About the statistics
Name and topic
Name: Causes of death (discontinued)
Division for Health Statistics
Definitions of the main concepts and variables
The data are coded according to the International Classification of Diseases. Since 1996, causes of death have been classified and coded according to the detailed list in the International Statistical Classification of Diseases and Related Health Problems, ICD-10, Volume 1 (World Health Organization, Geneva 1992). Regulations and guidelines for mortality coding and selection of underlying cause of death are found using ICD-10 Volume 2 (World Health Organization, Geneva 1993). The rules and principles for mortality coding in ICD-9 are found as appendices in NOS C 490 Causes of Death 1995. For more information about the coding WHO Family of International Classifications .
The disease or external cause of injury that initiated the chain of fatal events leading directly to death shall be registered as the underlying cause of death. In addition, up to six causes (conditions that directly led to the death and contributing causes) can be coded from the year 1996. The underlying cause of death is used in the international cause of death statistics.
Underlying cause of death is defined as: 1. The disease or injury that initiated the chain of fatal events leading directly to death or 2. The external circumstances of the accident or violence that was the cause of the fatal injury
The condition listed on the end line of part I of the death certificate is usually the underlying cause of death. The ICD 10 rules and guidelines can, however, result in a different condition (disease or external cause of injury) being selected as the underlying cause of death. To distinguish between these two possibilities, ICD-10 has introduced the concept of originating antecedent cause for the condition listed on the completed line at the bottom of part I. The underlying cause of death is used in the international causes of death statistics.
Immediate cause of death is defined as the disease, injury or condition directly leading to death and which was caused by the underlying cause of death.
Complication is defined as the condition that is a direct consequence of the underlying cause of death (immediate and intermediate cause of death) and by contributing cause of death, reference is to the condition that may have contributed to the death but is not in a direct causal relation with the disease or condition that caused the death.
Age is the number of years at the time of death.
Standard classifications are the Classification of diseases, injuries and causes of death, ICD-9, Norwegian edition, for the year 1995, and the Classification of diseases, injuries and causes of death, ICD-10, English edition, for the years 1996- 2005.
The cause of death statistics differ from ICD-10 (English edition) in the following cases: Fourth category for R99 (other poorly-defined and unspecified causes of death); R99.0 Cause of death cannot be established (physician or other expert has not been able to establish the cause of death) R99.8 Cause of death not specified (the form/document lacks information on the cause of death, e.g. document from abroad) R99.9 No information (the death is recorded in the National Population Register and not reported to the Cause of Death Registry).
Cases of hip fracture (fractura femoris) where the external circumstance is unknown or unspecified are coded with W19 (unspecified fall). From 2009 will such cases, according to ICD-10, be coded X59 (unspecified accident) as the underlying cause of death (ICD-10, Volume 3, 1994, page 600).
The publication NOS C 490 Causes of Death 1995 contains an overview of cases where the Norwegian causes of death statistics deviate from the English edition of ICD-9, as well as from guidelines for the selection of underlying causes. The Norwegian edition of ICD-9 was used in 1986-1995, while the Norwegian edition of ICD-8 was used in 1969-1985.
Statistics are provided at national and county level.
Frequency and timeliness
Annual final figures
Causes of death are reported annually to the Nordic Medical Statistical Committee (NOMESCO), Eurostat, OECD and the World Health Organization (WHO).
The data are stored on individual level by the personal identification number. The register is encrypted.
Background and purpose
The causes of death regulation §1-3 states that the purpose of the Cause of Death Registry is:
1. To monitor causes of death and highlight changes in causes of death over time. 2. To provide a basis for compiling national, regional and local causes of death statistics. 3. To promote and provide a basis for scientific research, and 4. To provide a basis for information and knowledge for planning, and for ensuring and developing the quality of the health service and the health administration.
As a member of the World Health Organization, Norway has, since 1 January 1951, been obliged to prepare official statistics on causes of death in accordance with the International Classification of Diseases (ICD) and the principles and guidelines that apply to coding causes of death. Furthermore, information on the causes of death from the death certificate is to be provided within the framework set by the World Health Organization as shown in appendix A.
