Correspondence to: Associate Professor Paul Torzillo, suite 420/RPAH Medical Centre, 100 Carillon Avenue, Newtown, New South Wales 2042. Fax: (02) 9516 3759; e-mail: firstname.lastname@example.org
Introduction: Many of the health problems faced by rural and remote Aboriginal people have been attributed to a poor living environment. In the mid 1980s we began a process of defining problems with the immediate living environment that would affect health. These related particularly to safety, washing and hygiene practice.
Methods: Between January 1999 and November 2006 we undertook a standardised and detailed assessment of housing in Aboriginal communities. This involved an initial assessment of 250 items in each house and living area, focusing on performance and their impact on these healthy living practices. At the first survey-fix we implemented a limited cost repair of non-functioning health hardware and then six months later returned to the communities for a repeat assessment to examine improvement in performance.
Results: Between January 1999 and November 2006 we assessed 4,343 houses in 132 communities in four States and the Northern Territory during survey-fix 1 (SF1) and have repeated that survey-fix assessment (SF2) in 3,448 houses in 112 of those communities. This survey demonstrates extraordinarily poor performance of Aboriginal houses. In the survey period, 71,869 items referred for repair by survey teams were inspected by licensed electricians or plumbers and 49,499 of these have so far been fixed. Only 10% of these house items requiring repair were due to vandalism or misuse.
Conclusion: Improvements in the living environment for Aboriginal people will require a sustained commitment to the planning, funding and implementation of maintenance programs in addition to adherence to the design, construction and supervision detail outlined in the National Indigenous Housing Guide.
For three-quarters of a century political and public health commentators have been linking the poor health of Aboriginal people to their living environment.1–3 However, during this time there have been no substantive attempts to detail the elements of the living environment that are likely to contribute to poor health, what health problems they are likely to cause and how they might be corrected. In 1985 on the Anangu Pitjantjatjaraku Lands, in the north-west corner of South Australia, we began a process of addressing these questions. In 1986, we undertook a detailed study of the living environment of Aboriginal people (Anangu) in this region. In this project we assessed the safety of the houses and determined a range of healthy living practices (HLPs) that would be necessary for anyone living in this environment, regardless of background, if they were to keep themselves and their family healthy.4 These were:
1Washing people, especially children.
2Washing clothes and bedding.
3Removing waste safely.
5Reducing the impact of crowding.
6Reducing the impact of animals, particularly dogs, as vectors of disease.
7Reducing the impact of dust.
8Improving temperature control.
9Reducing minor trauma.
These healthy living practices were prioritised on the basis of existing public health knowledge and their likely impact on health status. We placed life-threatening safety issues highest, followed by washing and waste disposal then cooking and food storage.
We then examined the housing stock and immediate surrounding living environment to determine if there was functioning ‘health hardware’ (a term borrowed from the late Professor Fred Hollows) necessary to carry out these healthy living practices. During this work we defined reasonable targets for each of the HLPs. For example, we developed an objective that mothers should be able to wash their children once a day and wash hands and face frequently all year round. Our assessment focused on testing the functional capacity of households to conduct these HLPs. The report from this project prescribed a range of design and implementation recommendations for the provision of health hardware.4 A major finding of this work was that maintenance programs were crucial to sustaining any health-supporting function of housing. When maintenance programs were absent or failed, housing infrastructure often became a health hazard.
Following this work we further developed a detailed process of testing and fixing the health hardware in community housing. Funding for subsequent community projects was provided predominantly from portfolios within the Commonwealth Government but also the NSW Department of Health. Between 1999 and 2006 we undertook 132 projects and here we describe findings from that work.
Community selection and project offer
Community councils and Indigenous community housing providers (CHPs) were approached about the projects by a range of representative bodies including Territory and State Indigenous housing organisations; regional Aboriginal councils or public health units, and in some cases individual Indigenous community councils made an unsolicited approach to have a project in their community. An initial community visit was then undertaken where the obligations, benefits and limitations of the projects were outlined to community councils and CHPs, who were then asked if they wished to participate. If the community representative organisation agreed and formally requested involvement then preliminary information about local housing management and maintenance practices, housing stock and essential services was obtained in order to facilitate planning for the implementation of the survey-fix process. At this stage the project managers initiated discussion about recruitment of local Aboriginal people for the project teams.
The survey-fix process
Each community project team consisted of an experienced project manager, a small number of trained staff to co-ordinate data and lead survey-fix teams, licensed tradespersons to conduct the urgent health and safety fix/repair works, mainly of electrical and plumbing items, and a majority of local community members. The project managers had a professional background in building, architecture or environmental health. All team members were issued with standard tools and testing equipment and underwent standardised training using a range of training materials developed for this work. The emphasis was on standardised, reproducible testing of all aspects of health hardware in the living environment. Each house/living area had 250 items assessed including lights, power outlets, water outlets and hot water temperature, shower, basin, bath, toilet, laundry, and kitchen areas. This allowed the collection of consistent data across projects and communities as well as accurate assessment of change over time when subsequent surveys were undertaken in the same communities.
