Bulletin 14 - Issues of injury classification
![[Previous]](/images/arr_prev.gif)
![[Next]](/images/arr_next.gif)
![[Top]](/images/arr_top.gif)
Issues of injury classification
Injury deaths and hospitalisations are classified according to the
International Classification of Diseases external causes codes, commonly known
as E-codes[3]. These have a limited capacity for describing the injury event
and are useful for broad epidemiological study. The level of detail of causes
varies from broad category to broad category, with the greatest detail
occurring in the on-road transport and poisoning by pharmaceuticals categories.
E-codes are most suitable for assessing injury patterns in western
industrialised countries. Many Aboriginal and Torres Strait Islander
peoples' injuries fall into categories where there is little detail
(e.g. falls). The overall size of a problem can be assessed, but a detailed
understanding of the causes cannot be obtained.
In addition to the difficulties with E-codes, it is apparent that culture
affects the way in which information about an event is presented to
investigators or clinicians and the way this is interpreted through coding.
Injury by violence is known to be severely under-reported in hospitalisation
data for non-Aboriginal women. These women are reluctant to report the true
nature of the cause, and medical staff are often reluctant to record details of
assault in medical records[9]. Violence is recognised by Aboriginal and Torres
Strait Islander peoples as a key concern. In this milieu, violence in domestic
settings, in particular, is less likely to be hidden. It is likely, therefore,
that the reluctance to report violence is less among Aboriginal and Torres
Strait Islander women. The reported rate is therefore likely to be closer to
the experienced rate.
The rate of reported violence resulting in hospitalisation among Aboriginal and
Torres Strait Islander women is much higher than that of non-Aboriginal women,
resulting in a very high rate ratio (Table 4). It is possible that a
substantial component of this ratio is due to systematic differences in
reporting and coding practices. The size of this bias is not known and will be
difficult to determine. This type of reporting bias is less likely to occur for
deaths due to the more detailed investigations undertaken as part of the
coronial process. The overall age-standardised rate ratio for interpersonal
violence related deaths is 10.8 (Table 3). This suggests that, while the
hospitalisation rate ratio may be inflated by under reporting in the
non-Aboriginal community, the level of interpersonal violence experienced by
Aboriginal and Torres Strait Islander men and women is comparatively high and
is worthy of particular attention.
Aboriginal and Torres Strait Islander peoples have raised the question of how
certain classes of events are coded in both deaths and hospitalisations data.
Traditional punishment may result in injury requiring hospitalisation and,
occasionally, in death. External causes codes could identify this as violence
related or as injury during legal intervention. It is uncertain which category
is used, although the low numbers of injuries classed as injury during legal
intervention, suggests that these cases may be coded as violence.
Alternatively, Aboriginal and Torres Strait Islander peoples may be unwilling
to identify traditional punishment to non-Aboriginal people and may provide
information that would result in these events being classed as accidents.
Clearly, it is important to understand such issues better and to develop ways
of obtaining information and coding it to reflect the cause accurately.
3. World Health Organization. International classification of diseases (1975
revision), Geneva: WHO, 1977.
9. Sherrard J, Ozanne-Smith J, Brumen IA, et al. Domestic violence: patterns
and indicators. Melbourne: Monash University Accident Research Centre, 1994;
Report No. 63.
|