Kirsten Vallmuur

Electric personal mobility devices, including e-scooters, e-bikes, e-skateboards and similar devices, have proliferated worldwide since 2017, with many jurisdictions introducing public e-scooter hi...

Objective To describe the personal, crash and claim characteristics of Aboriginal and Torres Strait Islander people who submitted a claim under the New South Wales (NSW) Compulsory Third Party (CTP) scheme and explore the differences in characteristics of submitting a claim and associated factors. Methods This linked data study used NSW police-reported crash data linked with NSW CTP claims data from 2018 to 2023. Descriptive statistics summarised claimant characteristics, and multivariable logistic regression models identified factors associated with lodging a claim. Results From 2018 to 2023, 1954 Aboriginal and Torres Strait Islander people lodged a claim under the NSW CTP scheme. Most claimants were male (52.7%) and lived in areas of low socioeconomic status. Higher odds of claiming were associated with being female (OR 1.16, 95% CI 1.01 to 1.33), older age (OR 1.01, 95% CI 1.01 to 1.02, per 1-year increase) and living in metropolitan areas (OR 1.65, 95% CI 1.34 to 2.01) and holding a standard licence (OR 3.83, 95% CI 2.83 to 5.19). Sustaining moderate injuries (OR 0.27, 95% CI 0.23 to 0.31) was associated with lower odds of claiming compared with serious injuries. Conclusions CTP claim rates among Aboriginal and Torres Strait Islander people in NSW are low relative to their burden of road transport injury, suggesting that they may face barriers in accessing CTP compensation. Improving support, especially for those in rural areas and young drivers, and simplifying the claims process could enhance access to compensation for Aboriginal and Torres Strait Islander people, potentially improving recovery outcomes and reducing disparities.

Accurate and efficient injury coding is critical for effective injury surveillance. Machine learning (ML) models trained on manually-coded historical injury records have been used to predict the injury codes efficiently and with reasonable accuracy but their accuracy has been limited for complex narratives and rare causes of injury. In this study, we examined performance of Large Language Models (LLMs) to predict three injury codes: cause-of-injury, product-involved, and nature-of-injury, on 100 injury cases randomly selected from the National Electronic Injury Surveillance System database. The prediction performance of LLM (ChatgGPT-3.5) was compared with a traditional ML (Logistic Regression) and a neural network model (Multilayer Perceptron). We observed that LLM was better than the other two models in terms of effectively (a) extracting syntactic relationships, (b) handling misspellings and common acronyms, and (c) deciphering semantic information from the text with reasonable accuracy, even when the narratives were not in a proper grammatical format.

Purpose: Serious injuries or illnesses impose a significant burden on the affected individuals. This study examined the long-term recovery of health-related quality of life (HRQoL) after a serious injury or illness and quantified the economic costs attributable to the quality-adjusted life-years (QALYs) lost to serious injury or illness. Methods: Data were sourced from the Household, Income and Labour Dynamics in Australia (HILDA) database. The propensity score matching-difference-in-differences (PSM-DiD) method was used to explore the recovery of HRQoL over a six-year follow-up period. HRQoL was measured by the short-form six-dimension utility index, physical health, and mental health. Results: We found that those who suffered any serious injuries or illnesses had a decline of 0.020 in QALYs during the year of event, reducing further until a trough of 0.032 by the third year and thereafter began to recover. Further analyses revealed that although serious injuries or illnesses had a significantly greater loss in physical health, the loss in mental health persisted over a longer period. Finally, the economic cost associated with the lost in QALYs due to serious injuries or illnesses was estimated at $685 per person during the year of event and increased to a peak of $1,250 per person by the end of the third-year post-exposure. Conclusion: Our findings highlight that follow-up care designed to mitigate the impacts of a serious injury or illness on people’s HRQoL should consider cost-effective strategies that are long-lasting and support those affected throughout at least, the first three years of their injuries or illnesses.

