on behalf of the PRESTO investigators.
Prehospital stroke scales are suggested to estimate the likelihood of proximal intracranial large vessel occlusion (LVO) in patients with suspected stroke symptoms. We aimed to compare the prehospital assessment by paramedics to the in-hospital NIHSS.
We analyzed data from the PRESTO study, a prospective multicenter observational cohort study of suspected stroke patients transferred by ambulance. Paramedics scored nine prehospital stroke scale items in the field. We calculated Spearman’s correlation between the prehospital assessment and in-hospital NIHSS assessed at the emergency department, stratified by diagnosis. Secondary analyses were the sensitivity and specificity for each prehospital item against the corresponding NIHSS item (a score of 0 vs. ≥1).
We included 1036 patients; 120 (12%) patients had an ischemic stroke due to LVO. Correlation between the prehospital assessment and NIHSS was 0.72 (95% CI: 0.69–0.75). Correlations subdivided per diagnosis were 0.81 (95% CI: 0.71–0.88) for patients with intracranial hemorrhage, 0.72 (95% CI: 0.63–0.80) for patients with LVO, 0.68 (95% CI: 0.62–0.73) for patients with non-LVO ischemic stroke, 0.64 (95% CI: 0.56–0.70) for patients with stroke mimic and 0.27 (95% CI: 0.13–0.40) for patients with TIA. The prehospital items “Arm” and “Leg” had the highest sensitivity (97%, 94%) with the lowest specificity (63%, 62%) (Table). “Neglect” had the lowest sensitivity (34%) with the highest specificity (93%).
Prehospital assessments by paramedics are strongly correlated with in-hospital assessments, but not in patients with TIA. Only neglect is missed frequently by paramedics. Our results indicate that paramedics can adequately assess neurological deficits in the field.
Not applicable