Background and Aims
Mechanical thrombectomy demands reliable pre-hospital triage by paramedics. Although optimal score has not been established, it is not clear what level of competency exists between paramedics for diagnosis of major neurological deficits. The aim of our study was to define inter-rater reliability between paramedics and stroke neurologists for the presence of severe hemiparesis.
Method
Prospective, double phase, multicentre study. During the first phase (April – October 2016), paramedics were educated to diagnose mild or severe hemiparesis via Internet e-learning. Their participation in the training was voluntary. During the second phase (August-October 2017) paramedics were re-educated via webinars and examined straight after. Agreement for degree of hemiparesis [NIHSS, item 4 and 5, scoring 0-2 (none or mild) versus3-4 (severe)] between paramedics (assessed during pre-hospital) and neurologists (assessed immediately after stroke centre admission) was analysed using the unweighted Kappa index.
Results
During the first phase consecutive 435 (47% of all) patients (215 men [51%], average age 73 and during the second phase 71 patients (51%) (35 men [50%], average age 74 were evaluated for the presence of hemiparesis by both paramedics and neurologists. In both phases the agreement between the paramedics and neurologists was moderate – in the first one kappa 0.54 (95%CI 0.46-0.62) in the second kappa 0.43 (95% CI 0.27-0.63).
Conclusion
In our study, reliability of assessment of hemiplegia by paramedics was only moderate and was not improved by repeated training. More effective trainings of paramedics are needed especially with the prospect of clinical need for pre-hospital triage for mechanical thrombectomy.

INTER-RATER RELIABILITY BETWEEN PARAMEDICS AND NEUROLOGISTS IN ASSESSMENT OF HEMIPLEGIA IN ACUTE STROKE DID NOT IMPROVE AFTER RE-EDUCATION