Data sources
Queensland Suicide Register (QSR). The QSR is a public health surveillance suicide mortality database which includes all suspected and confirmed suicide fatalities which have occurred in Queensland since 1990 [26]. The QSR sources data from police reports from the Queensland Police Service and the Coroners Court of Queensland; and post-mortem examinations, toxicology reports, and coronial findings from the National Coronial Information System (NCIS). The Queensland Registry of Births, Deaths, and Marriages supplies information on Indigenous status missing from primary data sources. The QSR includes information on a wide range of demographic, psychosocial, psychiatric, medical, contextual, and behavioural aspects of suicide cases. This information is used to classify the probability of suicide is classified as either ‘unlikely’, ‘possible’, ‘probable’ or ‘confirmed’ using a decision-tree [26]. This study included suicides classed as ‘probable’ or ‘confirmed’. The current study used data from the QSR on all suicides of First Nations people during the period 2001–2015 in Queensland from the QSR. The last known or usual residential address is recorded in the QSR and was further coded as Statistical Area Level Two (SA2) for this study. Statistical Area Level Twos (SA2) are the second smallest unit of the Australian Statistical Geography Standard used in the census of which there were 526 Queensland SA2s in the 2011 Census, they cover the entire state and have an average population of 10,000 people. To calculate suicide rates by SA2, the 2011 Census was used as the closest census to the midpoint of the period of analysis. It is acknowledged that significant cultural, historic, and social differences exist between the many First Nations on these lands. However, due to the limitations of available population denominator data which does not further disaggregate, comparisons presented include all Aboriginal and Torres Strait Islander individuals as 'First Nations/Indigenous Peoples.'
SA2 Social Indicators for the Indigenous Population (Synthetic estimates). Estimates were accessed from the Australian Urban Research Infrastructure Network (AURIN) system. The National Centre for Social and Economic Modelling (NATSEM) produces these synthetic estimates using a spatial microsimulation model combining geographic information from the 2011 Census with the richer data from the 2008 National Aboriginal and Torres Strait Islander Social Survey (NATSISS) to gain reliable small-area estimates of Indigenous social indicators. The NATSISS collects information on demographic, social, cultural, environmental, and economic indicators, including geography, language and cultural activities, social networks and support, health and well-being, education, employment, income, personal safety and experiences of discrimination. It is completed by Aboriginal and Torres Strait Islander residents aged 15 years and older [48]. Due to the sensitivity and potential for identification, confidentialised unit record files (CURFs) from the survey dataset are not available. Instead, a synthetic unit record file built with observations summing to the population totals from the survey was extracted using the ABS Survey Tablebuilder system [49]. A spatial microsimulation approach was then applied to this synthetic unit record file using a generalised regression reweighting algorithm that reweights the synthetic unit record file to small-area Census benchmarks. These benchmarks include characteristics like income, age, sex and education (see [50] for full benchmarks list). The results are validated against the aggregated NATSISS survey data. The model automatically eliminates any SA2 that fails to achieve the threshold of acceptable error (the Total Absolute Error from the model is greater than the population of the SA2), in which case the model would not provide an estimate [49], [50]. The SA2s in Queensland that did not meet the threshold were predominantly industrial and commercial areas in Brisbane.
The indicators selected for this study from this model (and originally from the NATSISS 2008) were:
Discrimination score reflects the percentage of Aboriginal and Torres Strait Islander residents reporting racial discrimination in the last 12 months. The scope includes discrimination in the workplace, community, educational institutions, recreational settings, court setting, treatment by the police, health professionals or government agency staff when seeking any public services. In these SA2s, between 9.4% and 50.9% of the First Nations population reported discrimination in the previous year. Areas with high levels of discrimination were those in which 30% or more of the population reported discrimination in the previous year.
Indigenous language use score reflects the percentage of Aboriginal and Torres Strait Islander residents who report speaking an Indigenous language in the home. The proportion of the First Nations population speaking an Indigenous language ranged from 0–96.41%. Due to the positively skewed language use distribution, areas were delineated as high and low based on being either above or equal to and below 5% of First Nations persons speaking an Indigenous language.
Australian Standard Geographical Standard - Remoteness Areas (ASGS-RA). The Remoteness Areas measure remoteness or conversely accessibility of areas to goods and services using five categories - major cities, inner regional, outer regional, remote, and very remote [51]. Inner regional, outer regional, remote, and very remote were merged into a single variable ‘Regional and Remote areas’ for comparison with ‘Metropolitan areas’ (major cities).
Index of Relative Socioeconomic Advantage and Disadvantage (IRSAD). The IRSAD is a Socio-Economic Indexes for Areas (SEIFA) index ranking areas in Australia according to comparative socioeconomic advantage and disadvantage and includes variables related to education (year 11 or lower), occupation (professional, low skills, management) disabilities, overcrowded housing, and income [52]. SA2 Quintile ranks were used for this study with the three lowest quintiles classified as ‘low.’