Acute Stroke Mimics: Etiological Spectrum and Efficacy of FAST, BE FAST, and the ROSIER Scores

Abstract

Background and Purpose: Stroke mimics constitute a good number of patients referred as acute strokes within the window period for acute therapies. Proper triaging can avoid unnecessary imaging and even thrombolytic therapies in these patients. This study looked at the etiological spectrum of acute stroke mimics presenting within the 4.5 hours therapeutic window. We also evaluated the FAST, BE FAST, and the ROSIER tools in picking true strokes. Methods: Prospective study conducted over a 2-year period. Results: Acute stroke mimics constituted 328/1635 (20%) of referrals for acute stroke after screening by the neurology stroke team. Focal and generalized seizures with transient weakness and peripheral vertigo were the commonest acute stroke mimics; followed by metabolic causes and psychiatric disorders. Females were more in the stroke mimic group ( P = .02). Ischemic heart disease and atherosclerotic risk factors (except diabetes mellitus) were significantly higher among true strokes. In total, 4 (1.2%) of the stroke mimics were treated with IV thrombolysis. Diagnostic accuracy for different stroke differentiating tools are as follows: FAST (sensitivity 85.9%, specificity 52.8%, odds 6.8), BE FAST (sensitivity 97.0%, specificity 31.4%, odds 14.9), and ROSIER scale (sensitivity 85.7%, specificity 59.4%, odds 8.7%). Conclusions: Acute stroke mimics can constitute up to 20% of cases evaluated as acute strokes by neurology stroke teams. None of the triaging tools appear to have enough accuracy. A proper history and clinical examination should be given priority over fixed protocols whenever acute stroke mimic are suspected especially before administering acute costly interventions. Auditing stroke mimics is important to improve acute stroke pathways.