2026 WA Health & Care Hackathon - Challenges

WA Health & Care Hackathon 2026

Challenge Information

Challenge 1 · HBF

Predicting cancellations

Can you help HBF predict cancellations?

Explore which factors predict different cancellation types, how early predictions can be made, and which interventions might suit different groups.

1. Predictive variables
Which variables are the most predictive of each cancellation type?
2. Prediction granularity
How granularly can you predict different types of cancellation?
3. Early prediction
How early can you predict each type of cancellation?
4. Segments & interventions
Segment these outputs - What interventions will suit each group?

Data: More than 12 million records covering approximately 24 months, at monthly intervals.
Suggested approaches: Prototype model, notable discoveries and intervention approach design.

Challenge 2

Alternative care pathways

Can you help identify and act on patterns that may indicate opportunities for alternative care pathways for mental health and aged care services in our hospitals?

Explore patterns in mental health emergency service use (2A) and how age and repeat ED presentations among adults aged 65+ relate to admissions and inpatient resource use (2B).

2A. Mental health services
What accounts for the greatest use of mental health emergency services (attendance and admission), and is this primarily driven by the number of consumers requiring care, recurrent episodes of care, or other underlying patterns of presentation?
2B. Adults aged 65+
Among adults (65+) who present to ED, how does age and presentation frequency relate to hospital admission and the broader inpatient resource burden?

Data: Synthesised Department of Health Data Collection Extract.
Suggested approaches: Recommendation models, cohort analysis, next-best-pathway modelling and simulation. The overall challenge theme also includes UX and design thinking.

Challenge 3

Clinician leave coverage

When a clinician asks for leave, how might an AI agent recommend the best approach to cover, or explain exactly why those dates won’t work?

Explore how an AI agent could recommend cover for clinician leave or explain why the requested dates cannot be accommodated.

Data: Synthetic versions of staffing data (role, ward and classification), roster data, leave request scenarios, and business rules and processes.
Suggested approaches: Agents, workforce simulations and recommendation engines. Consider who your user is.