Objectives: To examine the relative differences in injury hospitalisation cost estimates from hospital administrative data when using (1) hospital reimbursement based on National Efficient Price (NEP) versus local hospital costings methods, (2) inflation factors based on Consumer Price Index (CPI) versus health group index and (3) different healthcare funder’s perspectives. Design: Retrospective population-based cohort study of linked Queensland Hospital Admitted Patient Data Collection dataset and National Hospital Cost Data Collection data. Setting: All admitted injury-related care episodes occurring within a major trauma hospital in Queensland during 1 January 2012 to 31 December 2017. Outcome measure: Total in-patient hospitalisation cost associated with an episode of care. Results: Injury without catastrophic/severe complications or comorbidities was the most frequently occurring Australian Refined Diagnosis-Related Groups (AR-DRG) over the 6-year period, while rehabilitation with catastrophic complications or comorbidities was the most expensive ($37 938, 95% CI $36 067 to $39 809). Among the top 10 AR-DRGs, seven had NEP-based cost estimates substantially higher than the hospital-reported costs, with differences varying between 2.6% and 43.0%. CPI-inflated costs were significantly lower than health group index-inflated estimates, with observed differences between 7.7% (95% CI 6.9% to 8.7%) and 11.9% (95% CI 10.8% to 13.1%) for the same AR-DRG. Finally, cost estimates were significantly higher for care funded by private health insurers compared with care funded by either the public insurer or compulsory third-party injury insurers, with differences varying significantly between 8.4% (95% CI 7.2% to 30.1%) and 55.0% (95% CI 53.3% to 56.9%) for the same AR-DRG. Care funded by compulsory third-party injury insurers, however, incurred the highest cost for the most expensive AR-DRGs. Conclusion: There were considerable discrepancies in cost estimates for common injury-related hospitalisations depending on type of costing method used, inflation metrics applied and healthcare funder’s perspective adopted. These factors need to be considered when evaluating hospital cost in Australia’s health system using administrative data.

Bloodstream infection (BSI) can be a serious complication among injured patients. Understanding the burden of injury-related BSI is important for early detection and implementing appropriate treatments to improve patient outcomes. Incidence rates and risk factors are important measures that help provide insights into the burden of injury-related BSIs and early diagnosis of patients. In this review, the aim was to comprehensively summarize incidence rates and risk factors for injury-related BSIs from scientific literature. Four electronic databases (PubMed, CINAHL [via EBSCOhost], Embase, and Web of Science) were searched. There were no limitations on the language. Studies reporting the incidence rates or risk factors associated with incidence or adverse outcomes from injury-related BSIs were included. Database searches returned 9830 articles, of which 48 were included. Incidence rates of injury-related BSIs ranged from 0.71 to 27.4 episodes per 1000 patient-days. A total of 237 potential factors associated with the development and/or outcomes of injury-related BSIs were identified and classified into 8 broad categories: demographics, prognostic scores, burn extent, clinical and patient health factors, biomarkers, resource utilization and treatments, pathogens and injuries, and mechanisms. Older age, male sex, higher injury severity score, longer length of stay, greater total body surface burn area, and inhalation injuries were the most frequently reported risk factors. This review identified a large variation in reported incidence rates but no population-based studies. Many factors have been associated with injury-related BSIs; however, the direction of association and effect sizes vary across the studies, which can be attributed to the differences in study design.

The recent COVID-19 pandemic stimulated unprecedented linkage of datasets worldwide, and while injury is endemic rather than pandemic, there is much to be learned by the injury prevention community from the data science approaches taken to respond to the pandemic to support research into the primary, secondary and tertiary prevention of injuries. The use of routinely collected data to produce real-world evidence, as an alternative to clinical trials, has been gaining in popularity as the availability and quality of digital health platforms grow and the linkage landscape, and the analytics required to make best use of linked and unstructured data, is rapidly evolving. Capitalising on existing data sources, innovative linkage and advanced analytic approaches provides the opportunity to undertake novel injury prevention research and generate new knowledge, while avoiding data waste and additional burden to participants. We provide a tangible, but not exhaustive, list of examples showing the breadth and value of data linkage, along with the emerging capabilities of natural language processing techniques to enhance injury research. To optimise data science approaches to injury prevention, injury researchers in this area need to share methods, code, models and tools to improve consistence and efficiencies in this field. Increased collaboration between injury prevention researchers and data scientists working on population data linkage systems has much to offer this field of research.