91ɫ��Ƭ

National Injury Surveillance for Actionable Research - Emergency Department

$2.98M MRFF funded National Critical Research Infrastructure Grant (2024-2029)

NISAR-ED is developing a scalable national approach to injury surveillance using routinely collected emergency department data. The program combines epidemiology, clinical data science, artificial intelligence, natural language processing and ICD-11 classification to identify injury mechanism, intent, activity, location and other external causes of injury.

I initiated and lead the program as Chief Investigator and Project Lead, working with the AIHW, ACCC Product Safety Branch, Australasian College for Emergency Medicine, MUARC, state and territory health departments and other research and policy partners.

The program also provides an applied research platform for student projects in injury intent, self-harm, family and domestic violence, data quality, ICD-11 classification, product safety and clinical AI.

$2.51 million MRFF Primary Health Care Research Grant, (2024–2029)

The PACE-RURAL project is evaluating an evidence-based pathway of care for people with chronic musculoskeletal conditions living in rural, regional and remote Australia. The project uses allied-health telehealth and a digital clinical resource to improve access to care and health outcomes.��I am a contributing investigator and lead the health-economic evaluation of the stepped-wedge clinical implementation trial. My work includes assessing the costs, outcomes and cost-effectiveness of the intervention.��The project also supports student research in reproducible data management, data quality assurance, clinical trial data preparation and the planned cost-utility analysis.

Research collaboration: road safety, linked data and injury outcomes

I am a contributing investigator and research collaborator on a national road-safety research program led by Professor Rebecca Ivers. The project uses linked crash, emergency department, hospital, mortality, ambulance, workers’ compensation and other health datasets to examine predictors and consequences of serious road injury in New South Wales.

My contribution draws on my expertise in injury epidemiology, emergency department data, linked health data and population-based injury outcomes. I am contributing to the research program, related publications and interpretation of findings for road-safety policy and prevention.

Trauma registries, clinical notes and large language models

I am a research collaborator on a multidisciplinary project examining whether locally deployed large language models can extract injury characteristics, mechanism of injury and injury severity from emergency department clinical notes with .

The project compares information extracted from clinical notes with gold-standard trauma registry data at a major Australian trauma centre. It forms part of a broader program of work examining how artificial intelligence and natural language processing can improve the efficiency, completeness and scalability of injury surveillance and trauma registry processes.

Text mining, self-harm and suicide among people with intellectual disability

I am a Chief Investigator on a 91ɫ��Ƭ Medicine & Health and Australian Human Rights Institute-funded project examining health inequalities and self-harm and suicide behaviours among people with intellectual disability.

The project uses text-mining methods to identify and better understand self-harm and suicide-related presentations and behaviours in health data. It aligns with my broader research program in injury surveillance, self-harm prevention, clinical data science and the use of routinely collected data to identify populations whose health needs may otherwise remain poorly captured.

Self-harm, suicidal ideation and injury intent classification

I supervise and collaborate on a program of research examining how intentional self-harm, suicidal ideation and injury intent are identified and classified in emergency department data.

This work includes comparison of ICD-10-AM short-list approaches with ICD-11 classification, evaluation of keyword-based and machine-learning methods, human-in-the-loop validation and development of improved approaches to identifying self-harm presentations in routinely collected clinical data.

Current and recent student projects include:

• Hasan Rakibul: AI-enabled classification of injury intent to improve national self-harm surveillance.
• Joanne Miles: comparison of ICD-10-AM and ICD-11 classification of intentional self-harm and suicidal ideation presentations.
• Matthew Young: comparison of keyword search and machine-learning methods for identifying self-harm presentations.
• Kaehla Robinson: human-in-the-loop validation of natural-language-processing-derived self-harm and suicidality classifications.
• Madeleine Walker: health-service use following intentional self-harm among older Australians.

Family and domestic violence surveillance

I supervise research examining the use of longitudinal emergency department data to improve identification and surveillance of family and domestic violence.��This work focuses on the limitations of routine clinical documentation and coded data, and on how data-linkage and data-science methods may support earlier identification of violence-related presentations and better understanding of patterns of service use.

Current student research includes:

• Xin (Mia) Ma: improving the early identification of family and domestic violence through longitudinal analysis of Northern Territory emergency department data.

