Validation of a shortened FAST-ED algorithm for smartphone app guided stroke triage

Area of Science:

• Neurology
• Emergency Medicine
• Digital Health

Background:

• Large vessel occlusion (LVO) recognition scales are crucial for identifying acute ischemic stroke (AIS) patients requiring immediate transfer to a comprehensive stroke center (CSC).
• Pre-hospital stroke recognition tools aim to optimize patient routing and reduce delays in treatment.

Purpose of the Study:

• To validate a smartphone application utilizing a shortened Field Assessment Stroke Triage for Emergency Destination (FAST-ED) for pre-hospital LVO-related AIS triage.
• To compare the diagnostic accuracy of the app-based FAST-ED with existing LVO recognition scales.

Main Methods:

• Retrospective analysis of 2815 patients with confirmed stroke or stroke mimics admitted by emergency medical services.
• Evaluation of the predictive accuracy of a digital, app-based shortened FAST-ED (FAST-ED App) for LVO-related AIS.
• Comparison of the FAST-ED App's performance against original FAST-ED, RACE, CPSS, 3-ISS, and NIHSS scales.

Main Results:

• The FAST-ED App demonstrated high predictive accuracy for LVO-related AIS (Area under ROC = 0.887), comparable to the original FAST-ED and RACE scales.
• The app significantly outperformed CPSS, 3-ISS, and NIHSS in identifying LVO-related AIS.
• A FAST-ED App score of 4 or higher showed high sensitivity (77%) and specificity (87%) for detecting LVO-related AIS.

Conclusions:

• The validated smartphone app-based FAST-ED provides a high-quality tool for pre-hospital stroke triage.
• This digital approach effectively identifies patients with LVO, potentially streamlining transport to comprehensive stroke centers.