3317.0.55.001 - Information Paper: External Causes of Death, Data Quality, 2005  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 11/04/2007  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All

Initiatives to Address External Causes of Death Data Quality Issues

The ABS first highlighted a number of data quality issues in relation to external causes of death in the publication Causes of Death, Australia, 2004 (cat. no. 3303.0) on 14 March 2006. The ABS continues to develop ongoing strategies for identifying and managing concerns and issues regarding quality. The process to improve the quality of causes of death data will be progressive and continuous. Results achieved from the Indigenous Mortality Strategy will also have an impact on the quality of causes of death data.


DATA SOURCES

Registrars of Births Deaths and Marriages

In 2005, the ABS commenced a program to develop better relationships with Registrars of Births, Deaths and Marriages (RBDMs), with the aim of increasing the quality of birth and death statistics. This program has included a range of activities including developing a national standard set of data items for collection of birth and death statistics and providing ABS assistance to RBDMs in areas such as form content and design. The program has had a number of successes including delivery of data to the ABS in standard file formats.

Coroners and National Coronial Information System (NCIS)

The timeliness, quality and comprehensiveness of information on the NCIS database are important factors affecting the quality of ABS external causes of death data.

The ABS has commenced a major collaborative effort with coroners (and the NCIS team) to ensure that completed cases are able to be entered into the NCIS at the coroners' more promptly. This collaboration is expected to continue over the next few years, with the aim of identifying and addressing the issues (e.g. educational, operational, or technical). Initial effort will be concentrated in those jurisdictions with the greatest delays and those most likely to be able to produce improvements. It is likely that the issues will vary according to the jurisdiction, and progress will be faster in some jurisdictions than in others. This initiative will allow for a close working relationship with coroners which will also facilitate ABS monitoring of completion rates and increase knowledge of contributing factors.

NCIS data entry timeliness, quality and comprehensiveness has been steadily improving over the past few years, however constraints around data entry resources and organisational structures at coronial offices have limited the impact/improvements NCIS has been able to achieve in these areas in the past.

The NCIS has developed a strategic plan (NCIS, 2007a) for the five year period 2007-2012. One of the goals of this strategic plan is to improve the NCIS database integrity by:

  • improving the timeliness of data entry and case closure on the NCIS
  • improving the quality and comprehensiveness of data on the NCIS
  • encouraging and supporting the enhancement and addition of certain reports available on the NCIS.

The specific strategies and initiatives that are particularly relevant to the achievement of these objectives are:
  • the appointment of qualified, dedicated coders to perform NCIS data entry in each coroners office around Australia
  • the facilitation of real time coding of coronial cases as the investigation progresses
  • the implementation of an annual training and accreditation program for NCIS coders to achieve a national coding competency standard
  • inclusion of full text documents that are not currently available on NCIS for some jurisdictions
  • more detailed information on deaths which occurred in non-metropolitan areas
  • support and encouragement for the improvement of consistency and detail within various coronial documents which are made available via the NCIS (e.g. coronial findings relating to unidentified remains; narratives of police reports provided to a coroner)
  • consideration of proposals for adding additional documentation to the NCIS. These include proposals for automated upload of additional fielded information from electronic police forms which report the death to the coroner and the implementation of a drugs module project which will give more easily accessible information about types, levels and contribution of drugs to reported deaths (sourced from toxicology laboratory systems).


DATA PROCESSING

New mortality processing system

For the production of causes of death statistics from the 2006 reference period, the ABS will be using a new mortality processing system. This new system will incorporate the latest version of the MMDS software (footnote 2) and also contains a new suite of quality control features.

Moving to a later version of MMDS will allow the ABS to take advantage of increased functionality, particularly with regard to the SuperMICAR component of the MMDS system (which automatically encodes medical cause of death information). Concurrently with the move to the new version of MMDS, the ABS is undertaking a redevelopment of all the components of mortality processing. Many functions which are currently performed manually will be automated. This will result in coders spending more time coding and less time on administrative computing tasks e.g. moving files/records through the MMDS software.


