Summary
This paper explores the most appropriate method of identifying those admitted patient separations in national hospital data for which palliation was a substantial component of the care provided. Three different approaches to identifying such separations are considered: using ‘Care type’ information only (that is, the customary approach), using diagnosis information only, and using both ‘Care type’ and diagnosis information.
What do the coding and collection rules tell us?
An examination of the data collection rules for the ‘Care type’ data item indicates that, by definition, the principal clinical intent or treatment goal of separations with a ‘Care type’ of Palliative care is palliation. Likewise, by definition, the principal clinical intent or treatment goal for those separations with a diagnosis code of Palliative care is palliation, although this intent may have been applicable for part, not necessarily all, of the separation.
What do the results from analyses of data tell us?
Analyses of admitted patient data for 1999–00 to 2008–09 show that which of the three approaches is chosen to identify palliative care separations has a substantial impact on the number of separations identified. For example, in 2008–09, compared with using only ‘Care type’ information, 73% more palliative care separations would be identified across Australia if only diagnosis information was used; correspondingly, 77% more separations would be identified using both ‘Care type’ and diagnosis information.
The 2008–09 data also show state and territory differences in the approach used to assign Palliative care as a ‘Care type’ and/or a diagnosis code. In Queensland, Western Australia and the Australian Capital Territory, there was virtually a one-to-one correspondence between the recording of Palliative care as the ‘Care type’ and as a diagnosis. This was not the case for the remaining five jurisdictions. In these five jurisdictions, the number of separations having a ‘Care type’ and/or a diagnosis code of Palliative care was at least double the number with only a ‘Care type’ of Palliative care.
Do the types of separations identified differ?
An examination of patient characteristics (such as diagnosis) and the nature of the separations (such as length of the separation) showed, in a few instances, some variability in the coding pattern of palliative care separations. However, there were no clear, systematic differences such as those found when state and territory data were considered.
What approach is recommended?
In order to identify those admitted patient separations for which palliation was the principal clinical intent or treatment goal for either part or all of the separation, the use of both ‘Care type’ and diagnosis information is recommended. This approach was endorsed by the national Palliative Care Working Group at its meeting in March 2011. It will be used in further work by the AIHW when the aim is to capture those admitted patient separations for which palliative care was a substantial component of the care provided.