Differing Definitions of First-Ever Stroke Influence Incidence Estimates More than Trends: A Study Using Linked Administrative Data.

INTRODUCTION: Researchers apply varying definitions when measuring stroke incidence using administrative data. We aimed to investigate the sensitivity of incidence estimates to varying definitions of stroke and lookback periods and to provide updated incidence rates and trends for Western Australia (WA). METHODS: We used linked state-wide hospital and death data from 1985 to 2017 to identify incident strokes from 2005 to 2017. A standard definition was applied which included strokes coded as the principal hospital diagnosis or the underlying cause of death, with a 10-year lookback used to clear prevalent cases. Alternative definitions were compared against the standard definition by percentage difference in case numbers. Age-standardised incidence rates were calculated, and age- and sex-adjusted Poisson regression models were used to estimate incidence trends. RESULTS: The standard definition with a 10-year lookback period captured 31,274 incident strokes. Capture increased by 19.3% when including secondary diagnoses, 4.1% when including nontraumatic subdural and extradural haemorrhage, and 8.1% when including associated causes of death. Excluding death records reduced capture by 11.1%. A 20-year lookback reduced over-ascertainment by 2.0%, and a 1-year lookback increased capture by 13.3%. Incidence declined 0.6% annually (95% confidence interval -0.9, -0.3). Annual reductions were similar for most definitions except when death records were excluded (-0.1%, CI: -0.4, 0.2) and with the shortest lookback periods (greatest annual reduction). CONCLUSION: Stroke incidence has declined in WA. Differing methods of identifying stroke influence estimates of incidence to a greater extent than estimates of trends. Reductions in stroke incidence over time are primarily driven by declines in fatal strokes.

The Allure of Big Data to Improve Stroke Outcomes: Review of Current Literature.

PURPOSE OF REVIEW: To critically appraise literature on recent advances and methods using "big data" to evaluate stroke outcomes and associated factors. RECENT FINDINGS: Recent big data studies provided new evidence on the incidence of stroke outcomes, and important emerging predictors of these outcomes. Main highlights included the identification of COVID-19 infection and exposure to a low-dose particulate matter as emerging predictors of mortality post-stroke. Demographic (age, sex) and geographical (rural vs. urban) disparities in outcomes were also identified. There was a surge in methodological (e.g., machine learning and validation) studies aimed at maximizing the efficiency of big data for improving the prediction of stroke outcomes. However, considerable delays remain between data generation and publication. Big data are driving rapid innovations in research of stroke outcomes, generating novel evidence for bridging practice gaps. Opportunity exists to harness big data to drive real-time improvements in stroke outcomes.

Low dose CT detected interstitial lung abnormalities in a population with low asbestos exposure.

BACKGROUND: The use of low dose CT (LDCT) chest is becoming more widespread in occupationally exposed populations. There is a knowledge gap as to heterogeneity in severity and the natural course of asbestosis after low levels of exposure. This study reports the characteristics of LDCT-detected interstitial lung abnormalities (ILA). METHODS: The Asbestos Review Program offers annual LDCT, health assessments, and pulmonary function tests to an asbestos-exposed cohort. Asbestosis was defined using the Helsinki Consensus statement and the presence of ILA defined using a protocol for occupational CT reports. At least two of three pulmonary function tests: forced expiratory volume in 1 s (FEV1 );​ forced vital capacity (FVC); and diffusion capacity for carbon monoxide (DLco) were required for analysis of physiological decline. RESULTS: From 1513 cases, radiological ILA was present in 485 (32%). The cohort was 83.5% male with a median age of 68.3 years and a median (IQR) asbestos exposure of 0.7 (0.09-2.32) fiber/ml-year. A mixed occupation, mixed asbestos fiber cohort comprised the majority of the cohort (65.8%). Of those with ILA, 40 (8.2%) had an FVC decline of ≥10% and 30 (6.2%) had a DLco decline of ≥15% per year. Time since first exposure, increasing tobacco exposure and reported dyspnea were independently associated with the presence of ILA. CONCLUSIONS: In this population with relatively low asbestos exposure, LDCT-detected ILA that fits criteria for asbestosis is common, but physiological decline is not. This mild chronic stable phenotype of asbestos-associated ILA contrasts with the traditionally accepted views that asbestosis requires high exposures.

Autoantibodies and cancer among asbestos-exposed cohorts in Western Australia.

