together with the data:

Infographic showing Daily deaths by place of occurence and sex

Infographic showing Top 15 causes of death in Spain

In 2023 a report called "Causes of death two years after the start of the pandemic" was also published.

The data of the statistics on Causes on death are released annually. Provisional data are published biannually. We publish a set of tables at national level and at regional level with different variables of classification (age group, sex, high level of education, place of residence...).

Apart from information on underlying cause of death we include some tables with information on multiple causes of death.

These tables can be found on INE website.

Some infographics are published together with the data too. In 2023 a report called "Causes of death two years after the start of the pandemic" was also published.

Both the press release and the tables can be found on INE website.

The microdata of causes of death statistics can be provided to users when they justify the need of the microdata. They have to inform about the study/project for which the microdata on causes of death are needed.

In case it is duly justified, they must sign a confidentiality compromise. The variable of place of occurrence and the variable of place of residence are only provided when the number of inhabitants is superior to 10.000 in order to guarantee the confidentiality.

Together with the main tables and the press release that are published on the INE website, a dynamic infographic is published with the most frequent causes of death at national and regional level.

In 2023 a report called "Causes of death two years after the start of the pandemic" was also published.

Apart from the data provided at the INE website the information is published at regional level in the oficial web site of the institutions involved in the statistic. (Regional Statistical Offices)

Users and experts disseminate data through different articles in scientific magazines.

At national level the methology and the metadata are available together with the tables on the INE website.

100%.

The methology mentioned in previous points that is publised together with the data is enough complete.

The team responsible for the Statistics on causes of death in Spain is a steady team with great experience that ensures quality in all process.

Many efforts at national level from the medical sector are made to garantee physicians to complete correctly the death certificates. IRIS automatic codification is used for all Regional Mortality Registers in order to garantee the quality and the comparability. The information received is continuously revise with validation rules and comparing with previous years in other to detect any problem as soon as possible.

The direct contact with users (Health administrations, researchers...) helps to receive their feedback in terms of quality, in general, they express their satisfaction for the high level of quality of the information provided.

The recent improvement is the inclusion of the forensics in the information circuit. They provide directly the infomation of deaths with judicial intervention using a software provided by National Stistical Institute.

First stepts are being taking  to implement e-death certificate, it will be a quality gain.

There has not been any quality assessment in the recent years. The assessment is continous.

The Statistic on causes of death is highly demanded in our country: Ministry of Health, Regional Health Administrations, Researches, Students, Journalists...

Apart from Eurostat we send data to international organizations such as OECD and WHO.

The Ministry of Health need the information of relevant indicators for Public Health interest. The information required by researchers is based on the project they are working on. Journalists demand information of interest at the moment in which they make the request.

If there is no confidentiality issues all the information demanded by users are provided by the NSI. The variables provided are the strictly neccesary for their needs.

The demand has considerably increased with the pandemic situation.

 We maintain fluid communication with users, we received their feedback constantly.

 The Statistics on Causes of death is compulsory, it is included in the National Statistical Plan.

1. For mandatory variables:

 All mandatory variables are sent to Eurostat

2. For voluntary variables:

 100%

3. For additional variables:

• External CoD: 100%
• Place of occurrence for external CoD: 100%  (69,0% with a code other than 9 (unspecified place)).
• Activity for external CoD

The procedures of data collection, coverage control and revision of the errors guarantee the accuracy.

The statistic on causes of death is based on the information declared on the death certificates by physicians. For legal reasons, all deaths must be register in the Civil Register. This fact guarantees the coverage.  The coverage control is carried out by the provincial delegation of National Statistical Institute.

The use of validation rules, both errors and warnings, allows to analyze the quality of the data. The codification is carried out by experts from the Regional Mortality Registers. These experts usually contact with physician in order to improve or clarify the information written on the death certificate. As consequence the information is more accurate.

Deaths with Judicial intervention are informed by the Ministry of Justice from the forensic report. From 2020 deaths this information is provided directly from the forensics, improving the accuracy in these cases.

Not applicable. Data collection is from administrative sources.

Not applicable.

 Non-sampling error such as demographic variables are solved using the data of the Population Register. In case of medical errors or missing information we try to contact with physicians or check personal clinical histories.

For legal reasons, all deaths must be register in the Civil Register. This fact guarantees the coverage. The coverage control is carry out by the provincial delegation of National Statistical Institute.

There is no over-coverage.

There is no under-coverage sauf some weird exception, when the Civil Register send the death certificate very late and the Civil Register is not computerized.

Not applicable. Data collection is from administrative sources.

Possible errors related with demographic variables can be recorded. These kind of errors are validated and corrected by NSI with Population Register information.

In case of missing information from medical part, Physician are contacted in other to complete the information on causes of death.

 The non response error is minimun. In such cases, we try to recover the information through the physicians, the hospital or other data bases

 The number of units from which it is not possible to obtain the underlying cause of death is 0.84%.

