Chhabra, Nikita

Introduction: Recent studies report similar outcomes with antiplatelet drugs and intravenous thrombolysis (IVT) in patients with minor nondisabling ischemic stroke. These patients still meet emergent triage criteria and can increase the burden on stroke systems through prehospital stroke activation. Prehospital large vessel occlusion (LVO) scales can improve triage of LVO patients but have not been evaluated as tools to predict IVT administration. We sought to determine if common LVO scales can predict IVT use in a large academic telestroke practice. Methods: This retrospective study included all video telestroke consults performed in a large academic telestroke network from 2019-2020. Patients were excluded if they were on anticoagulation, presented beyond 4.5 hours from last known well, or family declined IVT due to goals of care. We recorded presenting NIHSS and IVT recommendations. NIHSS was used to calculate 7 LVO scales (RACE, G-FAST, FAST-ED, VAN, PASS, 3I-SS, and C-STAT). Scale performance was assessed by calculating area under curve (AUC) and compared to NIHSS total score using the DeLong test. Youden cutpoint was established for each scale and subsequently used to calculate sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy. Results: A total of 513 patients were included; 205/513 (40%) were recommended IVT. Median (IQR) NIHSS was 1 (0-3) in non-IVT patients and 6 (4-11) in IVT patients. The RACE and G-FAST scales were noninferior to the NIHSS in predicting IVT recommendation (Table 1). Conclusion: The RACE and G-FAST scales may be useful triage tools to identify patients needing emergent stroke evaluation for consideration of IVT.

Introduction: Telestroke enables timely and remote evaluation of patients with acute stroke syndromes. However, stroke mimics represent more than 30% of this population. Given the resources required for management of suspected acute ischemic stroke, several scales have been developed to help identify stroke mimics. Our objective was to externally validate four mimic scales (Khan Score [KS], TeleStroke Mimic Score [TS], simplified FABS [sFABS], and FABS) in a large, academic telestroke network. Methods: This is a retrospective, IRB exempt study of all patients who presented with suspected acute stroke syndromes and underwent video evaluation between 2019-2020 at a large academic telestroke network. Detailed chart review was conducted to extract both the variables needed to apply the mimic scales the final diagnosis confirmed by final imaging and discharge diagnosis (cerebral ischemic vs stroke mimic). Overall score performance was assessed by calculating area under curve (AUC). Youden cutpoint was established for each scale and used to calculate sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy. Results: A total of 1043 patients were included in the final analysis. Final diagnosis of cerebral ischemia was made in 662/1043 patients (63.5%) and stroke mimic was diagnosed in 381/1043 patients (36.5%). To predict stroke mimic, TS had the highest AUC (68.3) and PPV (73%); KS had the highest accuracy (68.6%), sensitivity (86.6%) and NPV (61.5%); and FABS had the highest specificity (54.8%) (Table 1). Conclusions: While each scale offers unique strengths, none was able to identify stroke mimics effectively enough to confidently apply in clinical practice. There remains need for significant clinical judgment to determine the likelihood of stroke mimic at presentation.

Introduction: Prehospital telestroke is increasingly utilized in mobile stroke units and telemedicine-enabled ambulances for acute ischemic stroke (AIS). While the NIHSS remains the standard for telestroke assessment, simpler scales have been validated to help paramedics recognize large vessel occlusion (LVO) strokes but have not been evaluated in a traditional telestroke network. We aimed to determine the accuracy of common LVO scales in a large academic telestroke practice. Methods: This retrospective study included all telestroke consults performed in a large academic telestroke network from 2019 to 2020. Patients were excluded if no NIHSS or vessel imaging was performed. We recorded presenting NIHSS, site of vessel occlusion, and discharge diagnosis. LVO was defined as an acute occlusion of the internal carotid artery and/or M1 middle cerebral artery. The NIHSS was used to calculate 7 LVO scales (RACE, C-STAT, FAST-ED, 3I-SS, PASS, VAN, and G-FAST). Diagnostic performance was assessed by calculating sensitivity, specificity, negative predictive value (NPV), positive predictive value (PPV), and accuracy using the established thresholds of each scale. These results were compared to NIHSS at thresholds of 6, 8, and 10. Area under curve (AUC) was calculated using c-statistics by treating scales as continuous variables. Results: A total of 625 patients were included, 373/625 (59.7%) of which were diagnosed with AIS at discharge. LVO was identified in 78/625 (12.5%) patients. VAN was the most sensitive prehospital stroke scale (83.3%), whereas 3I-SS ≥ 4 was the most specific (95.2%). Both the RACE and FAST-ED scales demonstrated superior accuracy and AUC compared to the NIHSS (Table 1). Conclusions: To our knowledge, this is the first study assessing LVO scale accuracy in a large telestroke network. As use of prehospital telestroke grows, this study highlights several scales that may allow for faster prehospital triage decisions without sacrificing accuracy.

Introduction: Telestroke facilitates emergent assessment of patients with suspected acute ischemic stroke (AIS) to identify candidates for acute interventions. The diagnostic accuracy of the telestroke consultation of a large academic hub and spoke telemedicine network has yet to be investigated. The goal of our study is to determine the diagnostic accuracy of the initial telestroke consultation. Methods: This is an IRB-exempt retrospective study including all patients evaluated for cerebral ischemia via video telestroke consultation in a large academic hub and spoke telemedicine network from 2019 to 2020. Detailed chart review was conducted to identify initial suspected diagnosis and final diagnosis. Cerebral ischemia was defined as AIS and transient ischemic attack (TIA). All other diagnoses were defined as stroke mimics. Data was organized into continuous and categorical variables. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, area under the curve (AUC) and likelihood ratio (LR+) for the initial telestroke consultation diagnosis were calculated while using the final diagnosis as the gold standard diagnosis. Results: A total of 1043 patients met inclusion criteria. Final diagnosis of cerebral ischemia was made in 63.5% of all patients (539/1043 with AIS,123/1043 with TIA). Stroke mimic was diagnosed in 36.5% patients (381/1043). The sensitivity and specificity of telestroke evaluation for diagnosis of cerebral ischemia were 97.1% and 81.4%, respectively. PPV was 90.1%, and NPV was 94.2%. Overall diagnostic accuracy was 91.4% with a LR+ of 5.21 and AUC of 0.89. The most common stroke mimics were metabolic encephalopathy (13.7%, 52/381), migraine (10.8%, 41/381) and seizure (10.5%, 40/381). Conclusions: This study highlights the high diagnostic accuracy of telestroke providers to diagnose cerebral ischemia and stroke mimic in a large academic hub and spoke network. Further research is needed to investigate methods to improve our diagnostic accuracy further, as well as explore the application of tele-neurology in the non-stroke setting.