Asbestos exposure is associated with many adverse health conditions including malignant mesothelioma and lung cancer as well as production of autoantibodies. Autoantibodies may serve as biomarkers for asbestos exposure in patients with cancer, and autoimmune dysfunction has been linked to increased rates of various cancers. The aim of this study was to examine the hypothesis that autoantibodies are more frequent in asbestos-exposed individuals with either lung cancer or mesothelioma than those without these conditions. Asbestos-exposed individuals from Western Australia who had lung cancer (n = 24), malignant mesothelioma (n = 24), or no malignancy (n = 51) were tested for antinuclear autoantibodies (ANA) using indirect immunofluorescence and specific extractable nuclear autoantibodies (ENA) employing a multiplexed addressable laser bead immunoassay. Contrary to the hypothesis, data demonstrated that individuals without malignancy were more likely to be positive for ANA compared to those with cancer. However, autoantibodies to histone and Ro-60 were found to be associated with lung cancer. These results support a possible predictive value for specific autoantibodies in the early detection of lung cancer and/or in our understanding of the role of autoimmune processes in cancer. However, further studies are needed to identify specific target antigens for the antibodies.

Trends in incidence, mortality rates, and survival of colorectal cancer in Western Australia from 1990 to 2014: a retrospective whole-population longitudinal study.

PURPOSE: This study aimed to investigate the trends in colorectal cancer (CRC) incidence and mortality rates among the Western Australian (WA) population. This study further compared the trends with the timing of the implementation and rollout of the National Bowel Cancer Screening Program (NBCSP) and examined the survival predictors in CRC cases. METHODS: This study was a whole-population, retrospective longitudinal study and included all individuals with a confirmed histological diagnosis of primary invasive CRC diagnosed in WA from 1990 to 2014 (n = 25,932). The temporal trends were assessed by Joinpoint regression models and Kaplan-Meier survival curves were used to asses 5-year survival. Predictors of survival were examined using multivariable Cox proportional hazard regression models, adjusting for age of diagnosis. RESULTS: The overall CRC incidence showed an upward trend between 1990 and 2010 (annual percent change (APC) = 1.1%); then, there was a downward trend from 2010 to 2014 (APC = - 5.0%). In younger people (< 50 years), the incidence rate increased steadily (APC = 0.9%) over the study period. The overall CRC mortality trend increased from 1990 to 1999 (APC = 1.6%), decreasing after that (APC = - 2.1%). Younger people had better CRC-related 5-year survival than older people (HR = 0.81, 95%CI 0.75-0.87, p = < 0.001). CONCLUSION: This study found that CRC incidence and mortality rates decreased among older people over the last 10 years in Western Australia. However, incidence continues to rise for younger people. Hence, more widespread adoption of the screening program, and potential preventive and early diagnostic strategies should become key priorities for the CRC control in WA.

Association between diesel engine exhaust exposure and lung function in Australian gold miners.

Previous studies have reported that miners (and other workers) exposed to high levels of diesel engine exhaust (DEE) have an increased risk of lung function decline. The main objective of this study was to evaluate associations between exposure to different components associated with DEE in relation to lung function across a 12-h working shift. Eighty underground gold miners and twenty surface miners completed spirometry and questionnaires at the beginning and end of their 12 h work shift. Personal exposure to elemental carbon (EC), volatile organic compounds (VOCs), nitrogen dioxide (NO2), particle size and particle number were monitored during their shift. Multiple regression models were used to examine the associations between DEE and lung function, adjusting for a range of covariates. Underground miners were exposed to higher levels of EC, VOCs, NO2, and particle number and larger mean particle size than surface miners. Cross-shift reduction in Z-score value of FEV1/FVC in underground miners was statistically significantly greater than those of surface miners. The cross-shift change in Z-score value of FEV1/FVC was associated with exposure to higher concentration of EC and particle number, but not with VOCs, NO2 and particle size. Occupational exposure to diesel engine exhaust in current Australian gold mines is substantial. Exposures were higher in underground miners and had a negative association with their lung function over a single 12-h shift.

Pleural Plaques and the Risk of Lung Cancer in Asbestos-exposed Subjects.

