Improving Prehospital Stroke Diagnosis Using Natural Language Processing of Paramedic Reports.

Background and Purpose: Accurate prehospital diagnosis of stroke by emergency medical services (EMS) can increase treatments rates, mitigate disability, and reduce stroke deaths. We aimed to develop a model that utilizes natural language processing of EMS reports and machine learning to improve prehospital stroke identification. Methods: We conducted a retrospective study of patients transported by the Chicago EMS to 17 regional primary and comprehensive stroke centers. Patients who were suspected of stroke by the EMS or had hospital-diagnosed stroke were included in our cohort. Text within EMS reports were converted to unigram features, which were given as input to a support-vector machine classifier that was trained on 70% of the cohort and tested on the remaining 30%. Outcomes included final diagnosis of stroke versus nonstroke, large vessel occlusion, severe stroke (National Institutes of Health Stroke Scale score >5), and comprehensive stroke center-eligible stroke (large vessel occlusion or hemorrhagic stroke). Results: Of 965 patients, 580 (60%) had confirmed acute stroke. In a test set of 289 patients, the text-based model predicted stroke nominally better than models based on the Cincinnati Prehospital Stroke Scale (c-statistic: 0.73 versus 0.67, P=0.165) and was superior to the 3-Item Stroke Scale (c-statistic: 0.73 versus 0.53, P<0.001) scores. Improvements in discrimination were also observed for the other outcomes. Conclusions: We derived a model that utilizes clinical text from paramedic reports to identify stroke. Our results require validation but have the potential of improving prehospital routing protocols.

Door-In-Door-Out Process Times at Primary Stroke Centers in Chicago.

STUDY OBJECTIVE: Acute stroke patients often require interfacility transfer from primary stroke centers to comprehensive stroke centers. Given the time-sensitive benefits of endovascular reperfusion, reducing door-in-door-out time at the primary stroke center is a target for quality improvement. We sought to identify modifiable predictors of door-in-door-out times at 3 Chicago-region primary stroke centers. METHODS: We performed a retrospective analysis of consecutive patients with acute stroke from February 1, 2018 to January 31, 2020 who required transfer from 1 of 3 primary stroke centers to 1 of 3 affiliated comprehensive stroke centers in the Chicago region. Stroke coordinators at each primary stroke center abstracted data on type of transport, medical interventions and treatments prior to transfer, and relevant time intervals from patient arrival to departure. We evaluated predictors of door-in-door-out time using median regression models. RESULTS: Of 191 total patients, 67.9% arrived by emergency medical services and 57.4% during off-hours. Telestroke was performed in 84.2%, 30.5% received alteplase, and 48.4% underwent a computed tomography (CT) angiography at the primary stroke center. The median door-in-door-out time was 148.5 (interquartile range 106 to 207.8) minutes. The largest contributors to door-in-door-out time, in minutes, were CT to CT angiography time (22 [7 to 73.5]), transfer center contact to ambulance request time (20 [8 to 53.3]), ambulance request to arrival time (20.5 [14 to 36]), and transfer ambulance time at primary stroke center (26 [21 to 35]). Factors associated with door-in-door-out time were (adjusted median differences, in minutes [95% confidence intervals]): CT angiography performed at primary stroke center (+39 [12.3 to 65.7]), walk-in arrival mode (+53 [4.1 to 101.9]), administration of intravenous alteplase (-29 [-31.3 to -26.7]), intubation at primary stroke center (+23 [7.3 to 38.7]), and ambulance request by primary stroke center (-20 [-34.3 to -5.7]). CONCLUSION: Door-in-door-out times at Chicago-area primary stroke centers average nearly 150 minutes. Reducing time to CT angiography, receipt of alteplase, and ambulance request are likely important modifiable targets for interventions to decrease door-in-door-out times at primary stroke centers.

Postural Balance Performance of Children with ADHD, with and without Medication: A Quantitative Approach.

tAtention Deficit Hyperactivity Disorder (ADHD) significantly impairs the performance of students during their early school years. However, the extent of the difference in postural stability between children with ADHD and Typically Developed (TD) children, are still not fully understood. Also, it is clinically important to investigate possible effects of medication on their performances. We recruited 38 children between the ages 6 to 12 with ADHD (15 medically naïve, 23 with current treatment history) and 25 age-matched as TD. Postural stability performance was assessed in four conditions: 1-eyes open, 2-eyes closed, 3-eyes open on foam and 4-eyes closed on foam. A cost-effective mobile force platform was used to evaluate postural performance quantitatively. Results revealed a lower stability performance in the medically-naïve participants especially in trials with eyes open on foam. No significant difference in performance was seen between children with medical treatment history and the TD group. Results suggest that using medication can effectively enhance postural stability for children with ADHD. We found portable balance assessment tools helpful in identifying changes in motor performances among children with ADHD. Our results indicate that further research is needed to understand the exact implications of ADHD on postural stability under different sensory stimuli.