Analytical approach
Data, methods, and responsible interpretation
The research combines qualitative and observational methods to understand EMS allocation as a complex public-sector healthcare system, with careful attention to interpretability, privacy, and responsible use of data.
Data sources
- Routinely collected EMS and dispatch data
- Emergency medical dispatch metadata
- Timestamps and response-time intervals
- Priority levels and call categories
- On-scene triage and clinical risk classification
Study designs
- Register-based cohort studies
- Qualitative interviews
- Retrospective observational analyses
- System-level performance analysis
- Interpretability-first modelling
Analytical principles
- Distributional analysis
- Attention to tail delays
- Heterogeneity and subgroup patterns
- Transparent limitations
- Privacy-by-design reporting
Responsible interpretation
The analyses are designed to support system understanding. Observational data can reveal patterns and associations, but they must be interpreted with attention to selection, prioritisation, missing information, and the fact that response time is often shaped by the same urgency assessments that later appear in the outcome pathway.
Where machine learning is used, it is not presented as a deployable triage or dispatch system. Its role is to make nonlinear patterns, heterogeneity, and system-level relationships easier to inspect and discuss responsibly.