Rationale: Asbestos exposure is associated with a dose-dependent risk of lung cancer. The association between lung cancer and the presence of pleural plaques remains controversial.Objectives: To define the relationship between pleural plaques and lung cancer risk.Methods: Subjects were from two cohorts: 1) crocidolite mine and mill workers and Wittenoom Township residents and 2) a mixed-asbestos-fiber, mixed-occupation group. All subjects underwent annual review since 1990, chest X-ray or low-dose computed tomography scan, and outcome linkage to national cancer and mortality registry data. Cox regression, with adjustment for age (as the underlying matching time variable), was used to estimate hazard ratios (HRs) for lung cancer incidence by sex, tobacco smoking, asbestos exposure, presence of asbestosis, and pleural plaques.Measurements and Main Results: For all 4,240 subjects, mean age at follow up was 65.4 years, 3,486 (82.0%) were male, 1,315 (31.0%) had pleural plaques, and 1,353 (32.0%) had radiographic asbestosis. Overall, 3,042 (71.7%) were ever-smokers with mean tobacco exposure of 33 pack-years. In total, 200 lung cancers were recorded. Risk of lung cancer increased with cumulative exposure to cigarettes, asbestos, and presence of asbestosis. Pleural plaques did not confer any additional lung cancer risk in either cohort (cohort 1: HR, 1.03; 95% confidence interval, 0.64-1.67; P = 0.89; cohort 2: HR, 0.75; 95% confidence interval, 0.45-1.25; P = 0.28).Conclusions: The presence of pleural plaques on radiologic imaging does not confer additional increase in the risk of lung cancer. This result is consistent across two cohorts with differing asbestos fiber exposures and intensity.

The Wittenoom legacy.

The Wittenoom crocidolite (blue asbestos) mine and mill ceased operating in 1966. The impact of this industry on asbestos-related disease in Western Australia has been immense. Use of the employment records of the Australian Blue Asbestos Company and records of the Wittenoom township residents has permitted two cohorts of people with virtually exclusive exposure to crocidolite to be assembled and studied. Follow-up of these two cohorts has been conducted through data linkage with available hospital, mortality and cancer records. The evolution of asbestos-related disease has been recorded and, with the establishment of exposure measurements, quantitative exposure-response relationships have been estimated. There has been an ongoing epidemic of mortality from lung cancer and malignant mesothelioma and, less so, from asbestosis. Wittenoom crocidolite was used extensively in asbestos-cement products in Western Australia. As a result, the state has recorded a higher malignant-mesothelioma mortality rate than in any other Australian state and in any defined general population in the world. Thus, the legacy of Wittenoom has extended beyond the mine and the town, and is still evident more than 50 years after the closure of the mine.

Measurement of urinary 1-aminopyrene and 1-hydroxypyrene as biomarkers of exposure to diesel particulate matter in gold miners.

Metabolites of polycyclic aromatic hydrocarbons measured in human samples are often used as biomarkers of exposure to diesel engine exhaust (DEE). The aim of this study was to assess the changes in urinary levels of 1-aminopyrene (1-AP) and 1-hydroxypyrene (1-OHP) and their relationship with Elemental Carbon (EC), as a component of diesel engine exhaust exposure, among a hard-rock gold-mining population. Urine samples were collected at the beginning and end of a 12-hour work shift from 100 underground and above ground gold miners. Miners were fitted with personal exposure monitoring equipment to quantify exposure to DEE, measured as Elemental Carbon (EC), across their 12-hour work shift. General linear regression assessed associations of the post-shift urinary 1-AP and 1-OHP concentrations with EC, controlling for age, gender, the pre-shift biomarker level, Body Mass Index (BMI), days on current shift, time in mining, smoking status and second-hand smoke exposure. The concentrations of 1-AP and 1-OHP increased significantly across a 12-hour mining work shift. Moreover, consistent with the sensitivity analysis, the concentration of 1-AP was significantly associated with EC after adjustments. Urinary 1-OHP, but not 1-AP was significantly associated with current smoking. Urinary 1-AP may be a more robust and specific biomarker of DEE than 1-OHP.

Validation of an Asbestos Job-Exposure Matrix (AsbJEM) in Australia: Exposure-Response Relationships for Malignant Mesothelioma.