All projects involved an initial survey-fix visit (SF1) and a subsequent identical process 6-12 months later (SF2) to assess improvement in health hardware as a consequence of the fix work and document issues to be further addressed. We had determined the minimum functioning health hardware that would be required to carry out each of our healthy living practices and to ensure basic infrastructure safety. For example, in addressing the problem of washing adults we tested and recorded performance of 19 items associated with the shower area of the house. We determined that seven of these were critical in order to be able to use the shower in a sustainable fashion over time (see Table 1). Thus the shower area would have to pass each of these seven key tests for the house to qualify as able to provide ‘the ability to wash people: shower’. A similar codification of measurable function was developed for each of the other critical healthy living practices.
Table 1. The seven critical tests for a house to ‘pass’ assessment for a functioning shower.
Hot water delivered to shower
Cold water delivered to shower
Hot water temperature in acceptable/safe range (>45° and<62°)
Hot taps functioning
Cold taps functioning
Shower rose functioning
Shower drainage of standard water load occurs
Role of local community staff, electrician and plumber
The local community staff on the survey-fix teams had the capacity to immediately fix 48 of the 250 assessed items. Tradespeople were present on the first day of all projects. As each house survey was completed the survey forms were returned to the site office where data were entered and then job orders printed for tradespersons to start work on recorded failures within 24 hours of the initial assessment. As part of this process all licensed tradespersons were required to classify every failed item that was fixed as being due to: routine maintenance; faulty installation or specification; or damage, overuse or misuse.
During each day of the survey-fix stage data collected by teams on the survey sheets were checked and then entered on to a computer established in a local community project office. Following this, data validity and missing data check processes were performed.
The database that has been developed has on its main functional specifications the generation of:
1An immediate, prioritised job list for the electricians and plumbers.
2A detailed record of the status of all items checked on all houses and thus a mechanism for generating a detailed report back to the community.
3A future scope of works for both building and essential service needs in each community.
4A chart demonstrating point-in-time results at the conclusion of each stage.
The system design is optimised to assess the state of housing in the selected communities and to efficiently assess changes in status following fix work.
Follow-up work after initial survey-fix assessment (SF1)
Projects were not initiated unless funds could be first secured for significant upgrade of housing and infrastructure as identified in the survey process. For each community the project manager would determine a scope of works at the conclusion of the survey and then proceed to organise contract letting and the work to proceed. The second survey-fix assessment (SF2) was a means of determining the impact of both the initial survey-fix process as well as this subsequent major upgrade work.
We report here on data collected from projects performed in 132 communities surveyed between January 1999 and November 2006. This involved 4,343 houses in 132 projects in four States and the Northern Territory assessed at SF1 and 3,448 houses in 112 of those projects so far completed for SF2. The funds available for assessment/fix/repair work ranged from $3,500 to $7,500 average per house. Overall 1,141 (78%) of total staff on these projects were local community Aboriginal or Torres Strait Islander people. The geographic distribution of these projects is shown in Figure 1 .
Since the focus of these projects is on functioning health hardware, we have presented the data to reflect the proportion of house/living areas in which the critical number of tests were passed for a series of safety issues or identified healthy living practices. We have chosen to present the data on infrastructure safety items and on the key items for hygiene practice and nutrition. We consider these of most interest to the public health readership. In the near future detail on pass/fail criteria for all categories will be placed on a website, www.healthabitat.com. In addition, we have indicated the proportionate improvement in housing performance as determined at the SF2 following the implementation of the project (see Figures 2 and 3).
This survey demonstrates extraordinarily poor performance of Aboriginal houses. At SF1 only 11% of houses passed national standard assessment for electrical safety. In 50% of houses it was not possible to wash a child in a tub or bath. A functioning shower was available in only 35% of houses. The criteria for functioning nutritional hardware included: storage space for food, preparation bench space, functioning stove and sink. Only 6% of houses met these criteria at initial assessment (SF1).
Health hardware performance at SF2 varied between regions depending primarily on baseline state of housing and the amount of funding available for fix work after SF1.
During the projects, 71,869 items assessed as requiring fixing were inspected by licensed electricians or plumbers (a small proportion by other tradespeople) and so far 49,499 of these have been fixed. The items were categorised by relevant tradespeople as requiring repair or replacement for the following reasons: routine maintenance – 65%; faulty installation or equipment – 25%; and damage or misuse – 10%.
These data demonstrate the detail of how Aboriginal housing fails to provide the basic requirements for healthy living. At SF1 only 11% of houses passed a standard assessment for electrical safety. In 50% of houses it was not possible to wash a child in a tub or bath. A functioning shower was available in only 35% of houses. Adequate facilities to store, prepare and cook meals were present in only 6% of houses.
