Journal Watch: The VAN Scale and LVO Strokes

Reviewed This Month

Paramedic Utilization of Vision, Aphasia, Neglect (VAN) Stroke Severity Scale in the Prehospital Setting Predicts Emergent Large Vessel Occlusion Stroke

Authors: Birnbaum L, Wampler D, Shadman A, et al. 

Published in: J NeuroIntervent Surg, 2021; 13: 505–8

Mechanical thrombectomy is an incredibly effective therapy for large vessel occlusion (LVO) strokes. However, it is only indicated within 24 hours of the patient’s last known well time. Studies have shown that each hour mechanical thrombectomy is delayed is associated with declines in patient outcome. Identifying LVO strokes in the prehospital environment is an important step in saving ischemic brain tissue that is not already infarcted. While many stroke scales have been developed, most have not been fully studied, including prospective and retrospective evaluations, when performed by EMS. 

The study we review in this month’s Journal Watch was a prospective prehospital cohort study. This study was designed to validate the use and efficacy of the VAN (vision, aphasia, neglect) stroke assessment. The authors indicated they selected VAN to study in their stroke system of care because of its “simplicity to teach and perform.” Unlike other stroke scales, VAN does not require the calculation of a score. Instead, for VAN to be positive, a patient must present with weakness and have one or more of the following: a forced gaze or lost vision on one side; inability to understand or produce language; or ignoring one side of the body (typically the left) when both sides are touched at the same time. 

While VAN has been validated in EMS and shown some promise in identifying LVO strokes, this study was the first to prospectively evaluate it in the prehospital environment. 

Background

Prior to the start of the study, paramedics were required to complete VAN training. Training included in-person learning, interactive live video, and access to recorded material. Following each training session, paramedics were required to triage three fictional LVO cases and complete a 10-question quiz. 

The study was conducted from June 1, 2017 to December 31, 2019. During this time, all San Antonio Fire Department and Acadian Ambulance Service paramedics were required to perform a VAN assessment whenever a stroke alert was initiated and the patient was transported to one of three comprehensive regional stroke centers. Stroke alert criteria included onset of symptoms within six hours, any abnormal findings on the Cincinnati Prehospital Stroke Scale (CPSS), and blood glucose within 60–600 mg/dL. Because existing stroke alert criteria required onset of symptoms to be within six hours, patients with a last known well time greater than that were excluded from the analysis. 

The investigators obtained stroke alert criteria as well as results from the VAN and CPSS from prehospital electronic medical records. Outcome data was obtained from the comprehensive stroke centers and included the initial National Institutes of Health Stroke Scale score, advanced neuroimaging to determine LVO status, mechanical thrombectomy performance, and final diagnosis. 

The analysis focused on the performance of the VAN compared to an NIHSS score of 6 or more for the presence of LVO stroke alone and a combined outcome of LVO and any intracranial hemorrhage (ICH). The investigators calculated a positive predictive value (PPV), sensitivity, negative predictive value (NPV), specificity, and overall accuracy. As a brief reminder, PPV is the percentage of patients with a positive test who actually have the disease; sensitivity is the ability of a test to correctly classify an individual as having the disease; NPV is the percentage of patients with a negative test who do not have the disease; and specificity is the ability of a test to correctly classify an individual not having the disease. These are all common metrics used in evaluating diagnostic medical tests. 

Results

During the study period there were 386 stroke alert patients, and 290 (75%) had complete data. Only those patients with complete data were included in the analysis. The average age was 59 years. There were 193 (67%) with a positive VAN and 168 (58%) with an NIHSS of 6 or more. There were 76 patients with a final diagnosis indicating non-LVO stroke, 68 with LVO stroke, 37 with intracranial hemorrhage, 28 with transient ischemic attacks, 23 seizures, 18 altered mental status, nine migraines, nine conversions, and 22 final diagnoses classified as other. There were 29 (43%) LVO strokes that had mechanical thrombectomy performed. 

VAN and NIHSS of 6 or more showed a statistically significant association with LVO stroke alone as well as LVO and ICH combined. For LVO stroke, VAN had a PPV of 29%, a sensitivity of 81%, an NPV of 87%, a specificity of 38%, and an accuracy of 48%. NIHSS of 6 or more had a PPV of 35%, a sensitivity of 85%, an NPV of 92%, a specificity of 51%, and an accuracy of 59%. For LVO or any ICH, VAN had a PPV of 45%, a sensitivity of 83%, an NPV of 81%, a specificity of 43%, and an accuracy of 57%. NIHSS of 6 or more had a PPV of 54%, a sensitivity of 86%, an NPV of 88%, a specificity of 58%, and an accuracy of 68%. 

This was an interesting study that adds to the published evidence that prehospital EMS assessments can be effectively taught and implemented in stroke systems with multiple EMS agencies and comprehensive stroke centers. Limitations for this study included exclusion of records that did not have complete data, exclusion of patients with a last known well time greater than six hours, and an average age of 59. While the specificity of VAN to identify for LVO alone was suboptimal, its specificity and sensitivity improved when evaluating the combined outcome of LVO and ICH. The results of this study support the use of VAN to identify ICH and as an effective prehospital ICH bypass tool. Nevertheless, the investigators indicated VAN is favored in their system for its high sensitivity and simplicity to learn and teach.   

Antonio R. Fernandez, PhD, NRP, FAHA, is a research scientist at ESO and serves on the board of advisors of the Prehospital Care Research Forum at UCLA.