OBJECTIVES: An asbestos job-exposure matrix (AsbJEM) has been developed to systematically and cost-effectively evaluate occupational exposures in population-based studies. The primary aim of this study was to examine the accuracy of the AsbJEM in determining exposure-response relationships between asbestos exposure estimates and malignant mesothelioma (MM) incidence (indirect validation). The secondary aim was to investigate whether the assumptions used in the development of the original AsbJEM provided accurate asbestos exposure estimates. METHODS: The study population consisted of participants in an annual health surveillance program, who had at least 3-month occupational asbestos exposure. Calculated asbestos exposure indices included cumulative asbestos exposure and the average exposure intensity, estimated using the AsbJEM and duration of employment. Asbestos and MM exposure-response relationships were compared between the original AsbJEM and its variations based on manipulations of the intensity, duration and frequency of exposure. Twenty-four exposure estimates were calculated for both cumulative asbestos exposure and the average exposure intensity using three exposure intensities (50th, 75th and 90th percentile of the range of mode exposure), four peak durations (15, 30, 60 and 120 min) and two patterns of peak frequency (original and doubled). Cox proportional hazards models were used to describe the associations between MM incidence and each of the cumulative and average intensity estimates. RESULTS: Data were collected from 1602 male participants. Of these, 40 developed MM during the study period. There were significant associations between MM incidence and both cumulative and average exposure intensity for all estimates. The strongest association, based on the regression-coefficient from the models, was found for the 50th percentile of mode exposure, 15-min peak duration and the doubled frequency of peak exposure. Using these assumptions, the hazard ratios for mesothelioma were 1 (reference), 1.91, 3.24 and 5.37 for the quartiles of cumulative asbestos exposure and 1 (reference), 1.84, 2.31 and 4.40 for the quartiles of the average exposure intensity, respectively. CONCLUSION: The well-known positive exposure-response relationship between MM incidence and both estimated cumulative asbestos exposure and average exposure intensity was confirmed. The strongest relationship was found when the frequency of peak exposure in the AsbJEM was doubled from the originally published estimates.

Are children more vulnerable to mesothelioma than adults? A comparison of mesothelioma risk among children and adults exposed non-occupationally to blue asbestos at Wittenoom.

OBJECTIVES: The presence of asbestos in public buildings is a legacy of past asbestos use in many developed countries. Of particular concern is the amount and current condition in schools and the vulnerability of children to mesothelioma. Our aim was to compare the risk of mesothelioma between those exposed to blue asbestos as children and as adults at Wittenoom. METHODS: Public sources were used to establish the Wittenoom residents' cohort. Mesothelioma incidence rates per 100 000 person-years at risk were derived for those first exposed to asbestos at Wittenoom as children (<15 years) or adults separately. Proportional hazards survival models examined the slope of the exposure-response relationship between asbestos exposure and incidence of mesothelioma in different sex and age groups. RESULTS: The mesothelioma rate was lower among those first exposed as children (76.8 per 100 000) than those first exposed as adults (121.3 per 100 000). Adjusting for cumulative exposure to asbestos and sex, those exposed as adults had a greater risk of mesothelioma (adjusted HR 2.5, 95% CI 1.7 to 3.7). The slope of the exposure-response relationship did not differ between those exposed as children and those exposed as adults. CONCLUSION: We found no greater susceptibility to mesothelioma among those first exposed to asbestos as children than those first exposed as adults. However, given the long latency of mesothelioma, and the greater years of life yet to be lived by the Wittenoom children, it is likely that there will be more cases of mesothelioma in the future among those first exposed as children.

Increasing Incidence of Colorectal Cancer in Adolescents and Young Adults Aged 15-39 Years in Western Australia 1982-2007: Examination of Colonoscopy History.

AIMS: To examine trends in colorectal cancer (CRC) incidence and colonoscopy history in adolescents and young adults (AYAs) aged 15-39 years in Western Australia (WA) from 1982 to 2007. DESIGN: Descriptive cohort study using population-based linked hospital and cancer registry data. METHOD: Five-year age-standardized and age-specific incidence rates of CRC were calculated for all AYAs and by sex. Temporal trends in CRC incidence were investigated using Joinpoint regression analysis. The annual percentage change (APC) in CRC incidence was calculated to identify significant time trends. Colonoscopy history relative to incident CRC diagnosis was examined and age and tumor grade at diagnosis compared for AYAs with and without pre-diagnosis colonoscopy. CRC-related mortality within 5 and 10 years of incident diagnosis were compared for AYAs with and without pre-diagnosis colonoscopy using mortality rate ratios (MRRs) derived from negative binomial regression. RESULTS: Age-standardized CRC incidence among AYAs significantly increased in WA between 1982 and 2007, APC = 3.0 (95% CI 0.7-5.5). Pre-diagnosis colonoscopy was uncommon among AYAs (6.0%, 33/483) and 71% of AYAs were diagnosed after index (first ever) colonoscopy. AYAs with pre-diagnosis colonoscopy were older at CRC diagnosis (mean 36.7 ± 0.7 years) compared to those with no prior colonoscopy (32.6 ± 0.2 years), p < 0.001. At CRC diagnosis, a significantly greater proportion of AYAs with pre-diagnosis colonoscopy had well-differentiated tumors (21.2%) compared to those without (5.6%), p = 0.001. CRC-related mortality was significantly lower for AYAs with pre-diagnosis colonoscopy compared to those without, for both 5-year [MRR = 0.44 (95% CI 0.27-0.75), p = 0.045] and 10-year morality [MRR = 0.43 (95% CI 0.24-0.83), p = 0.043]. CONCLUSION: CRC incidence among AYAs in WA has significantly increased over the 25-year study period. Pre-diagnosis colonoscopy is associated with lower tumor grade at CRC diagnosis as well as significant reduction in both 5- and 10-year CRC-related mortality rates. These findings warrant further research into the balance in benefits and harms of targeted screening for AYA at highest risk.