There are five main mechanisms by which improvement in the functioning of health hardware may improve health status:
• Reduction in electrocution and electrical fires.
• Reduction in infectious disease, especially in children, many of which have an impact on health in later life.
• Provision of the essential prerequisites for improved nutrition.
• Improved control of the living environment and reduction in the daily ‘problem list’ for Aboriginal people in communities.
• Providing employment and transferable competencies.
Our work focuses on measuring and improving the functioning of health hardware in Aboriginal housing and not on measuring health outcomes. Measuring these outcomes in disparate, small denominator communities is fraught with multiple methodological problems. In our view this should not detract from the implications of these findings since there are two centuries of public health literature on the relationship between housing and health.5,6 The bulk of this evidence confirms that improved housing improves health. What varies is effect size and the precise nature of the health benefit demonstrated in a very heterogeneous literature, all of which has methodological limitations.
Major improvements in mortality occurred during the 19th century in concert with improvements in hygiene and health hardware infrastructure.5 The broad conclusions from this period are consistent with the large number of hygiene and health studies performed in developing communities and other disadvantaged populations.7
In the early 20th century there was considerable interest and publication on housing in the medical literature.8–13 None of this work involved controlled trial interventions, but its strength was trying to define the problems with housing for the poor and a focus on change.
Particularly in the past 30 years, studies have also emphasised the deleterious mental health effects of poor housing. In a cohort study with 33-year follow-up, Marsh et al. concluded that poor housing in childhood leads to poorer mental and physical health in later life. They also reported a “dose-response effect”.14 Another more recent cohort study also found that the adverse health consequences of poor childhood housing was independent of socio-economic status.15
In a systematic review of the relationship between housing and health, Thompson et al. reviewed the literature from 1887 to 2000.6 Reflecting the major methodological, logistic and ethical difficulties in such work, they found only 11 prospective studies and only six of these with a control group. Their conclusion is likely to apply to work in Indigenous communities:
“… many studies showed health gain but small study population and a lack of control for confounders limits generalisability”.6
In our view, it is unlikely that any feasible study in Australia would add substantially to the existing, albeit imperfect, evidence. Sixty years ago in a conclusion that remains salient for researchers and policy makers in Australia today, Britten observed:
“… the inability to define the precise influence of the various elements of bad housing must not be an excuse for failing to make progress in improving housing conditions”.11
At the present time there is a barely muted inference from a small group of researchers, bureaucrats and politicians that further expenditure on Aboriginal housing should only occur if these small communities are able to prove a direct and immediate health benefit. In our view, this is both scientifically illogical and socially unethical.
Detailed repair data from this survey repudiates the long-standing myth that housing in these communities is poor because “Aborigines destroy their houses”. We are not aware of any data that support this widespread contention. In contrast, our data and conclusions come from the assessment of independent tradespersons (primarily electricians and plumbers) who repaired and categorised the reason the repair was required for 41,885 items identified. Only 10% of these were categorised as due to householder damage, overuse, misuse or vandalism.
A major strength of this work is that the survey process is linked to an immediate ‘fix’ action, which means that householders obtain an immediate benefit from the project. This has resulted in strong support for the work by community members. The methodology has been copyrighted and all funding agencies are required to sign a licensing agreement. We have found this a powerful lever to prevent government agencies from reducing or changing initial commitments once projects have started and is a key factor in maintaining program integrity and quality.
This work has important implications for housing and health policy for Indigenous people around Australia.
1The major causes of ‘house failure’– i.e. failure to deliver functioning health hardware – are lack of routine maintenance and poor initial construction and not, as commonly attributed, a failure to consider issues of cultural appropriateness.
2New housing projects need to ensure that construction, supervision and inspection are adequately planned and funded. This is not the situation at present and it is a major reason for house failure and housing stock loss.
3Maintenance programs must be regionally planned, funded and involve local community staff if loss of housing stock is to be prevented. Prioritisation of maintenance funding should be directed at health hardware.
4Sustained maintenance programs potentially provide a far more reliable source of long-term local employment than new housing projects.
5Health education and promotion messages that exhort Aborigines to take up ‘healthy behaviour’– eat healthy food, wash your children daily – are unlikely to be successful unless sustainable improvements in health hardware can be achieved.
6Detailed analysis of maintenance items and experience from these programs has provided an extensive evidence base for the National Indigenous Housing Guide.16 This should be used for all future housing design and construction in Aboriginal and Islander communities.
The public health fraternity generally responds to the social determinants agenda in Aboriginal health by urging government to provide ‘more’ or ‘better housing’. We think this approach lacks both sophistication and effectiveness. Like health, it is the detail of housing policy that matters and public health practitioners who advocate for better housing need to understand those details if they are to work effectively in this area.
These data provide a detailed picture of the state of health hardware in Aboriginal communities across Australia. Importantly, the standardised, objective and reproducible assessment provides a mechanism for determining accurately whether change occurs over time and hence the value of different interventions and housing programs.