Prehospital scale to differentiate intracerebral hemorrhage from large-vessel occlusion patients: a prospective cohort study

Evaluating scales to detect large vessel occlusion (LVO) could aid in considering early referrals to a thrombectomy-capable center in the prehospital stroke code setting. Nevertheless, they entail a significant number of false positives, corresponding to intracranial hemorrhages (ICH). Our study aims to identify easily collectible variables for the development of a scale to differentiate patients with ICH from LVO. We conducted a prospective cohort study of stroke code patients between May 2021 and January 2023. Patients were evaluated with CT/CT-Angiography at arrival. We compared clinical variables and vascular risk factors between ICH and LVO patients. Out of 989 stroke code patients, we included 190 (66.7%) LVO cases and 95 (33.3) ICH cases. In the multivariate analysis, headache (odds ratio [OR] 3.56; 1.50-8.43), GCS < 8 (OR 8.19; 3.17-21.13), SBP > 160mmHg (OR 6.43; 3.37-12.26) and male sex (OR 2.07; 1.13-3.80) were associated with ICH, while previous hypercholesterolemia (OR 0.35; 0.19-0.65) with LVO. The scale design was conducted, assigning a score to each significant variable based on its specific weight: +2 points for SBP > 160, + 1 points for headache, + 1 points for male sex, + 2 points for GCS < 8, and - 1 points for HCL. The area under the curve was 0.82 (0.77-0.87). A score ≥ 4 exhibited a sensitivity of 0.10, a specificity of 0.99, a positive predictive value of 0.21, and a negative predictive value of 0.98. We present the development of a prehospital scale to discriminate between ICH and LVO patients, utilizing easily detectable variables in the prehospital setting.