In 1951, Norway compiled causes of death statistics according to ICD-5 for the first time. The ICD has been revised several times: 1951-1968 (ICD-6 and 7) 1969-1985 (ICD-8) 1986-1995 (ICD-9) 1996 (ICD-10)
Since 1996, the cause of death statistics have been classified and coded according to the World Health Organizations International Statistical Classification of Diseases and Related Health Problems, 10th revision. The 10th revision of the ICD is based on Jacques Bertillons 1893 international classification of causes of death. The traditional structure in the ICD has been retained in the 10th revision, however the former numerical code system has been replaced with an alphanumeric system. This expands the framework by more than double compared to ICD-9. Not all available codes in ICD-10 are currently in use, and it is possible to include later revisions without changing the alphanumeric system.
ICD-10 consists of 21 chapters. Four chapters (I, II, XIX and XX) extend over more than one character. Chapter I-XVII covers diseases, conditions and symptoms. Chapter XX covers external causes of death. In the mortality statistics, Chapter XIX (S00-T98) is not to be used as the underlying cause of death.
In 2005, Norway started using the automatic system for coding which is named ACME. ACME is an automatic system for coding that selects the underlying cause of death. The coding is based on international rules and guidelines. The object of implementing ACME is to achieve a more uniform coding practice among countries, which will lead to more comparable statistics internationally. Norway has implemented a semi-automatic version of ACME.
Users and applications
The statistics are formed to provide health authorities with an overview of health conditions in Norway. Furthermore, data on causes of death are released for scientific/research and analytic purposes. The release of identifiable individual data to scientists is subject to the relevant permissions from the Directorate for Health and Social Affairs and the Data Inspectorate.
Users of the causes of death data include the OECD, Eurostat, the World Health Organization (WHO), the Nordic Medical Statistical Committee (NOMESCO), the Medical Birth Registry of Norway, university hospitals, medical researchers connected to central and local hospitals, health authorities, the Cancer Registry of Norway and different public institutions. Other users include students, journalists and the general public.
Coherence with other statistics
The number of deaths in causes of death statistics does not completely coincide with the number of deaths in Statistics Norway’s population statistics. The difference used to be more extensive due to different routines regarding delayed reporting.
The differences in the number of deaths in recent years are due to the fact that population statistics are not including civilian death certificates received after the end of January the following year, while causes of deaths statistics includes medical death certificates received until just before publishing the results, for example for the 2010 statistics until the end of September 2011. On the other hand population statistics are including deaths that occurred in previous years, with a change in birth year and death year for the deceased so that age at death will be correct.
The different routines have, for the last twenty years, resulted in a yearly difference in the number of deaths by plus/minus 100 with a few exceptions. Please note that even for a statistical year with an equal number of deaths the two statistics can include a different group of (population of) deaths, for instance regarding age, gender and so on for approximately 200 persons. Nevertheless, over a time span of several statistical years the populations with be approximately identical.
Road traffic accidents involving personal injury
Deaths caused by road traffic accidents in causes of death statistics differ on some points from road traffic accident statistics based on police reports, see http://www.ssb.no/english/subjects/10/12/20/
•The accidents statistics that has the Police as data source, includes only accidents reported to the Police
•The Police reports includes all accidents in Norway, in public and private roads, streets or areas open for the public, while causes of death statistics includes all traffic accident deaths by persons registered as residents in the National Population Register at the time of death, regardless whether the accident occurred in Norway or abroad.
•A minimum of one vehicle must be involved in the accident to be included in the Police records. Civilian and military motor vehicles, railroad and other vehicles on tracks and non engine driven vehicles, are regarded as a vehicle.
•The Police records are including both deceased in the accident and deaths that occurred up to 30 days after the accident. Causes of death statistics include deaths occurring after the accident regardless of the time span after the accident. In the statistics based on the Police records, suicides and accidents due to indisposition are excluded.
Accidents at work
Information on fatal accidents at work are registered in many different registries in Norway. In the Causes of Death Registry these accidents are registered as accidents at work when the physician who fills out the death certificate marks “Yes” to whether this is an accident at work or not. A fatal accident at work is defined as an accident that leads to death during work activity that generates an income. Suicide and murder are not included. There is an extensive underreporting of fatal accidents at work to the Causes of Death Registry.
