Patient and Process Factors Associated With Type of First Neuroimaging and Delayed Diagnosis in Childhood Arterial Ischemic Stroke

Abstract

Objectives: In-hospital factors contribute more to delayed diagnosis of childhood arterial ischemic stroke (AIS) than prehospital factors. We aimed to explore process and patient factors associated with type of and timing to neuroimaging in childhood AIS in the emergency department (ED). Methods: This was a retrospective hospital registry–based study of children with AIS, presenting to an Australian tertiary pediatric ED between January 2003 and December 2012. Neuroimaging data and timelines of care were also collected from referring hospitals for transferred patients. Results: Seventy-one AIS episodes and 19 transient ischemic attacks were recorded. The majority (56%) were initially seen at a referring hospital. Patients underwent computed tomography (CT) as first scan more frequently than magnetic resonance imaging (MRI) as first scan (61% vs. 32%) at both the referring and the tertiary hospitals. Time to first scan as CT was significantly shorter compared with MRI (median = 1.5 hours vs. 10.9 hours, p < 0.001). MRI was performed more often at the tertiary hospital (92.5% vs. 26%, p = 0.001). Median time to performance of diagnostic MRI was 15.1 hours (interquartile range = 7.1–23.5), with no significant difference between patients first presenting to a referring hospital and those directly accessing the tertiary center. Patient characteristics including age, past medical history, conscious state, focal symptoms, and signs on arrival were not associated with the type of first neuroimaging or time to diagnostic MRI. Patients presenting during weekends were less likely to receive an MRI as first scan (odds ratio [OR] = 0.3, 95% confidence interval [CI] = 0.1–0.8), while time to MRI was significantly longer for children presenting after hours (5 pm–8 am; median = 17.6 hours vs. 8.4 hours, p = 0.026). MRI overall and as first scan was associated with a higher use of sedation than CT (OR = 6.5, 95% CI = 1.3–32.9; and OR = 3.9, 95% CI = 1.3–11.8), particularly for children younger than 5 years of age (OR = 12.5, 95% CI = 3–52.4). Conclusions: Strategies to improve rapid diagnosis of pediatric stroke should include shared regional hospital networks protocols to optimize local imaging strategies and where possible rapid transfer to the tertiary center. Future priorities should include development of pediatric ED physician decision support tools to differentiate stroke from mimics and the development and implementation of rapid ED imaging stroke protocols to improve access to confirmatory MRI scanning.