Injury classification, ICD-11 and emergency department data quality

I lead and supervise research examining the quality and information content of routinely collected emergency department data for injury surveillance.��This includes the use of structured and free-text clinical information, presenting-problem data, injury intent, external-cause classification and ICD-11 injury and extension codes. The work aims to identify what information is available in routine emergency department data, what is missing, and how data-quality improvements and artificial intelligence can support surveillance-quality outputs.

Current student research includes:

• Ryan (Zongrui) Dai: quantifying the information content of emergency department data for injury surveillance.
• Joanne Miles: comparing ICD-10-AM and ICD-11 approaches to classification of self-harm and suicidal ideation.
• Hasan Rakibul: improving classification of injury intent using artificial intelligence and epidemiological methods.

Reproducible health-data management and quality assurance

I supervise research focused on the preparation, documentation and quality assurance of clinical trial and health-service datasets.��This work recognises that reproducible data workflows are essential for reliable analysis, transparent research and future implementation. It includes data dictionaries, cleaning and validation scripts, quality-control reporting and preparation of analysis-ready datasets.

Current student research includes:

• Gina Liao: development of a reproducible patient-data management, quality-assurance and preparation pipeline for the PACE-RURAL trial.

One Health research and training

I supervise a public health capstone project examining the development of One Health research and training in Australia and New Zealand over the past decade.��The project will replicate and extend earlier work examining the scope, disciplinary breadth and integration of One Health research and education across human, animal and environmental health.

Current student research includes:

• Jismy Baiju: systematic review of One Health research and training in Australia and New Zealand: ten years on.

Injury prevention, product safety and emerging hazards

I contribute to research examining how injury surveillance, clinical data and artificial intelligence can identify emerging product and recreational safety hazards.��This work includes product-related injury, playground and recreational injury, emerging consumer-product risks and the use of emergency department data to inform prevention and regulatory action. It connects my research through NISAR-ED with the work of government and industry partners, including the ACCC Product Safety Branch.

Planned student research includes:

• Clinical AI Honours project, commencing 2027: detection of emerging product safety hazards using clinical AI and emergency department data, undertaken with the ACCC Consumer Product Safety Branch as a NISAR-ED partner organisation.

Systematic review of spinal subdural haemorrhage in abusive head trauma

I supervise a Master of Public Health systematic review examining the incidence of spinal subdural haemorrhage among young children with abusive head trauma.��The project synthesises the available evidence on this important but relatively uncommon injury pattern and considers its implications for clinical assessment, imaging and child-protection practice.

Current student research includes:

• Dr Jenny Sohn (Paediatric Fellow): systematic review of the incidence of spinal subdural haemorrhage among young children with abusive head trauma.

Public health evaluation and health economics

I supervise and contribute to applied research evaluating public health programs, clinical interventions and health-service pathways.��My work includes economic evaluation, implementation outcomes, resource use, cost-effectiveness and translation of evidence into policy and practice. Current projects include the economic evaluation of PACE-RURAL and qualitative analysis of patient-reported information relevant to healthcare utilisation and costs.

Current student research includes:

• Charles Houssarini: qualitative evaluation and classification of free-text patient responses relevant to healthcare utilisation and costs, contributing to the planned cost-utility evaluation of PACE-RURAL.

Digital transformation, work health and safety, and human rights

Following a panel presentation at the World Safety Conference, I am developing a collaborative paper with ��(Hellenic Mediterranean University) and ��(Maastricht University) examining safety and health in the digital transformation of work. Our current work��considers how artificial intelligence, automation, platform work and algorithmic management are changing occupational health and safety risks. It develops a sociotechnical perspective that connects injury prevention, public health, worker rights, professional autonomy and regulatory protection.

Broader injury epidemiology and health-service research

My broader research program also includes traumatic spinal cord injury, emergency and acute-care pathways, road safety, workplace injury, self-harm and suicide prevention, family and domestic violence, product safety, playground and recreational injury, health economics and the evaluation of health-service interventions.

Across these activities, my research focuses on improving the quality and usefulness of health data so that injury and other forms of preventable harm can be identified earlier, understood more accurately and addressed through evidence-based prevention and policy.

��

��

��

��

��

��