TECHNICAL NOTE 2 CAUSES OF DEATH - REVISIONS PROCESS


INTRODUCTION

1 An ongoing issue for the Cause of Death collection has been that the quality of the data can be affected by the length of time required for the coronial process to be finalised and the coroner case closed. For some time, these concerns have been raised by key users of Causes of Death (COD) data regarding the quality of external causes data (e.g. deaths due to intentional self-harm (Suicide), homicide, Sudden Infant Death Syndrome (SIDS) and motor vehicle accidents). The ABS have addressed these data quality concerns in two ways.

• First, by increasing the length of time from the end of the reference period to publication of data from 11 to 15 months to allow for a longer time period to receive information on coroner certified deaths; and
• Second by introducing a process of revisions to COD data. This technical note will discuss in detail the revisions process which will be applied to COD data.

2 Up to and including deaths registered in 2006, ABS Causes of death processing was finalised at a point in time. At this point not all deaths registered in the reference year which were referred to the coroner, had been investigated, the case closed and relevant information loaded to the National Coroners Information System (NCIS). The coronial process can take several years if an inquest is being held or complex investigations are being undertaken. In these instances, the cases remain open on the NCIS. Impacts on data quality for coroner certified deaths in terms of specificity of ICD code are:

• general increases in the length of coronial investigations
• increases in the workload of coroners
• the timeliness of input of coronial findings to the NCIS.

3 To address these data quality issues, all coroner certified deaths registered after 1 January 2007 are now subject to a revisions process. The revisions process enables the use of additional information relating to coroner certified deaths either 12 or 24 months after initial processing. This increases the specificity of the assigned ICD-10 codes over time. As 12 or 24 months of time has passed since initial processing, further information may have become available to the ABS about the causes of these deaths.

4 2007 (revised) data which has been subjected to the revisions process has been published in this release. In 2011, after a second application of the revisions process, 2007 (final) data will be released in the 2009 (preliminary) Causes of Death publication. See Technical Note 3: 2007 Revisions for further information on the impact of the first application of the revisions process to 2007 data.


REVISIONS PROCESS

5 All coroner certified deaths registered after 1 January 2007 are now subject to a revisions process after 12 and 24 months. The COD revisions process allows the ABS to code any further information loaded to the NCIS after the preliminary coding took place. In many cases, the coronial cases will be closed, with the coroner having determined the underlying cause of death and allowing the ABS to code a more specific cause of death. If the case remains open on the NCIS, ABS will investigate and use additional information from police reports, toxicology reports, autopsy reports and coroners findings to assign a more specific cause of death to these open cases. It is expected that all open cases will be investigated in this manner from 2008 onwards.

6 ABS will undertake two rounds of revisions only for each year's data. It is anticipated that the quality of the data will be higher after two revision processes. In each round of revisions, those cases which were open at the time of publication the year before will be reassessed by ABS coders to take into account any new information available on the NCIS. This will affect both cases which have been subsequently closed, and those which remain open.

7 As 2007 reference year data is the first series of data to undergo the revisions process, a review of those 2007 cases which remain open on the NCIS at the end of the second round of revisions (January 2011) will be undertaken. The results of this review will be used to determine whether the ABS will undertake further revisions to the 2007 data. It is anticipated that the quality of the 2007 data will be very high after being subjected to two rounds of revisions. If this is the case, the ABS will not undertake any further revisions for that reference year. COD data for each year will therefore be released three times, as preliminary, revised and final. The release schedule is demonstrated in the table below.

Table 1, Causes of Death Release Schedule
Data Reference Year	Anticipated Release Schedule
2007 Preliminary	March 2009
2007 Revised	March 2010
2007 Final	March 2011
2008 Preliminary	March 2010
2008 Revised	March 2011
2008 Final	March 2012

8 For all deaths registered from 1 January 2008, all coroner certified deaths with a case status of open or closed will be reviewed during the processing of the current reference year's data. Table 2 below is a summary of the business rules used by the ABS in the application of the revisions process to coroner certified deaths. During subsequent revisions (2007 onwards), all coroner certified deaths that are open at the end of processing for the previous two reference years will be reviewed.

9 These business rules will be applied for all open and closed coroner cases for all releases from now on.

Table 2, Business Rules for 2008 and beyond for revision of coroner certified deaths
Data Type	Coroners Cases Affected	Process
Preliminary, Revised and Final	All - open and closed	Code all open and closed coroners cases for the reference year based on ICD-10 coding rules, including the use of additional information: police, toxicology and autopsy reports.