To evaluate the use of dispatch narratives for risk stratification in suspected cardiopulmonary emergencies.
Approach:
Study Design: This study analyzes unstructured data from emergency dispatch narratives to enhance risk assessment in EMS.
Data Utilization: Natural language processing techniques are used to extract relevant clinical patterns from emergency call narratives.
Key Findings:
Dispatch narratives contain critical clinical indicators often missed by structured data.
Natural language processing improves the assessment of unstructured data in emergency dispatch.
Current EMS systems lack tools for effective risk stratification based on initial call data.
Interpretation:
The study highlights the importance of utilizing detailed narrative accounts from emergency calls to improve prehospital decision-making and risk stratification.
Limitations:
Prior research has focused on detailed clinical narratives rather than time-sensitive dispatch communications.
Previous NLP models in EMS have primarily addressed isolated high-acuity conditions, lacking a comprehensive view of various cardiopulmonary presentations.
Conclusion:
This study proposes a framework to support early decision-making in EMS using unstructured data from emergency calls.