The Causes of Death Registry includes all dead persons who were registered in the Central Population Register at the time of death regardless whether the death occurred in Norway or abroad. Different supervisory authorities are making records of fatal accidents at work on land (The Norwegian Labour Inspection Authority), aviation (Civil Aviation Authority - Norway), shipping and fishing (Norwegian Maritime Directorate) and petroleum activity on Norwegian shelf (Petroleum Safety Authority - Norway). The registration of accidents at the supervisory authorities includes both residents and non- residents and is mainly based on reports from employers, on the Police’s accident registration and on information about accidents at work found in media.
The Causes of death registry is a central health register sanctioned by Act no. 24 of 18 May 2001 on personal health data filing systems and the processing of personal health data (Personal Health Data Filing System Act), Section 8, cf. the Regulations on the collection and processing of personal health data in the Causes of death registry of 21 December 2001 (Causes of death regulations).
Regulation (EC) No 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work and the Commission Regulation (EU) No 328/2011 implementing Regulation (EC) No 1338/2008 of the European Parliament and of the Council on Community statistics on public health and health and safety at work, as regards statistics on causes of death.
The statistics cover all persons registered by the National Population Register as inhabitants of Norway at the time of their death, regardless of whether the death took place in Norway or abroad.
Data sources and sampling
Statistics on causes of death are prepared on the basis of medical death certificates sent to Statistics Norway by public health officers. Additional information is routinely obtained from the Cancer Registry of Norway, the Medical Birth Registry of Norway, Statistics Norway's statistics on road traffic accidents and the results of autopsies from hospital and forensic laboratories.
In 1995-2002, the doctor reports were used to code approximately one third of all deaths. From 2003 there was a change in the use of data sources from the Cancer Registry and information from additional sources was reduced to just below 20 per cent. In the remaining cases, the only information used is the information given on the death certificate issued by a medical practitioner or a report of death abroad. Data sources of information:
Death certificate: civil information, a cause of death diagnosis form is completed using WHO guidelines, giving the name of the attending physician where possible and is signed by the issuing physician.
The Cancer Registry of Norway: contains information on the certainty of the diagnosis, whether the tumour is benign or malignant, the metastasis status and the time of diagnosis, registration and any changes if applicable.
The Medical Birth Registry: contains data on the health of the mother before and after birth, procedures during the birth and complications in connection with the birth, and the condition of the baby right after the birth.
Report on road traffic accidents: data from police reports on whether the deceased was the driver or passenger and what kind of vehicle was involved in the accident as well as type of accident. From 2005 this source is no longer used.
Autopsies: information about which hospital conducted the autopsy, autopsy number, laboratory number, cause of death diagnosis, and often also a conclusion/assessment.
Forensic autopsies: information on external circumstances in the event of accidents, murder, suicide and other causes of death.
Additional information: forms with additional questions are sent out when the death certificates are incomplete, incorrectly filled in or information is missing. The cause of death statistics unit cooperates with medical consultants, and medical questions are discussed.
The statistics include all persons registered by the National Population Register as living in Norway at the time of their death, regardless of whether the death took place in Norway or abroad.
Collection of data, editing and estimations
Whenever a death occurs, a medical death certificate must be issued by a medical practitioner or local police officer to the public health officer (appendix A). The death certificate is sent to the Probate Court, which issues new death certificates that apply to the administrative and practical aspects of a death. The Probate Court provides the report of death to the local population registry. The records in the population register are forwarded to the National Population Register at the Directorate of Taxes. The Probate Court forwards the death certificate to the chief municipal medical officer at the place of death. The municipal medical officer forwards the death certificate to Statistics Norway. Statistics Norway processes the data on behalf of the Norwegian Institute of Public Health
Deaths registered from medical death certificates are compared with deaths registered in the National Population Register. This process ensures that the statistics are exhaustive. Statistics Norway reminds the municipal medical officers to obtain missing information or copies of previously issued certificates that have not reached Statistics Norway by sending them letters. This process takes place 3-4 times per year. Deaths that still lack a medical certificate and for which no other information is available, are registered with unknown cause of death.
