Death investigation is the responsibility of each individual Canadian province and territory—there is no overarching federal authority. As a result, each province and territory has developed their own system and legislation to fulfill the mandate of investigating deaths that are unexpected, unexplained, or as a result of injuries or drugs. Two different death investigation systems have developed in Canada: the Coroner’s system and the Medical Examiner’s system. The Coroner’s system is used in the majority of provinces and territories. It is a system that is centuries old and originated in Great Britain. It is found throughout the world in countries that were former British colonies, including Canada. The Medical Examiner’s system (used in Alberta, Manitoba, Nova Scotia, and Newfoundland and Labrador) is just over one century old and originated in the United States. Although there are some differences between the two systems, the ultimate goal of each is the same—to investigate certain deaths defined in their legislation and establish the identity of the deceased together with the cause of death and the manner of death.
Almost all Canadian Coroner and Medical Examiner systems have some provision for going beyond an investigation of the death to a public “inquisitional” hearing, referred to as an Inquest or Public Inquiry. One of the primary purposes of this type of hearing is to develop recommendations for the prevention of similar deaths in the future without making any findings of fault.
It is clear from the foregoing that C/MEs are a rich source of information with respect to deaths that are of great public interest, including all injury- and drug-related deaths.
As has been previously noted, C/MEs are required to categorize deaths according to what is called the cause of death and the manner of death, both of which are reflected on the death certificate. The cause of death is defined as the disease or injury that initiates the chain of events ending in death (with no implication of any time limit). The manner of death is also referred to as the means by which death occurs. The five manners are:
Natural: All deaths where a disease initiates the chain of events ending in death.
Accident: All deaths where an injury initiates the chain of events ending in death and there is no element of intent in the circumstances leading to the injury.
Suicide: All deaths where a self-inflicted injury initiates the chain of events ending in death and where the decedent intends to cause their own death.
Homicide: All deaths where an injury initiates the chain of events ending in death and there is evidence to indicate some intent on the part of another individual to cause harm.
Undetermined: All deaths where investigation is unable to attribute one of the previous manners are categorized as undetermined. Note that in such instances, the cause of death may be known.
Some Canadian jurisdictions also use an “Unclassified” manner of death, but there is significant variability in how each jurisdiction defines and utilizes this manner.
An important consideration for both the cause and manner of death is that these are not facts but represent the opinion of the certifier. As with any opinion, there are bound to be differences between individuals certifying deaths that occur under similar circumstances and the strength of any opinion is dependent upon several factors, including the training, experience, biases, and integrity of the certifier.
Prior to the CCMED, the only comprehensive national data on fatal injuries were those contained in Statistics Canada’s Canadian Vital Statistics Death Database (CVS-D). The CVS-D collects demographic and medical (cause of death) information annually from all provincial and territorial vital statistics registries on all deaths in Canada. Provincial and territorial death registration forms include a medical certificate of cause of death section, completed by a physician or a coroner or medical examiner. The cause of death variable in the CVS-D is classified according to the World Health Organization’s “International Statistical Classification of Diseases and Related Health Problems” Tenth revision (ICD-10). There are approximately 230,000 deaths of Canadian residents registered in Canada each year.
It is important to understand that the majority of deaths in Canada are caused by natural diseases that have been diagnosed by a physician; such that when death occurs the decedent’s physician can complete a death certificate that documents the cause of death. These deaths do not require any involvement by a C/ME. The remaining deaths are unexplained natural deaths, where a physician doesn’t know the cause of death, and deaths caused by injuries or drugs. The latter are subdivided into four main categories referred to as manners of death: accidents (or unintentional injuries), suicides, homicides, and undetermined deaths (where there is considerable doubt about what the correct manner of death is). These deaths must be reported to and investigated by a C/ME.
