From the Melbourne Brain Centre and Department of Neurology, Royal Melbourne Hospital, Australia (H.Z., L.P., S.C., E.R., P.S., N.Y., S.M.D., B.C.V.C.); Ambulance Victoria, Melbourne, Australia (K.S., S.B., M.S.); The Florey Institute of Neuroscience and Mental Health, University of Melbourne, Australia (L.C., N.Y.); Department of Epidemiology and Preventive Medicine, and Department of Community Emergency Health and Paramedic Practice, Monash University, Australia (K.S., M.S.); and Discipline of Emergency Medicine, University of Western Australia, Australia (K.S., S.B.).
Stroke. 2018 Apr;49(4):945-951. doi: 10.1161/STROKEAHA.117.019307. Epub 2018 Mar 14.
Clinical triage scales for prehospital recognition of large vessel occlusion (LVO) are limited by low specificity when applied by paramedics. We created the 3-step ambulance clinical triage for acute stroke treatment (ACT-FAST) as the first algorithmic LVO identification tool, designed to improve specificity by recognizing only severe clinical syndromes and optimizing paramedic usability and reliability.
The ACT-FAST algorithm consists of (1) unilateral arm drift to stretcher <10 seconds, (2) severe language deficit (if right arm is weak) or gaze deviation/hemineglect assessed by simple shoulder tap test (if left arm is weak), and (3) eligibility and stroke mimic screen. ACT-FAST examination steps were retrospectively validated, and then prospectively validated by paramedics transporting culturally and linguistically diverse patients with suspected stroke in the emergency department, for the identification of internal carotid or proximal middle cerebral artery occlusion. The diagnostic performance of the full ACT-FAST algorithm was then validated for patients accepted for thrombectomy.
In retrospective (n=565) and prospective paramedic (n=104) validation, ACT-FAST displayed higher overall accuracy and specificity, when compared with existing LVO triage scales. Agreement of ACT-FAST between paramedics and doctors was excellent (κ=0.91; 95% confidence interval, 0.79-1.0). The full ACT-FAST algorithm (n=60) assessed by paramedics showed high overall accuracy (91.7%), sensitivity (85.7%), specificity (93.5%), and positive predictive value (80%) for recognition of endovascular-eligible LVO.
The 3-step ACT-FAST algorithm shows higher specificity and reliability than existing scales for clinical LVO recognition, despite requiring just 2 examination steps. The inclusion of an eligibility step allowed recognition of endovascular-eligible patients with high accuracy. Using a sequential algorithmic approach eliminates scoring confusion and reduces assessment time. Future studies will test whether field application of ACT-FAST by paramedics to bypass suspected patients with LVO directly to endovascular-capable centers can reduce delays to endovascular thrombectomy.
院前识别大血管闭塞（LVO）的临床分诊量表，在急救人员应用时特异性较低。我们创建了 3 步式救护车急性卒中治疗临床分诊（ACT-FAST），作为首个算法性 LVO 识别工具，旨在通过仅识别严重的临床综合征，以及优化急救人员的可用性和可靠性，来提高特异性。
ACT-FAST 算法由以下步骤组成：（1）单侧手臂向担架漂移<10 秒；（2）严重的语言障碍（如果右臂无力）或通过简单的肩部敲击试验评估的凝视偏差/偏盲（如果左臂无力）；（3）符合条件和卒中模拟筛查。回顾性验证了 ACT-FAST 检查步骤，然后由在急诊转运疑似卒中的文化和语言多样化患者的急救人员前瞻性验证，以识别颈内或近端大脑中动脉闭塞。然后，对接受取栓治疗的患者进行完整的 ACT-FAST 算法的诊断性能验证。
在回顾性（n=565）和前瞻性急救人员（n=104）验证中，与现有的 LVO 分诊量表相比，ACT-FAST 显示出更高的整体准确性和特异性。急救人员和医生之间的 ACT-FAST 一致性非常好（κ=0.91；95%置信区间，0.79-1.0）。由急救人员评估的完整的 ACT-FAST 算法（n=60）对可血管内治疗的 LVO 的识别具有较高的整体准确性（91.7%）、敏感性（85.7%）、特异性（93.5%）和阳性预测值（80%）。
3 步式 ACT-FAST 算法在识别临床 LVO 方面的特异性和可靠性均高于现有的量表，尽管仅需要进行 2 项检查步骤。纳入符合条件的步骤可以使识别有血管内治疗条件的患者具有较高的准确性。使用序贯算法方法可消除评分混淆并减少评估时间。未来的研究将检验是否可以通过急救人员将疑似 LVO 患者直接分诊至有血管内治疗能力的中心，从而减少血管内取栓的延误。
