Tracking deaths of people who have experienced homelessness: a dynamic cohort study in an Australian city.

OBJECTIVE: Life expectancy and rates of premature death are fundamental markers of health and social equity globally, and measures on which people experiencing homelessness face enormous disparities. However, unlike for other population groups with similar disparities, concerted government action to reduce homeless mortality is rare, partly due to a lack of reliable, timely data. Contemporaneous tracking of homeless deaths is required to render such deaths less invisible and measure trends over time. Drawing on multiple data sources as recommended by the US National Health Care for the Homeless Council's seminal Homeless Mortality Data Toolkit, we routinely and contemporaneously capture, verify and report on deaths occurring among people who have experienced homelessness in the Australian city of Perth. DESIGN: Dynamic cohort study. SETTING: Perth, Western Australia, Australia, between 2016 and 2022, with deaths examined between 2020 and 2022. PARTICIPANTS: For this study, the cohort comprised 8753 people who experienced homelessness in Perth, with 'recruitment' into the cohort governed by engagement with one or more local homelessness services and programmes over the period. OUTCOME MEASURES: Number and median age-at-death statistics. RESULTS: There were 360 deaths over the 3-year period, which is likely an undercount. The median age at death was 50 years, >3 decades below the current Australian median age at death of 82 years. Aboriginal people accounted for 30% of the deaths. CONCLUSION: The ongoing poor health and premature death of people who have experienced homeless are indictments on our society. Triangulation of multiple data sources is required to identify and monitor deaths among homeless populations. Timely, verified data on homeless mortality are important for galvanising action and accountability, and targets should be set to reduce the observed three-decade life expectancy chasm.

Vape stores in Western Australia: growth, proximity to schools and socio-economic gradient of density.

OBJECTIVE: To audit the number and location of dedicated e-cigarette retailers ("vape stores") in Western Australian (WA), measure proximity to schools and model the association between vape store density and socio-economic disadvantage. METHODS: Vape stores were identified via internet search and geocoded. Proximity to schools was measured. Regression modelling was used to investigate the association between vape store density and socio-economic disadvantage. RESULTS: 194 stores were identified, with 88% located within one kilometre of a school. In metropolitan WA, vape store density was nearly seven times higher in the most socio-economically disadvantaged areas than in the least disadvantaged areas (rate ratio 6.9, 95% confidence interval 3.4-15.5). CONCLUSIONS: There has been rapid, recent growth in the number of vape stores in WA, with most located within walking distance of schools. In metropolitan WA, vape store density is strongly associated with socio-economic disadvantage, mirroring the pattern observed globally for tobacco outlets. IMPLICATIONS FOR PUBLIC HEALTH: This is the first Australian study demonstrating that vape stores are more densely located in socio-economically disadvantaged areas. Vape stores' proximity to schools may increase young people's access and exposure to promotional signage. There is a need to address ready e-cigarette availability through strong regulatory and compliance measures.

Investigating local variation in disease rates within high-rate regions identified using smoothing.

Exploratory disease maps are designed to identify risk factors of disease and guide appropriate responses to disease and helpseeking behaviour. However, when produced using aggregatelevel administrative units, as is standard practice, disease maps may mislead users due to the Modifiable Areal Unit Problem (MAUP). Smoothed maps of fine-resolution data mitigate the MAUP but may still obscure spatial patterns and features. To investigate these issues, we mapped rates of Mental Health- Related Emergency Department (MHED) presentations in Perth, Western Australia, in 2018/19 using Australian Bureau of Statistics (ABS) Statistical Areas Level 2 (SA2) boundaries and a recent spatial smoothing technique: the Overlay Aggregation Method (OAM). Then, we investigated local variation in rates within high-rate regions delineated using both approaches. The SA2- and OAM-based maps identified two and five high-rate regions, respectively, with the latter not conforming to SA2 boundaries. Meanwhile, both sets of high-rate regions were found to comprise a select number of localised areas with exceptionally high rates. These results demonstrate how, due to the MAUP, disease maps that are produced using aggregate-level administrative units are unreliable as a basis for delineating geographic regions of interest for targeted interventions. Instead, reliance on such maps to guide responses may compromise the efficient and equitable delivery of healthcare. Detailed investigation of local variation in rates within high-rate regions identified using both administrative units and smoothing is required to improve hypothesis generation and the design of healthcare responses.

