Hospital care for Aboriginals and Torres Strait Islanders: appropriateness and decision making

We are well aware of the excess mortality of Australian Aboriginals and Torres Strait Islanders, their higher hospital admission rates and their longer duration of hospital stay.1,2 However, despite this, relative to their need, Aboriginals and Torres Strait Islanders underutilise specialist healthcare, both as inpatients3 and outside hospitals.4 This situation is exacerbated by demonstrable underfunding of primary care services for Indigenous Australians.4

In this issue of the Journal, Cunningham (page 58)5 reports an analysis of Australian hospital separation data which documents significantly fewer diagnostic and therapeutic procedures performed on patients identified as Indigenous. Cunningham took into account variables possibly affecting use of procedures, including diagnosis, age, sex and place of residence, and acknowledged the problem of incomplete identification of hospital patients as Indigenous. The adjusted data still show fewer recorded procedures in Indigenous compared with non-Indigenous inpatients. These differences are significant within certain disease and diagnostic groups, and of a magnitude which cannot be ignored. Cunningham's study is also consistent with a large body of research from other countries showing disparities according to ethnic group and gender in the use of procedures.6,7

Are the reasons for these disparities in use of procedures related to disease characteristics (including severity), the patient (including preferences and comorbidities), the clinician or the institutional setting? Are they appropriate? And what exactly do these findings indicate?

Ideally, decisions regarding medical care are based on evidence, or at least consensus opinion as reflected in a range of standard practices or options. Patient goals and values are then factored in, together with provider and institutional preferences and consideration of available resources. The end-result should be an individualised decision for each patient. Cunningham's findings suggest that, somewhere along this chain within hospitals, different decisions about use of procedures are being made that correlate with, but may not be caused by, ethnic origin.

It seems implausible that such significant and Australia-wide differences could implicate large numbers of individual clinicians and result from purely personal biases based on race. The disparities are more likely a result of subtler systemic practices, not ill-intentioned but still discriminatory, and almost invisible within an individual patient–provider encounter. The challenge for clinicians is to further dissect the available information and identify the true cause from the many possible contributory factors. For example, it is likely that in some disease conditions where aetiology is well identified in a population and where the disease is endemic that fewer investigations (and therefore procedures) are needed. Renal biopsy may be performed less frequently to investigate renal disease in Indigenous patients, in whom identified antecedent chronic diseases are endemic. Iron-deficiency anaemia may be treated with anthelmintic drugs rather than first confirming the cause. Alternatively, there may be situations in which a population is genuinely underserviced relative to need (eg, some forms of elective or semi-elective surgery).

Possible patient factors should also be identified and dealt with. These include stage of presentation, comorbidities, consent and anticipated postprocedural compliance. All may be reasons why a procedure is not performed, but all these factors can be ultimately overcome, in particular with greater recognition of the importance of primary care. The Royal Australasian College of Physicians has emphasised that adequate primary care is a prerequisite for effective specialist care.8

The uptake of services is more than "patient related" and cannot be separated from the provision of services; societal and institutional factors also structure the doctor–patient encounter within which decisions are made.9 Compliance is likely to improve when a patient's understanding and ownership of his or her disease is matched by a commitment (in the broadest sense) to providing high quality communication and a range of treatment options. The national underdevelopment of key services — Aboriginal Health Workers, Liaison Officers, and particularly interpreter services — shows a lack of appreciation by health institutions of the importance of involvement of Indigenous patients in decision making.

Finally, could clinician factors be a major issue in this disparity in the use of procedures? Do we try hard enough and devote sufficient resources to the early diagnosis of serious conditions, to ensuring stabilisation of comorbidities, to obtaining meaningful consent, and to ensuring good postprocedural care and compliance? Do we have preconceived and incorrect perceptions of cultural appropriateness? Are the expectations we factor into our decision making, often derived from our knowledge of group outcomes, appropriate for this individual patient? It is certainly interesting that for trauma and infectious diseases, in which the decision-making process may be less subjective, Cunningham found no difference overall in the likelihood of having a procedure. However, even within these groups, large differences could be seen at an individual "principal diagnosis" level.

To improve and structure our medical decision making, we need to continue to develop standardised protocols based on best evidence. Subsequent careful recording of consent processes, together with documentation of the decisions reached and the reasons behind them, will shed further light on the issues raised by Cunningham's data. Cunningham's article shows why clinicians need to contribute to the interpretation of population health information. As presented, these data do not identify the specific clinical procedures possibly denied to Aboriginal and Torres Strait Islander inpatients, and, although the findings have flagged the disparity in procedures, more information is needed for clinicians to take these concerns to the next level of analysis.

Specifically, these findings should prompt us to review the decision-making processes determining use of diagnostic and therapeutic procedures in Aboriginal and Torres Strait Islander inpatients. Specialist colleges, societies, hospital units and individual clinicians now have a responsibility to review their own data and establish whether the trend in differential use of procedures applies to their area and, if so, what is driving this difference. The next chapter in this story needs to tease out the connections between healthcare need, use of procedures and health outcomes. If healthcare services are to foster equity rather than further institutionalise inequity, inappropriate reasons for different use of procedures need to be identified and the problems rectified.