Cancer incidence in the Western Australian mining industry (1996-2013).

BACKGROUND: Miners are frequently exposed to established and potential carcinogens. We aimed to assess cancer incidence in miners relative to the general population and identify high-risk subgroups. METHODS: Incident cancers in Western Australian miners (n=153,922; 86% male) during 1996-2013 were identified. Indirectly standardised incidence ratios (SIRs) were calculated and mixed-effects Poisson models were used to calculate Incidence Rate Ratios (IRRs) to identify high-risk within-cohort subgroups. RESULTS: Compared with the general population, the overall cancer incidence in miners (n=4194 cases) was lower for both females (SIR:0.83, 95%CI:0.74-0.92) and males (SIR:0.96, 95%CI:0.93-0.99). Overall, cancer incidence did not differ by employment duration or employment commencement time. Ever-underground work was associated with lung cancer (IRR:1.81, 95%CI:1.11-2.93). Relative to multi-ore miners, IRRs for specific cancers were significantly different when exclusively mining: iron (prostate:0.73, 95%CI:0.56-0.94); gold (lung:1.77, 95%CI:1.04-3.01 and colorectum:1.70, 95%CI:1.16-2.51); and other metals (urinary tract:1.85, 95%CI:1.03-3.31 and leukaemia:0.36, 95%CI:0.14-0.96). CONCLUSION: Working underground emerged as a significant determinant of lung cancer risk in our contemporary mining cohort. Increased risks of lung, prostate, colorectal and urinary tract cancers and leukaemia were identified in miners of specific ores. These findings underline the importance of continued surveillance of the health and exposures of this relatively young cohort of miners.

Pre- and post-sentence mental health service use by a population cohort of older offenders (≥45 years) in Western Australia.

PURPOSE: Information on older offenders' mental health service (MHS) used before and after sentence is sparse. We therefore aimed to determine the 1-year prevalence of MHS use before sentence, and the likelihood and predictors of MHS use in the 5-year post-sentence period by first-time older adult offenders (≥45 years). METHODS: Pre- and post-sentence MHS use by a cohort of 1,853 first-time offenders over 45 years in Western Australia was determined through whole-population linked administrative data. Logistic regression models compared the 1-year pre-sentence MHS contacts between offenders and matched non-offenders. Cox proportional hazards regression models identified the socio-demographic, offending and pre-sentence health service variables that determined post-sentence MHS use. RESULTS: Older offenders were six times more likely to have used MHSs than non-offenders before sentence. Substance use was the most commonly treated disorder. Non-custodial offenders were twice more likely than prisoners to have been treated for any mental disorder and substance use disorders, and violent offenders were four times more likely to have attempted self-harm than non-violent offenders before being sentenced. The strongest predictors of post-sentence MHS contact were past psychiatric diagnosis in offenders with a pre-sentence MHS contact, and pre-sentence hospitalisation for attempted self-harm or physical illness, or being a male in those without. Discontinuity in MHS use after sentence by over half of the offenders with a prior contact was prominent. CONCLUSION: Better detection and treatment of mental disorders in older offenders to ensure continuity of care at all transition points through age-sensitive correctional and community-based MHSs is needed.

Predictors of post-sentence mental health service use in a population cohort of first-time adult offenders in Western Australia.