Improving the Efficiency of Geographic Target Regions for Healthcare Interventions.

Appropriate prioritisation of geographic target regions (TRs) for healthcare interventions is critical to ensure the efficient distribution of finite healthcare resources. In delineating TRs, both 'targeting efficiency', i.e., the return on intervention investment, and logistical factors, e.g., the number of TRs, are important. However, existing approaches to delineate TRs disproportionately prioritise targeting efficiency. To address this, we explored the utility of a method found within conservation planning: the software Marxan and an extension, MinPatch ('Marxan + MinPatch'), with comparison to a new method we introduce: the Spatial Targeting Algorithm (STA). Using both simulated and real-world data, we demonstrate superior performance of the STA over Marxan + MinPatch, both with respect to targeting efficiency and with respect to adequate consideration of logistical factors. For example, by design, and unlike Marxan + MinPatch, the STA allows for user-specification of a desired number of TRs. More broadly, we find that, while Marxan + MinPatch does consider logistical factors, it also suffers from several limitations, including, but not limited to, the requirement to apply two separate software tools, which is burdensome. Given these results, we suggest that the STA could reasonably be applied to help prevent inefficiencies arising due to targeting of interventions using currently available approaches.

Predicting Future Geographic Hotspots of Potentially Preventable Hospitalisations Using All Subset Model Selection and Repeated K-Fold Cross-Validation.

Long-term future prediction of geographic areas with high rates of potentially preventable hospitalisations (PPHs) among residents, or "hotspots", is critical to ensure the effective location of place-based health service interventions. This is because such interventions are typically expensive and take time to develop, implement, and take effect, and hotspots often regress to the mean. Using spatially aggregated, longitudinal administrative health data, we introduce a method to make such predictions. The proposed method combines all subset model selection with a novel formulation of repeated k-fold cross-validation in developing optimal models. We illustrate its application predicting three-year future hotspots for four PPHs in an Australian context: type II diabetes mellitus, heart failure, chronic obstructive pulmonary disease, and "high risk foot". In these examples, optimal models are selected through maximising positive predictive value while maintaining sensitivity above a user-specified minimum threshold. We compare the model's performance to that of two alternative methods commonly used in practice, i.e., prediction of future hotspots based on either: (i) current hotspots, or (ii) past persistent hotspots. In doing so, we demonstrate favourable performance of our method, including with respect to its ability to flexibly optimise various different metrics. Accordingly, we suggest that our method might effectively be used to assist health planners predict excess future demand of health services and prioritise placement of interventions. Furthermore, it could be used to predict future hotspots of non-health events, e.g., in criminology.

Multimorbidity among People Experiencing Homelessness-Insights from Primary Care Data.

BACKGROUND: Although the poor health of people experiencing homelessness is increasingly recognised in health discourse, there is a dearth of research that has quantified the nature and magnitude of chronic health issues and morbidity among people experiencing homelessness, particularly in the Australian context. METHODS: Analysis of the medical records of 2068 "active" patients registered with a specialist homeless health service in Perth, Western Australia as of 31 December 2019. RESULTS: Overall, 67.8% of patients had at least one chronic physical health condition, 67.5% had at least one mental health condition, and 61.6% had at least one alcohol or other drug (AOD) use disorder. Nearly half (47.8%) had a dual diagnosis of mental health and AOD use issues, and over a third (38.1%) were tri-morbid (mental health, AOD and physical health condition). Three-quarters (74.9%) were multimorbid or had at least two long-term conditions (LTCs), and on average, each patient had 3.3 LTCs. CONCLUSIONS: The study findings have substantial implications from both a health risk and healthcare treatment perspective for people experiencing homeless. The pervasiveness of preventable health conditions among people experiencing homelessness also highlights the imperative to improve the accessibility of public health programs and screening to reduce their morbidity and premature mortality.

