The Los Angeles Motor Scale as a predictor of angiographically determined large vessel occlusion.

Recent advances in time-sensitive mechanical thrombectomy for the treatment of emergent large vessel occlusion (ELVO) have changed the role of prehospital providers from simply identifying a stroke to identifying the likely presence of ELVO. No one method for identifying ELVO in the field has been demonstrated to be superior. We sought to describe how this might be best accomplished using the Los Angeles Motor Scale (LAMS) in concert with other physical exam findings by paramedics and emergency medical technicians (EMTs). We had paramedics and EMTs examine patients with suspected stroke in the hospital. We compared their exams to the standard neurologist exams and to the results of angiography. We performed multiple analyses to identify the exam elements that would best identify large vessel occlusions. Using LAMS with a threshold score of 4, sensitivity for stroke and ELVO, respectively, was 27% (95% CI 20-36%) and 42% (95% CI 30-55%). When a LAMS of 3 was used in concert with speech abnormality, sensitivity improved to 36% (95% CI 28-45%) and 61% (95% CI 48-73%). Specificity of this model was 70%, (95% CI 64-75%). Most striking was the negative predictive value of this model for ELVO: 90% (95% CI 85-93%) The LAMS or LAMS plus speech can be used to decrease the number of missed large vessel occlusions and to route suspected large vessel occlusions to thrombectomy-capable centers. Other, more complicated scales may have little additional benefit. This derivation data set is the first to use paramedics and EMTs as examiners prospectively and supports prehospital protocol change underway in New York City.

Patient characteristics affecting stroke identification by emergency medical service providers in Brooklyn, New York.

Early identification of stroke should begin in the prehospital phase because the benefits of thrombolysis and clot extraction are time dependent. This study aims to identify patient characteristics that affect prehospital identification of stroke by Long Island college hospital (LICH) emergency medical services (EMS). All suspected strokes brought to LICH by LICH ambulances from January 1, 2010 to December 31, 2011 were included in the study. We compared prehospital care report-based diagnosis against the get with the guidelines (GWTG) database. Age-adjusted logistic regression models were used to study that the effect of individual patient characteristics have on EMS providers' diagnosis. Included in the study were 10,384 patients with mean age 43.9 years. Of whom, 75 had a GWTG cerebrovascular diagnosis: 53 were ischemic strokes, 7 transient ischemic attacks, 3 subarachnoid hemorrhage, and 12 intercerebral bleeds. LICH EMS correctly identified 44 of 75 GWTG strokes. Fifty-one patients were overcalled as stroke by the EMS. Overall EMS sensitivity was 58.7 % and specificity was 99.5 %. Dispatcher call type of altered mental status, stroke, unconsciousness, and increasing prehospital blood pressure quartile were found to be significantly predictive of a true stroke diagnosis. Patients with a past medical history and EMS providers' impression of seizures were more likely to be overcalled as a stroke in the field. More than a third of actual stroke patients were missed in the field in our study. Our results show that the patients' past medical history, dispatcher collected information and prehospital vital sign measurements are associated with a true diagnosis of stroke.