Cincinnati Prehospital Stroke Scale Can Identify Large Vessel Occlusion Stroke

Richards, C. T.; Huebinger, R.; Tataris, K. L.; Weber, J. M.; Eggers, L.; Markul, E.; Stein-Spencer, L.; Pearlman, K. S.; Holl, J. L.; Prabhakaran, S.

Prehosp Emerg Care. 2018 Jan 4; 1-7


OBJECTIVE: Accurate prehospital identification of patients with acute ischemic stroke (AIS) from large vessel occlusion (LVO) facilitates direct transport to hospitals that perform endovascular thrombectomy. We hypothesize that a cut-off score of the Cincinnati Prehospital Stroke Scale (CPSS), a simple assessment tool currently used by emergency medical services (EMS) providers, can be used to identify LVO. METHODS: Consecutively enrolled, confirmed AIS patients arriving via EMS between August 2012 and April 2014 at a high-volume stroke center in a large city with a single municipal EMS provider agency were identified in a prospective, single-center registry. Head and neck vessel imaging confirmed LVO. CPSS scores were abstracted from prehospital EMS records. Spearman's rank correlation, Wilcoxon rank-sum test, and Student's t-test were performed. Cohen's kappa was calculated between CPSS abstractors. The Youden index identified the optimal CPSS cut-off. Multivariate logistic regression controlling for age, sex, and race determined the odds ratio (OR) for LVO. RESULTS: Of 144 eligible patients, 138 (95.8%) had CPSS scores in the EMS record and were included for analysis. The median age was 69 (IQR 58-81) years. Vessel imaging was performed in 97.9% of patients at a median of 5.9 (IQR 3.6-10.2) hours from hospital arrival, and 43.7% had an LVO. Intravenous tissue plasminogen activator was administered to 29 patients, in whom 12 had no LVO on subsequent vessel imaging. The optimal CPSS cut-off predicting LVO was 3, with a Youden index of 0.29, sensitivity of 0.41, and specificity of 0.88. The adjusted OR for LVO with CPSS = 3 was 5.7 (95% CI 2.3-14.1). Among patients with CPSS = 3, 72.7% had an LVO, compared with 34.3% of patients with CPSS

Read More on PubMed