RESUMEN
BACKGROUND: With newly-extended treatment windows for endovascular therapy in emergent large vessel occlusions, it is increasingly important to identify thrombectomy-eligible patients without overwhelming resources dedicated to acute stroke care. We devised a simple paradigm to classify patient's presenting neurologic symptoms to screen for large vessel occlusions. METHODS: We reviewed the presenting symptoms, imaging findings, and final diagnoses of consecutive emergency department stroke alert cases. Patients were classified based on their neurologic exams as focal objective, focal subjective, or nonfocal. Outcomes of final diagnoses of acute ischemic stroke and large vessel occlusions were compared across groups. Comparisons were made to other large vessel occlusion prediction scales. RESULTS: Of 521 patients, 342 (65.6%) were categorized as focal objective, 142 (27.2%) as focal subjective, and 37 (7.1%) as nonfocal. Ischemic stroke and large vessel occlusions were diagnosed in 114 (21.9%) and 27 (5.2%) of patients, respectively. Classification as focal objective significantly predicted stroke (odds ratio 3.77; 95% confidence interval 2.17-6.55) and captured all large vessel occlusions (P = .0001). The focal objective categorization was the only tool which achieved 100% sensitivity for large vessel occlusions (with a specificity of 36%) compared to other large vessel occlusion prediction tools. CONCLUSIONS: Patients who presented as stroke alerts without focal neurologic symptoms were unlikely to have large vessel occlusions. With high sensitivity, classifying patients' neurologic exams into focal objective versus subjective or nonfocal categories may serve as a useful tool to screen for large vessel occlusions and prevent unnecessary emergent workup in patients unlikely to be endovascular candidates.