The IHOPE project aims to disentangle the contributions of the individual, the neighbourhood and the hospital of admission on risk of hospitalisation, and risk of poor outcomes after hospital admissions, for Aboriginal people in New South Wales compared with non-Aboriginal people. The study aims to understand where there are overall disparities between Aboriginal and non-Aboriginal people, and also whether there are disparities at the hospital or neighbourhood level that could be targeted with specific interventions.
The IHOPE reference group includes representatives from Aboriginal Medical Services, the Aboriginal Health and Medical Research Council, and the NSW Ministry of Health. The reference group advises on: areas of research interest, interpretation of results processes for community engagement, disseminating results, and how the results can be best translated into policy and practice.
The study has looked at admissions for and outcomes after ischaemic heart disease, surgery for otitis media among children, transport injury rates across NSW, cataract procedure rates and potentially preventable hospitalisation rates. The outcomes we can look at include event rates (if captured by hospital and deaths data), mortality or admission/procedure rates, length of stay in hospital, readmission to hospital and survival after admission.
The IHOPE project uses linked population datasets (NSW Admitted Patient Data Collection and NSW mortality data), with the data linkage performed by the Centre for Health Record Linkage (CHeReL). The linked data for the APDC is from July 2000 to December 2008. The linked fact of death records are from July 2000 to December 2009 and cause of death records are from July 2000 to December 2007.
The main analysis technique is multilevel modelling. Multilevel modelling takes into account that individuals form groups that are related because they have, for example, received treatment at the same hospital or lived in the same area. People within these groups or clusters are more similar than those in other groups as they have similar exposures which are often unmeasured. Multilevel modelling can account for these unmeasured similarities within the groups, as well as measure the impact of the group level, e.g. individual, hospital and neighbourhood, on the health outcomes.
IHOPE findings will make an important contribution to knowledge regarding the factors and processes underlying poorer health outcomes for Aboriginal Australians. Through the reference group, the findings will inform current policy debates and guide appropriate intervention strategies at the hospital and area level to improve health outcomes for Aboriginal people.
For more information
Please contact Deborah Randall at the School of Medicine, University of Western Sydney, on (02) 4620 3829 or email@example.com.
Figure 1: Diagram of the possible levels and factors in the multilevel analysis of mortality, admission rates and hospital outcomes.
- Randall DA, Jorm LR, Lujic S, O'Loughlin AJ, Churches TR, Haines MM, Eades SJ, Leyland AH. Mortality after admission for acute myocardial infarction in Aboriginal and non-Aboriginal people in New South Wales, Australia: a multilevel data linkage study. BMC Public Health. 2012;12:281.
- Randall D, Jorm L, Leyland A, Lujic S, Churches T, Haines M, Eades S, O'Loughlin A. P2-252 Short- and long-term mortality of Aboriginal people after hospital admission for ischaemic heart disease: a data linkage study in New South Wales, Australia. J Epidemiol Community Health. 2011;65:A291.
Publications under review
- Randall DA, Jorm L, Lujic S, O’Loughlin A, Eades S, Leyland AH. Disparities in revascularization rates after acute myocardial infarction between Aboriginal and non-Aboriginal people in Australia. Under review.
- Falster MO, Randall DA, Lujic S, Ivers R, Leyland AH, Jorm LR. Disentangling the impacts of geography and Aboriginality on serious road transport injuries in New South Wales. Under review.
- Falster K, Randall D, Banks E, Eades S, McNamara B, Gunasekera H, Bambrick H, Reath J, Einarsdóttir K, Jorm L. Disparities in myringotomy and ventilation tube insertion procedures between Aboriginal and non-Aboriginal children in New South Wales. Population Health Congress, Adelaide 2012.
- Reinten T, Randall D, Lujic S, Leyland A, Jorm L. Disparities in cataract surgery between Aboriginal and non-Aboriginal people in NSW. Population Health Congress, Adelaide 2012.
- Randall D, Jorm L, Lujic S, Leyland A. Targeting interventions to improve disparities in incidence of acute myocardial infarction among Aboriginal Australians using multilevel modelling of linked data. International Data Linkage Conference, Perth 2012.
- Reinten T, Randall D, Lujic S, Jorm L. Design matters: exploring disparities in the incidence of cataract procedures between Aboriginal and non-Aboriginal people. International Data Linkage Conference, Perth 2012.
- Falster M, Lujic S, Randall D, Ivers R, Jorm L. Disentangling the impacts of remoteness and Aboriginality on serious road transport injuries in NSW – a data linkage study. Australasian Epidemiological Association Annual Conference, Perth 2011.
- Randall D, Jorm L, Lujic S, O’Loughlin A, Leyland A. Rates of coronary procedures for Aboriginal and non-Aboriginal patients admitted with acute myocardial infarction. Health Services and Policy Research Conference, Adelaide 2011.