BACKGROUND: Little is known about whether or how offenders use mental health services after sentence completion. AIM: This study aimed to determine the likelihood of such service use by adult (18-44 years) first-time offenders up to 5 years after sentence completion and possible predictor variables. METHODS: Pre-sentence and post-sentence mental health service use was obtained from whole-population linked administrative data on 23,661 adult offenders. Cox proportional hazard models were used to determine which socio-demographic, offending and pre-sentence health service variables were associated with such post-sentence service use. RESULTS: The estimated 5-year probability of any post-sentence mental health service use was 12% for offenders who had not previously used such services, but still only 42% for those who had. For the latter, best predictors of post-sentence use were past psychiatric diagnosis and history of self-harm; history of self-harm also predicted post-sentence use among new mental health services users and so also did past physical illness. Indigenous offenders had a greater likelihood of service use for any mental disorder or for substance use disorders than non-Indigenous offenders, irrespective of pre-sentence use. Among those with pre-sentence service contact, imprisoned offenders were less likely to use mental health services after sentence than those under community penalties; in its absence, socio-economic disadvantage and geographic accessibility were associated with greater likelihood of post-sentence use. CONCLUSIONS: Our findings highlight the discontinuity of mental healthcare for most sentenced offenders, but especially prisoners, and suggest a need for better management strategies for these vulnerable groups with mental disorders.

Pre-sentence mental health service use predicts post-sentence mortality in a population cohort of first-time adult offenders.

PURPOSE: With the high risk of death associated with mental disorders and their increased prevalence in offenders, it is judicious to investigate the risk of post-sentence mortality with respect to offenders' psychiatric treatment history. METHODS: Using linked administrative data for a whole-population retrospective cohort of first-time adult offenders (n = 25,537) sentenced to either prison or non-custodial orders in Western Australia, we determined the risk and baseline predictors of post-sentence mortality. RESULTS: Of 192 deaths within 2 years of sentence completion, deaths from injury/poisoning (55.6 %), cancer (13.3 %) and cardiovascular disorders (9.7 %) were the most common. Pre-sentence history of mental health service (MHS) contact doubled the risk of post-sentence all-cause and injury/poisoning-related mortality. Physical comorbidity was the strongest predictor of mortality irrespective of pre-sentence MHS contact. Baseline history of attempted self-harm and being an Indigenous male were associated with an elevated risk of death in offenders with a pre-sentence MHS contact. In offenders without a pre-sentence MHS contact, socio-economic disadvantage and incarceration almost doubled the risk of dying from any cause and injury/poisoning. CONCLUSIONS: Mortality risk in the 2 years following sentence completion is associated with pre-sentence health service use and a range of socio-demographic factors for both incarcerated and non-custodial offenders. The opportunity afforded by imprisonment could be exploited by provision of funding to identify and treat mental illness, impart preventive health education addressing modifiable risk factors and provide transitional care to community-based services, all of which may help reduce preventable post-sentence deaths. Diversion to non-custodial sentences is also a plausible option.

Pre-sentence mental health service use by adult offenders in Western Australia: baseline results from a longitudinal whole-population cohort study.

BACKGROUND: Little is known about community mental health service use prior to offending. AIM: This study aimed to investigate the 1-year community mental health service use by adult offenders (18-44 years) prior to their first ever criminal sentence in Western Australia. METHODS: Administrative mental health service data were compared between all offenders (n = 23,755) commencing their first ever criminal sentence in Western Australia during 1985-1994 and a matched community group of 21,977 non-offenders. RESULTS: Just over 8% of offenders had used mental health services prior to sentence compared with 1% of non-offenders. After adjusting for age, offenders were more likely to have used these services than non-offenders in all gender-race groups, but the effect was strongest for non-Indigenous women, who were over 12 times more likely to have used such services, and weakest among Indigenous men, who were about twice as likely to have used them as their non-offending peers. Service use for substance use disorder, the most common diagnosis, was about one and a half times more prevalent among Indigenous than non-Indigenous offenders, regardless of gender. For non-Indigenous offenders, prevalence of any mental health service contact was higher for violent than non-violent offenders, irrespective of gender. Service use was no different between offenders receiving custodial or non-custodial sentences in all gender-race groups. CONCLUSION: The higher likelihood of mental health service use by offenders in the year prior to their first ever sentence than by non-offenders suggests that, insofar as the disorder was relevant to offending, there were some opportunities for preventive measures during that time. Differential service use according to gender and Indigenous/non-Indigenous status is of concern. It would be important to understand more about this apparently unequal service access, not least because Indigenous populations tend to be over-represented in prison.