The number of units from which it is not possible to obtain the underlying cause of death is 0.84%.

In this statistic, unit non-response is the same as item non-responde

Processing error are not significant.

Not applicable.

Not applicable.

Not applicable.

From data collection with reference year 2011 onwards, Eurostat asks for the submission of final data at national and regional level and related metadata for the year N at N+24 months, according to the Implementing Regulation (EC) No. 328/2011, Article 4.

There is no problem of comparability between the regions in Spain. The reason is that NSI coordinates the statistic in the whole country providing the same guidelines and IT tools to all the regions.

For example, the NSI is the reponsible for the IRIS dictionary updates to garantee the comparability. All regions use the IT tools provided by NSI for data transmition, these tools include validation rules .

We have a forum integrating all Regional Mortality Resisters in which complicated cases that IRIS cannot solve are discuss to come to an agreement in the codification in other to garantee the comparability.

Not applicable.

 Apart from the ICD uptates that have influences in the comparability of the data, other issue is the change from manual codification to automatic codification IRIS. However, the bridge coding study carried out highlighted that the comparabily continued.

The time series of the Statistic on causes of death are comparable for the Underlying Cause of death when they are related to the same ICD clasification.

The Spanish NSI have data from 1900 always referred to internacional classifications. That means that for some causes of death the length of the series is 120 years (annual data).

However, from the avalaibility of the data (1900) there have been several revisions of the ICD.

The ICD versions used for the last years have been:

From 1968 to 1979 with ICD 8

From 1980 to 1998 with ICD 9

From 1999  with ICD 10

That means that the length of comparable time series with 4-digit codes is 25 years taking into account that from 1999 we are using ICD 10

However, taking into account the reduced list with aggregation of data and the equivalence between ICD 9 and ICD 10 it is possible to build comparable time series from 1980 (the lenght would be 44 years).

The coherence with demographic data is complete due to the fact that we work in a coordinated way.

We check information from the following datasets to confirm the coherence: Hospital Morbidity Survey, Mortal victims from General Traffic Department.

Not applicable, only annual data are available.

Not applicable.

There are no inconsistencies.

The statistical on causes of death is a complex statistic which involves several institutions and staff in the whole country. The cost and burden is reducing with the introdution of IT tool in every phase of the project.

Although we use paper death certificate, we use Optical Recognition System to get the information provided by doctors in the certificate. The information of deaths with judical intervention are provided by forensics through a IT tool developped by the NSI.

The use of the automatic codification IRIS reduces the burden although the more difficult cases must be codified manually. IT tools are used for the detection of errors.

No data revision policy.

No data revision practice.

Not applicable.

The raw statistical data used for compiling statistical aggregates comes from the death certificates provided by physicians, excepting the cases with external causes of death which are informed by a judge through the web death bulletin with judicial intervention. From 2020 the deaths with judicial intervention are informed directly by forensics using a NSI sofware.

NSI receive the death certificates monthly. In the cases of deaths with judicial intervention, we received the information daily, when the court fill in the web death bulletin or the forensic records the death in the software to send the information to the NSI.

Data are recovered from paper death certificates using Optical Recognition to get electronically the information.

Death with judicial intervention are reported using a NSI website.

IRIS is used to coding.

Table on certification (Percentage)

a) Stillbirths

Yes, in the case of stillbirths there are two possible certificates: the one used for births where it can also be informed the foetal deaths and the general death certificate.

b) Neonates

No, We don´t have a different death certificate for neonates cases.

Data validation is explained in the following points.

Description of coding procedure (central level, distributed among other bodies, etc.):

Regional Mortality Registers (MR) are in charge of coding procedure. The coordination of this codification is carried out by the National Statistical Institute (NSI).

NSI provide to MRs the tools to garantee the comparability (updated IRIS dictionary, validation rules...) and the sofware to garantee the secure data transmitions.

NSI is in charge of revising codes (incosistencies, validation rules...)

Description of the procedures to detect errors (i.e. errors such as potential inconsistency in the death certificate or error due to mistake when filling the deaths certificates):

The sofware desinged by NSI for data transmision after coding from the RMs includes validation rules and none of the files containing errors can be sent. The errors and warnings are listed when they occur. Apart from that, NSI use a different software to check again warnings and errors.

Description of the measures taken in order to solve detected errors:

Revision of the certificates containg errors.

Coding performed by a certifier:

Certifiers don´t write the code in Spain.

Estimation of the percentage of autopsy from which information is available for coding:

Not available

Description of double coding exercises and rate of codification errors for underlying cause of death:

Double coding exercises is not made. We use automatic coding.

ICD codes for the underlying cause (% of the Total)

Unknown country/region (%) for residents and non-residents who died in the country

Local Register is used for validation coverage

Not applicable.

Not applicable.

Not applicable

Not applicable.

None