The provincial and territorial C/MEs hold data on all deaths that they investigate in their jurisdictions. Depending on the province or territory, the percentage of deaths investigated by a C/ME can range approximately from 7% to 45% annually. Although the criteria for reporting deaths vary somewhat by jurisdiction, deaths caused by natural diseases account for about 61% of all C/ME cases annually.
Until the development of the CCMED, there was no central collection of C/ME data. When national data was required, particularly as it pertains to injury/drug deaths, researchers needed to visit each of the 13 chief C/MEs offices to consult the relevant records and data. The CCMED project grew out of the recognition of a need for a national source of accessible, standardized information on the circumstances in which fatal injuries occur. Through the aggregation and centralization of C/ME offices to consult the relevant records and data. The CCMED data, it will be easier to identify and link similar deaths across the country and reveal patterns of contributing factors in these deaths. The CCMED will also make it possible to obtain additional detail on deaths due to causes that are not specified to a unique code in the current version of the ICD but are important in Canada; for example, deaths involving the use of snowmobiles as a specific type of all-terrain vehicle, or the specific source of carbon monoxide (such as car exhaust) in intentional and unintentional carbon monoxide deaths.
Each province and territory has a list of data elements which they collect, store and report. A common data set was developed to allow the collection and aggregation of provincial and territorial C/ME data. This system allows provinces and territories to maintain their own data but also support data collection for the national minimum data set. A CCMED record contains the following information:
Certain variables, such as the activity at the time of the event leading to death, usual residence type, circumstances of injury, location of the event leading to death, and safety devices are based on extensive code sets with an expanded classification.
Each province and territory has a distinct system for managing their data, which varies in the degree of automation and the amount of detailed information stored. Provinces and territories either map their electronically stored data to the national data requirements to produce an output file for the CCMED or use the data capture tool developed at Statistics Canada to capture their cases and produce an output file. Prince Edward Island, Nunavut, Northwest Territories, and the Yukon are currently using the data capture tool.
The CCMED went into production on March 1st, 2008 and began collecting data from 2006. To date, not all jurisdictions are able to provide data: for a number of reasons, data from Nova Scotia, Manitoba and Newfoundland and Labrador and Nunavut are not covered in this report.
Data files received from the provinces and territories are sent to the CCMED electronically and go through an extensive verification process to ensure that only valid data are accepted onto the system. Cases with fatal errors are not loaded onto the database; reports are returned to the jurisdiction for validation and correction. The case must be re-submitted to the CCMED and go through the verification process before it can be loaded onto the CCMED. Minimal changes are made centrally to the data submitted by the provinces and territories. Steps are taken to reduce the percentage of cases that fail validation. The Statistics Canada data capture tool contains the same validation rules as the CCMED, and cases must clear validation prior to being available for transmission to the CCMED. For the provinces and territories with their own databases that require mapping to the CCMED, the data mapping program was developed and thoroughly tested in conjunction with Statistics Canada, which allowed for the identification and correction of major sources of error.
As a newly implemented database, the CCMED requires a thorough analysis of its data integrity and quality. This report provided the first opportunity to test the robustness of the database and to identify gaps in some of the variables—specifically, those describing the circumstances and activities surrounding the event leading to death. Corrective measures should be implemented in order to release their analytical potential.
This report covers the universe of “closed” C/ME cases and covers deaths that occurred in 2006 to 2008. When the year of death is unknown, the year in which the death was discovered is used. On the basis of these extraction parameters, a reconciliation exercise using 2006 to 2008 data found a strong agreement between the P/T and the CCMED data.
There are situations where two jurisdictions may investigate the same death. As an example, if an injury occurs in one jurisdiction and the patient is sent to a trauma centre in a different jurisdiction where they die, C/MEs in both jurisdictions might conduct investigations. For the purposes of the CCMED, the death record retained will be the one from the jurisdiction where the death occurred.
