How good is the identification of Indigenous people in hospital separation data?
As with other major health collections, the identification of Indigenous people in hospital records is incomplete [1, 2] . Statistics on hospitalisation provide some insights into ill-health in the population. They are, however, quite a poor reflection of the extent and patterns of treatable illness in the community, since they represent only illness that is serious enough to require hospitalisation. Even then, the comparability of hospitalisation statistics is limited by many factors, including State/Territory and regional variations in admission policies, and differential geographic accessibility of hospitals. More importantly, the incompleteness of Indigenous identification in most jurisdictions means that most comparisons of Indigenous and non-Indigenous hospitalisation under-estimate the true difference. As a result of this, a degree of caution must be exercised in the interpretation of hospital statistics, including those summarised here.
