Heart Stroke Diagnosis using AI Model

Introduction: Telestroke allows stroke expertise for thrombolysis decision making remotely using high-quality bidirectional audiovisual technology. Hypothesis: Intravenous tissue plasminogen activator (IVtPA) is administered via telestroke network to a proportion of patients without a stroke diagnosis (i.e. stroke mimic) Methods: Our academic comprehensive stroke program telestroke program includes 26 spoke Emergency rooms (ERs) through which IVtPA is administered throughout central Ohio. From July 1, 2016 to Sept 30, 2017, nearly all patients who received IVtPA at the outside hospital telestroke ERs were transferred to our institution for post-IVtPA care. Data was collected on final diagnosis, demographics, National Institutes of Health Stroke Scale (NIHSS), door to needle (DTN) time, and outcomes. Results: Among 270 acute ischemic stroke patients who received IVtPA via telestroke, we identified 64 (23.7%) with a stroke mimic diagnosis. Stroke mimics were younger (mean age 56.4 vs 68.2, p <0.0001), more likely female (60.9% vs 45.6%, p 0.03), and had higher DTN times (85.3 vs 69.9 minutes, p 0.0008). The increase in DTN was due to longer time to recommend by the telestroke neurologist for stroke mimic (65.0 vs 53.2 minutes, p 0.0034). The stroke mimic diagnosis included Migraine 26 (40.6%), Factitious disorder 12 (18.8%), Encephalopathy 7 (10.9%), and Unmasking 6 (9.4%). The stroke mimics did not differ from each other based upon initial NIHSS, DTN, or sex. Compared to the other stroke mimics, Migraine and Factitious disorder patients were younger (51.2 vs 63.9 years, p <0.0006), more likely to have a personal history of migraines (42.1% vs 0%, p < 0.0001), and more likely to have functional exam findings (42.1% vs 3.8%, p 0.0007). There were no hemorrhagic complications in the stroke mimic patients. Among all stroke mimics, 26 (40.6%) had a history of similar prior episodes and 10 (15.6%) would have future recurrence of another similar episode, with 2 patients receiving IVtPA again in the future (1 Migraine and 1 Factitious disorder). Conclusions: In a tertiary academic telestroke network, nearly one-quarter of patients receive IVtPA for a non-stroke diagnosis, with migraine and factitious disorder being the most commonly seen.

Download Full-text

Abstract T P390: Baseline Stroke Knowledge of 4th-6th Grade Children in New York City Public Schools Has No Appreciable Trend from 2005-2014

Stroke ◽

10.1161/str.46.suppl_1.tp390 ◽

2015 ◽

Vol 46 (suppl_1) ◽

Author(s):

James M Noble ◽

Cailey Simmons ◽

Mindy F Hecht ◽

Olajide Williams

Keyword(s):

New York ◽

New York City ◽

Public Schools ◽

Real World ◽

York City ◽

Hip Hop ◽

Expected Value ◽

Stroke Symptom ◽

Stroke Diagnosis ◽

Stroke Knowledge

Background and Purpose: To examine whether the baseline stroke knowledge of children in schools participating in our Hip Hop Stroke program has changed since its inception in late 2005. Methods: We gathered baseline stroke knowledge surveys from 2,839 students enrolled in the Hip Hop Stroke program from November 2005 through April 2014 with median annual enrollment of 344 (range 55 to 582). All students were enrolled in New York City public schools, in 4th through 6th grade. Students who left ≥3 questions blank were discarded; other blank answers were treated as missing. Data were analyzed using binomial, Chi-Square and regression analysis (SPSS v22.0). Results: Overall there was no consistent trend in baseline stroke knowledge by academic year. Overall, 28.4% of students recognized stroke occurred in the brain (expected value 25% [p<0.001], range from 13.8-61.2% for any given year). With stroke diagnosis provided, 85.5% of 1436 students knew to call 911, whereas only 59.6% of 1243 students knew to call 911 when given a hypothetical real-world stroke symptom scenario without stroke diagnosis included, p<0.001. For a composite assessment of knowledge including 4 stroke symptoms (blurred vision, facial droop, sudden headache, slurred speech), 1 distractor (chest pain), and urgent action plan (call 911), asked consistently since 2006, overall students scored a mean 2.86 (95% CI: 2.80-2.92; possible range 0-6, expected value 2.75), with annual scores ranging from 2.54-3.56. Conclusion: Stroke knowledge among elementary school students remains low and has not appreciably changed during the last 9 years. The use of hypothetical real-world stroke symptom scenarios may more accurately reflect intent to call 911 for stroke than the use of questions in which stroke diagnosis is given.

Download Full-text