CODING

Review of ABS external causes coding practices

Over the next couple of years, the ABS will work with experts, seeking their assistance in guiding the development of new business rules for ABS coding of external causes. While this exercise is not expected to result in dramatic changes in the data, the aim is to standardise and improve the rules.

The Suicide Coding Review Working Group has recently been established by the ABS to assist in improving the quality of national causes of death data. The working group is expected to help support and facilitate the following outcomes:
    • quality suicide data which meet the needs of users and ensures quality outputs and outcomes are delivered
    • suicide coding undertaken in accordance with the coding rules of the ICD-10
    • consistent application of suicide coding rules and practices among all stakeholders in Australia
    • best practice suicide coding practices are used by the ABS which are understood by providers and users of suicide data.

The Suicide Coding Review Working Group is undertaking a review of current ABS and other Australian organisations suicide coding practices. The review will include reviewing the following issues:
    • data sources
    • ABS application of ICD-10 suicide coding rules
    • coroner practices and constraints with regard to making a finding of suicide and how this relates to a statistical definition of suicide i.e. should the current dependence on the legal definition of suicide be changed to a statistical/classificatory approach
    • level of 'evidence' required to code a death as suicide
    • revision of suicide coding practices (e.g. possible use of a supplementary code to differentiate between undetermined intent and unknown intent).

The Suicide Coding Review Group may make recommendations for ABS consideration with regard to the issues investigated. Final responsibility and accountability for practices and data quality will rest with the ABS.

Coding of open coronial cases

An investigation will be undertaken in 2007, in conjunction with key external stakeholders including the NCCH, to determine options for the treatment of open coronial cases. Currently the ABS codes an open coronial case based on the information available at the cessation of processing for that particular reference year. This generally leads to the allocation of a non-specific code (which is not subsequently revised). Options will be investigated which would allow for open coronial cases to be allocated a code which specifically indicates that the case has still not been finalised. These cases could then be coded upon finalisation of the coronial investigation and revised data published by the ABS on a year of occurrence basis.


DATA ANALYSIS

Ongoing analysis and monitoring of data quality

The quality of external causes data has been the particular focus of careful validation and analysis in the last two reference years.

The ABS will continue to build on its existing quality assurance measures in order to monitor data quality as comprehensively as possible, and to inform users through explanatory material provided with statistical outputs and through publications such as this. The new ABS processing system will facilitate the measurement of quality through greater flexibility and internal access to the data, and allow further and more comprehensive studies to be undertaken.


RELEASE OF DATA

Increased availability of metadata

Over the next several years, the ABS plans to review the metadata (information about the data) available for causes of death statistics in order to allow users to make more informed use of the data, including ability to compare between different data collections. Examples of the metadata which is planned to be released include:
  • publication of editing rules applied to ICD-10 coding
  • review of the documentation accompanying release of unit record data
  • revision of information in the Directory of Statistical Sources on the ABS website
  • availability of data item level quality statements
  • increased provision of technical and explanatory notes in statistical outputs.

Review of outputs (including release dates)

The Causes of Death collection is a long-standing ABS collection. The main outputs are currently a detailed summary publication (Causes of Death, Australia, cat. no. 3303.0), a supplementary publication about suicides (Suicides, Australia, cat. no. 3309.0) and a unit record file (URF). There is also a large demand for information consultancy services. Data are not revised after release.

A review of causes of death data products will be undertaken by the ABS during 2007-08 to ensure that a suite of products is put together which better meets user needs, takes advantage of the wealth of information available from the collection, is readily accessible, released in a timely manner, easy to produce, and provides additional analytical material. The production of new outputs will be facilitated by the introduction of the new, more efficient, ABS processing system. Options for consideration could include early release of data for natural causes, or some form of revision of external causes data.

A proposal for a revised causes of death product suite is expected to be made available to users for comment in the next six months. To express interest in this review, contact the National Information and Referral Service on 1300 135 070 or email client.services@abs.gov.au.