Data on Norwegians short-term residence abroad are compiled by the Ministry of Foreign Affairs or in some cases the Directorate of Taxes. These deaths are included in the statistics, most often registered with unknown cause of death.
Information from the death certificate is compared with information from other sources like autopsies, the Medical Birth Registry or others.
The IT system has built-in validity and consistency controls.
All deaths are included in the data and aggregated by age groups and sex. The aggregated code list for causes of death compiled by Eurostat is frequently used.
The data are not published on individual level. The data are not published if a person may be indirectly identified through a number of variables.
Comparability over time and space
Comparing causes of death (at a detailed level) classified according to the various ICD revisions creates difficulties, both because of the differences in content and the changes in rules for selecting the underlying cause of death. In addition, the number of categories at the 4th character level has been almost doubled from 6 969 categories in ICD-9 to 12 420 in ICD-10, and there is no simple relation between the codes in the two revisions. The World Health Organization has published a translation between the 9th and 10th revision (Translator, Ninth and Tenth Revision, WHO, Geneva 1997), and this shows that only 2 200 codes are unique in both versions.
Eurostat has prepared a shortlist of ICD-10 for use in international comparisons. The list consists of 65 groups of causes of death that are also translated into ICD-9 and ICD-8.
Eurostat has requested that the European countries start working with an automatic system for coding. An automatic program for coding (ACME) selects the underlying cause of death according to the International Classification of Diseases. ACME is based on international rules and guidelines for mortality coding and selection of underlying cause of death (ICD-10 Volume 2). The object of implementing ACME is to achieve a more uniform coding among the coders and to produce comparable statistics internationally.
All Nordic countries use or have started to implement ACME. ACME is a computer application that establishes the underlying cause of death. ACME is developed and owned by the National Center for Health Statistics in the USA, and is used as a standard to increase the comparability of causes of death statistics between European countries. Fifteen countries already use ACME and several will start using it within the next ten years. Norway implemented a semi-automatic version of ACME in 2005.
The Cause of Death Registry in Statistics Norway carried out a pilot project in 2004 (Notater 2004/1). The objective of the project was to evaluate the current code practice in Norway compared to an automated system (ACME) for selection of underlying cause of death.
The Norwegian Cause of Death Registry collaborates with The Baltic and Nordic countries via the Nordic Centre for Classification of Diseases in Uppsala, and internationally with the WHO with regard to comparability and uniform coding. Every quarter, cases are coded in the various countries and the results compared, with ACME as the international standard. Currently the agreement rate (the percentage of corresponding coding) for Norway is approximately 70.
From 1996-2002, deaths caused by acute poisoning among known substance abusers were coded with ICD-10 F10-19 with the fourth character
0. Deaths where the deceased was not known to be a substance abuser were coded with ICD-10 X40-49 (accidental poisoning). Following recommendations from the WHO all acute deaths caused by substance abuse/poisoning were transferred to the accident chapter from 2003. This change has resulted in significant changes to the tables.
The WHO has made a list of codes that can cause pneumonia and are preferable in accordance with WHO rules and guidelines when pneumonia is the reported underlying cause of death. Senile dementia is among the codes preferable to pneumonia according to the new rules. This change has resulted in significant changes in the tables for 2003, 2004 and 2005.
Sources of error and uncertainty
The main source of uncertainty in the cause of death statistics, and thus potentially the largest source of error, is the cause of death examination, i.e. the diagnosis and the physician's reporting of the information on the death certificate. Additional information and contact with the physician are consequently of major importance for ensuring the quality of the information on cause of death.
Deaths registered on the basis of death certificates are compared with deaths registered in the National Population Register. This comparison ensures that the statistics are exhaustive. Reminders for missing death certificates are sent 3-4 times a year. Where no death certificate exists at the time of release of the statistics, these are classified as unknown cause of death.
Other possible errors in the data could be different interpretations of the death certificates and data entry errors.
The death certificate as primary data source may be unreliable because the diagnosis on the certificate simply are not correct or the reported sequence is incorrect.