Undercoverage is thought to be minimal and may occur due to lengthy delays in the investigation of certain types of deaths. It is expected that a low percentage of cases will fail validation and thus limit the possibility of undercoverage. However, if the jurisdiction does not re-submit cases that fail validation, these would not be part of the final data file.
There will be no instances of complete nonresponse. That is, the CCMED obtains at least some information for each identified case. In the rare instances where little information is known about the decedent, the fields will be left blank or coded to unknown. There are situations where a C/ME is notified of a death, and after an initial investigation it is decided that this is not a C/ME case. These cases are identified and are excluded from the database.
The CCMED is the only centralized source of standardized C/ME data in Canada. C/ME data are rich in detail on the circumstances leading to deaths, such as the location of the incident leading to the death, the activity at the time of the event leading to the death, whether there were any safety devices in use, or if this death was part of an incident leading to multiple deaths. This information will allow for better injury prevention research and analysis.
The CCMED provides additional detail on particular types of deaths that may not be specified in the current version of ICD-10 but are important to Canadians (such as snowmobile deaths or deaths due to carbon monoxide poisoning).
The collection of C/ME data on the national level will aid in revealing patterns of factors and circumstances that contribute to death through detection of similar deaths and “clusters of deaths”. In addition, the CCMED will greatly enhance the rate of information exchange between interested parties and address the information needs of a range of stakeholders, including: coroners and medical examiners, national information agencies, public health policy makers and researchers. In doing so, the CCMED will lead to better data collection around the country and will encourage the standardization and implementation of investigation protocols.
Despite these important advantages, the CCMED data has certain limitations. Different jurisdictions may have their specific definitions for certain data elements. Although an attempt is made to standardize data with the minimum dataset, the interpretation may vary within the different jurisdictions.
Individual C/ME offices collect different levels of detail for particular deaths according to the importance of certain types of deaths in their jurisdiction. The minimum dataset of the CCMED may have only one level of detail for the same types of death, such that there are instances where this will result in a loss of detail in the information that was collected at the C/ME office versus what was mapped and sent to the CCMED.
The CCMED will not be the best source of data for certain types of deaths. Since coroners and medical examiners only deal with a small percentage of deaths caused by natural disease, these will not be well reflected in the database. Information will not be collected on deaths of Canadians occurring outside the country. The CCMED data will not allow for the identification of all cases where alcohol or drugs were a factor in injury deaths. For example it won't necessarily capture the role that alcohol or drugs played in a motor vehicle accident where the driver of the other vehicle that caused a collision was intoxicated but survived. On the other hand, if the drivers of both vehicles died in the same incident, the CCMED will link the two deaths.
The official inception of the Australian National Coroner Information System (NCIS) in July 2000 makes Australia the first country in the world to have developed a national database of coroner information. The success of their system is evidenced by the number of third party users who had online subscription access to the NCIS in 2009-2010. This vast number of registered organizations includes various government and private industries: the Australian Department of Health and Ageing, the Consultative Council on Obstetric and Paediatric Mortality and Morbidity, Safe Work Australia, the National Drug and Alcohol Research Centre, the Australian Institute for Suicide Research and Prevention, the Australian Institute of Criminology, the Monash University Accident Research Centre, the Queensland Fire & Rescue Service, the Australian Bureau of Statistics, and many others. The NCIS has demonstrated the immense potential and relevance of coroner’s data in the area of injury prevention; NCIS data were critical in producing evidence to support the implementation of several significant Australian death and injury prevention initiatives.
The remainder of this report will profile the work of C/MEs in the nine provinces and territories for which the CCMED has received data and examine how caseload varies across jurisdictions. Effort will be made to put some of this information into greater context by using the information regarding the events and circumstances leading to death, and in doing so, highlight the strengths of the CCMED while identifying some of its limitations. The three appendices provide additional analysis and information to better put the observed results into context. Notably, Appendix C examines informally the differences in the circumstances under which the provincial and territorial coroners and medical examiners investigate deaths
Note to readers – Key information about the CCMED.