Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Machine Learning Improves Emergency Dispatch Prioritization During Ambulance Shortages
A Swedish emergency dispatch trial demonstrates that machine learning-assisted risk assessment can help nurses more accurately identify and prioritize high-risk patients when ambulance demand exceeds availability.
Background
Emergency medical dispatch centers face critical challenges when patient demand outpaces ambulance supply. Under these resource-constrained conditions, dispatchers must rapidly assess patient severity using clinical notes and caller information—a complex decision-making process conducted under significant time pressure.
Key Findings
The MADLAD trial tested a gradient boosting model at two Swedish dispatch centers from 2021–2024. Among cases where ambulances were limited:
- 68.3% of intervention patients received ambulances routed to the highest-risk individual (by National Early Warning Score) versus 62.5% in standard care
- Dispatchers accepted the model’s recommendations 80.9% of the time
- Model performance remained stable over four years with no degradation
- Intervention group patients experienced shorter dispatch times (29 vs 32 minutes)
Why It Matters
Though the absolute improvement is modest, this trial provides strong evidence that machine learning can enhance clinical decision-making under resource constraints. The model’s sustained performance across a real-world dispatch system suggests practical feasibility for implementation in diverse emergency settings.
Limitations
The authors emphasize that maximizing benefit requires combined strategies addressing both model accuracy and dispatcher compliance. Higher dispatcher training and institutional support may unlock greater clinical gains.
Original paper: Machine learning assisted differentiation of low acuity patients at dispatch: The MADLAD randomized controlled trial. — PLoS medicine. 10.1371/journal.pmed.1004770
