Inter-Rater Reliability of the FAST-ED in the Out-of-Hospital Setting.

Introduction: Patients experiencing a large vessel occlusion stroke (LVOS) may require endovascular-capable centers and benefit from direct transport to such facilities, creating a need for an accurate prehospital assessment. The Field Assessment Stroke Triage for Emergency Destination (FAST-ED) is a secondary scale to identify LVOS. Currently, there is limited prospective evidence validating the use of the FAST-ED in the prehospital environment. This study aimed to evaluate the inter-rater reliability of the FAST-ED between patient care providers in the prehospital setting. Methods: This prospective study was conducted between 4/1/2018 and 7/1/2018 in a single municipal EMS agency that staffs two providers per ambulance with at least one being a paramedic. Patients were included based on paramedic impression that the patient was both having a stroke and greater than 18 years old. Each provider independently performed and documented a FAST-ED assessment on eligible patients. Data analysis consisted of performing inter-rater reliability using Cohen's Kappa on the FAST-ED score between primary and secondary providers. The FAST-ED was analyzed on an item level, an aggregate level (cumulative of all items), and using the defined cut point of ≥4. A sub-analysis determined if inter-rater reliability changed across provider certification. Results: There were 231 patients included in this analysis with an average age of 68.5 years and 135 (58.4%) female. Inter-rater reliability varied across individual items in the scale from 90.1% agreement to 82.5%. When analyzing inter-rater reliability of the aggregate FAST-ED score, the scale demonstrated 70.1% agreement (Kappa 0.66), considered substantial agreement. FAST-ED scores were analyzed using a cut point of ≥4. When using this cut point, there was 92.2% (Kappa 0.81) agreement between primary and secondary caregiver, demonstrating almost perfect agreement. Agreement was substantial across provider certifications including paramedics and EMTS. Conclusion: This study demonstrated high inter-rater reliability of the FAST-ED scale when performed in the prehospital setting on patients suspected of having a stroke. There were minimal differences in reliability based on provider certification, and item level analysis indicated substantial inter-rater reliability.