Abstract TMP1: Predicting Scale Design and Derivation to Select Acute Ischemic Stroke Patients for Direct Angiosuite Intervention

Abstract

Introduction: Vascular imaging after head CT to confirm large vessel occlusion (LVO) for acute stroke patients requires additional time and delays recanalization. We developed the T hrombectomy A cute M echanical re P erfusion A ssessment ( TAMPA ) scale for selecting patients with LVO for direct angiosuite transfer and intervention to improve recanalization time. Methods: The TAMPA scale was developed from our prospectively collected “Get with the Guidelines” database. We included all “stroke alert” patients between 1/2017 and 8/2018 with vascular imaging and National Institutes of Health Stroke Scale scores between 5 and 25. We excluded patients with immediately obvious non-stroke diagnoses, those lacking subsequent vascular imaging, or those with incomplete records. Different variables were collected. The TAMPA scale receiver operating characteristics curve (ROC) was compared with the ROCs of other commonly used scales. Results: 571 eligible patients from 2115 “acute stroke alerts” were identified for developing the TAMPA scale. The scale was established with a combination of 5 items with a total score of 9: CT hyperdense sign, parenchymal hypodensity, lateralizing hemiparalysis, gaze deviation and speech disturbance. A cutoff of ≥ 4 yielded a sensitivity of 68.98%, specificity of 72.91%, false positive rate of 27.09%, and false negative rate of 31.02%. Compared with other scales, such as total NIHSS, C-stat/CPSSS, RACE, FAST-ED and 3I-SS, the TAMPA scale had the best ROC for the selected group of patients. Conclusions: The TAMPA scale accurately predicts presence of clinically amenable LVO in patients with moderate to severe ischemic stroke. Use of the TAMPA scale to identify high probability mechanical embolectomy candidates for direct transfer to the angiosuite could potentially reduce revascularization times and increase treatment rates.