Abstract WP293: Simplified Prehospital Prediction Models to Estimate the Likelihood of 4 Types of Stroke: 7-Item Japan Urgent Stroke Triage (Just-7) Score

Abstract

Background and Purpose: Prehospital prediction models to estimate the likelihood of several types of stroke (large vessel occlusion [LVO], intracranial hemorrhage [ICH], and subarachnoid hemorrhage [SAH], and other strokes) should be useful to transfer those with suspected stroke to appropriate facilities. Japan Urgent Stroke Triage score with 21 items had excellent predictive abilities, but we tried to simplify the score with parsimonious items and comparable predictive abilities. Methods: We conducted historical and prospective multicenter cohort studies from June 2015 to July 2017. We developed the prediction rule with select variables from JUST score for LVO, ICH, SAH and other strokes using historical cohort study with 2236 patients and validated the developed score using prospective cohort study with 964 patients. We used multivariable logistic regression models to develop the prediction models using the same variables for each stroke type. Result: In the historical cohort, there were 1150 stroke, including 235 LVO, 352 ICH, 107 SAH and 456 other stroke. We developed the score with 7 items (high blood pressure, arrhythmia, conjugate deviation, headache, dysarthria, disturbance of consciousness, paralysis of upper limbs) which showed area under the receiver operating curve (AUC) of 0.84 for any type of stroke, 0.89 for LVO, 0.79 for ICH, and 0.90 for SAH in the historical cohort. The score was validated with good predictive ability in the prospective cohort (AUC of 0.76 for any type of stroke; 0.81 for LVO, 0.73 for ICH, and 0.85 for SAH). Conclusions: The simplified 7-item JUST (JUST-7) score had good predictive ability and can help paramedics or primary care physicians to estimate the likelihood of different type of stroke and decide the referral hospital.