Modelling the Relationship between Rainfall and Mental Health Using Different Spatial and Temporal Units.

Drought is thought to impact upon the mental health of agricultural communities, but studies of this relationship have reported inconsistent results. A source of inconsistency could be the aggregation of data by a single spatiotemporal unit of analysis, which induces the modifiable areal and temporal unit problems. To investigate this, mental health-related emergency department (MHED) presentations among residents of the Wheat Belt region of Western Australia, between 2002 and 2017, were examined. Average daily rainfall was used as a measure of drought. Associations between MHED presentations and rainfall were estimated based on various spatial aggregations of underlying data, at multiple temporal windows. Wide variation amongst results was observed. Despite this, two key features were found: Associations between MHED presentations and rainfall were generally positive when rainfall was measured in summer months (rate ratios up to 1.05 per 0.5 mm of daily rainfall) and generally negative when rainfall was measured in winter months (rate ratios as low as 0.96 per 0.5 mm of daily rainfall). These results demonstrate that the association between drought and mental health is quantifiable; however, the effect size is small and varies depending on the spatial and temporal arrangement of the underlying data. To improve understanding of this association, more studies should be undertaken with longer time spans and examining specific mental health outcomes, using a wide variety of spatiotemporal units.

Impact of the modifiable areal unit problem in assessing determinants of emergency department demand.

OBJECTIVE: To examine the impact of the modifiable areal unit problem (MAUP) in an investigation of factors associated with ED demand in Perth, Western Australia, in 2016. Furthermore, to advocate a means of avoiding this impact. METHODS: ED presentations were classified as: urgent medical, non-urgent medical, urgent trauma or non-urgent trauma. In each group, sex-stratified, age-adjusted multivariate associations with socio-economic status and distance to the nearest ED and general practitioner (GP) were estimated. Modelling was undertaken using different sets of spatial units: Australian Bureau of Statistics (ABS) Statistical Areas Level 1 (SA1s) and numerous aggregate-level zonations of SA1s (ABS SA2s and others). RESULTS: Estimates obtained using the different units often varied widely: for seven (30%) of 24 strata defined by combinations of sex, ED type and covariate, the smallest and largest effect sizes differed in terms of direction; further, for 11 (65%) of the remaining 17 strata, the largest effect size was at least twice as high as the smallest. This demonstrates the MAUP's impact and that analyses based on a single set of spatial units are unreliable. To resolve the observed variation, we highlight the SA1-level estimates. CONCLUSIONS: When formulating interventions targeting reduced ED utilisation, policy planners should be guided by evidence based on analysis of appropriate spatial units. This ideal is undermined by the widespread lack of acknowledgement of the MAUP in studies examining drivers of ED demand using spatially aggregated data. To avoid the MAUP, only estimates obtained through examining a minimal geographic unit should be relied upon.

Overcoming inefficiencies arising due to the impact of the modifiable areal unit problem on single-aggregation disease maps.

BACKGROUND: In disease mapping, fine-resolution spatial health data are routinely aggregated for various reasons, for example to protect privacy. Usually, such aggregation occurs only once, resulting in 'single-aggregation disease maps' whose representation of the underlying data depends on the chosen set of aggregation units. This dependence is described by the modifiable areal unit problem (MAUP). Despite an extensive literature, in practice, the MAUP is rarely acknowledged, including in disease mapping. Further, despite single-aggregation disease maps being widely relied upon to guide distribution of healthcare resources, potential inefficiencies arising due to the impact of the MAUP on such maps have not previously been investigated. RESULTS: We introduce the overlay aggregation method (OAM) for disease mapping. This method avoids dependence on any single set of aggregate-level mapping units through incorporating information from many different sets. We characterise OAM as a novel smoothing technique and show how its use results in potentially dramatic improvements in resource allocation efficiency over single-aggregation maps. We demonstrate these findings in a simulation context and through applying OAM to a real-world dataset: ischaemic stroke hospital admissions in Perth, Western Australia, in 2016. CONCLUSIONS: The ongoing, widespread lack of acknowledgement of the MAUP in disease mapping suggests that unawareness of its impact is extensive or that impact is underestimated. Routine implementation of OAM can help avoid resource allocation inefficiencies associated with this phenomenon. Our findings have immediate worldwide implications wherever single-aggregation disease maps are used to guide health policy planning and service delivery.