- Randall D, Jorm L, Leyland A, Lujic S, Churches T, Haines M, Eades S, O’Loughlin A. Variation in outcomes for Aboriginal and non-Aboriginal people after admission for acute myocardial infarction. 3rd CRIAH Aboriginal Health Research Conference, Sydney 2011.
- Randall D, Jorm L, Lujic S, Banks E, Leyland A, Churches T, Eades S. Algorithms to improve Aboriginal identification in the NSW Admitted Patient Data Collection. 3rd CRIAH Aboriginal Health Research Conference, Sydney 2011.
- Falster K, Randall D, Banks E, Eades S, McNamara B, Gunasekera H, Bambrick H, Reath J, Einarsdóttir K, Jorm L. Management of otitis media: rates of surgical intervention in Aboriginal and non-Aboriginal children in NSW. 3rd CRIAH Aboriginal Health Research Conference, Sydney 2011
- Randall D, Jorm L, Leyland A, Lujic S, Churches T, Haines M, Eades S, O’Loughlin A. Variation in outcomes for Aboriginal and non-Aboriginal people after admission for acute myocardial infarction. Australasian Epidemiological Association Annual Conference, Sydney 2010.
Summary of the latest IHOPE Results:
|Where are the gaps?||Description||What are the implications?|
|Cataract surgery rates|
|Rates of cataract procedures||Aboriginal people are less likely than non-Aboriginal people in NSW to get cataract surgery, particularly in major cities.||There is evidence that Aboriginal people have higher rates of cataract than non-Aboriginal people, so increasing numbers of cataract surgeries for Aboriginal people is crucial. To increase the numbers of cataract surgeries provided, issues of availability, accessibility, cost, and cultural competency in each region, particularly in major cities, need to be improved.|
|Surgery rates for otitis media|
|Surgery rates||Aboriginal children have lower rates of myringotomy with ventilation tube insertion (MVTI), a treatment for recurrent acute otitis media and ‘glue ear’.||Despite some evidence that Aboriginal children have higher rates of otitis media, there are lower rates of this particularly treatment. However, this disparity is not apparent when looking within remoteness or socioeconomic strata; that is, Aboriginal and non-Aboriginal children in remote and more disadvantaged areas both have lower rates than those in major cities and more disadvantaged areas.|
|Rates by geographic area||Those in major cities and inner regional areas are more likely to get the procedures than those in remote and very remote areas.||As it is likely that otitis media is more prevalent in rural and more disadvantaged areas, the rates of MVTI do not appear to be related to need. They may be related to access and private health insurance. There may be over-servicing of those in major cities and less disadvantaged areas.|
|Rates by socioeconomic status||Those living in the least disadvantaged areas are most likely to get procedures.|
|Potentially preventable hospitalisations|
|Admissions for potentially preventable hospitalisations (PPH)||Aboriginal people have higher rates of admissions for PPHs than non-Aboriginal people, and the higher admission rates are, in turn, driving higher bed day rates and higher costs per person.||It is possible that the higher admission rates for Aboriginal people should be even higher given the higher prevalence of many of the conditions and access issues. More work is needed to determine if these admissions are, in fact, preventable.|
|Acute myocardial infarction|
|Age at first heart attack||Aboriginal people in NSW are being admitted with their first AMI at an earlier age than non-Aboriginal people.||Points to the importance of prevention and management of those with early heart disease symptoms.|
|Procedures in hospital||Aboriginal people admitted with AMI are less likely to receive a revascularisation procedure (CABG or angioplasty).||Related to the hospital of admission (Aboriginal people are less likely to be admitted to higher volume hospitals with specialist facilities), and also to the higher rate of conditions such as diabetes and renal failure that can interfere with the revascularization procedure.|
|Longer term survival||Aboriginal people are more likely to die within one year of their first AMI admission, after adjusting for age, sex and hospital of admission.||Points to the importance of follow-up care, and co-ordination of care between hospitals and GPs/AMSs. There might also be issues with medication compliance and general poorer health and multiple conditions. We plan to look into whether the lower procedure rates may also be involved.|
|Road traffic injuries|
|Risk of small vehicle injuries||Aboriginal people have a higher risk of small vehicle injuries across NSW, but this is because of where people live. An Aboriginal person does not have a higher risk of injury compared with a non-Aboriginal person living in the same area.||Overall risk (for both Aboriginal and non-Aboriginal people) is highest in regional areas, and safety campaigns and urban interventions are needed to address this.|
|Risk of bicycle and pedestrian injuries||Aboriginal people have a higher risk of bicycle and pedestrian injuries compared with others living in the same area.||Safety campaigns are needed in the higher risk areas. We are looking into ways of communicating these findings to local Aboriginal communities and councils in higher risk areas.|