Anxiety and alcohol in the working-age population are driving a rise in mental health-related emergency department presentations: 15 year trends in emergency department presentations in Western Australia.

OBJECTIVE: To evaluate age, gender and disease-specific trends in ED for mental health presentations over 15 years. METHODS: The study population consisted of residents of metropolitan Perth, Western Australia, presenting to Perth ED between 1 July 2002 and 30 June 2017. Population rates of mental health-related ED presentations per year were calculated. RESULTS: Rates of mental health ED presentations are significantly increasing in the working-age population for those with stress and anxiety-related diagnoses, particularly in younger females, and also for alcohol-related presentations for those aged 10-49 years, particularly in males. CONCLUSION: The present study demonstrates that increased rates of mental health-related ED presentations are driven by increased rates of presentation for stress and anxiety-related and alcohol-related presentations in both genders across the working-age population.

Burden of primary care-type emergency department presentations using clinical assessment by general practitioners: A cross-sectional study.

OBJECTIVE: To compare methods of assessment of the burden of primary care-type ED (PCTED) presentations against clinical assessment by general practitioners (GPs) in ED. METHODS: A cross-sectional study involving clinical assessment of patients presenting to four EDs in Western Australia. The GPs assessed patients who were likely to be discharged home from ED, and considered whether they could be managed in general practice. Patient presentations were defined by the GPs as: PCTED; PCTED if additional primary care resources were available; or not PCTED. RESULTS: GP researchers determined that 80% of patients assessed were PCTED presentations, with one-third of these considered PCTED presentations if additional resources were available. A high proportion of identified PCTED presentations included categories excluded by previous methods. Analysis of linked data found the cohort assessed to be of lower urgency, younger, and with a shorter length of stay than the average patient being discharged from ED. After accounting for potential bias, it is suggested that 20-40% of all ED presentations could be PCTED presentations. CONCLUSIONS: Previous methods determining the burden of PCTED presentations have not been validated. Many presentations excluded by previous methods were identified as manageable in general practice by GPs clinically assessing patients in ED. Improved validation of criteria used to identify PCTED presentations will enable appropriately designed interventions to reduce such events.

Reducing Bruzzi's Formula to Remove Instability in the Estimation of Population Attributable Fraction for Health Outcomes.

The aim of this study was to reconcile 3 approaches to calculating population attributable fractions and attributable burden percentage: the approach of Bruzzi et al. (Am J Epidemiol. 1985;122(5):904-914.), the maximum-likelihood method of Greenland and Drescher (Biometrics. 1993;49(3):865-872.), and the multivariable method of Tanuseputro et al. (Popul Health Metr. 2015;13:5.). Using data from a statewide point prevalence survey (Western Australian Point Prevalence Survey, 2014) linked to an administrative database, we compared estimates of attributable burden percentage obtained using the contrasting methods in 6 logistic models of health outcomes from the survey, estimating 95% confidence intervals using nonparametric and weighted bootstrap approaches. Our results show that instability can arise from the fundamental algebraic construction of Bruzzi's formula, and that this instability may substantially influence the calculation of attributable burden percentage and associated confidence intervals. These observations were confirmed in a simulation study. The algebraic reduction of Bruzzi's formula to the 2 alternative methods resulted in markedly more stable estimates for population attributable fraction and attributable burden percentage in cross-sectional studies and cohort designs with fixed follow-up time. We advocate the widespread implementation of the maximum-likelihood approach and the multivariable method.