235EMF Timeliness of Inter-Facility Transfer for Patients With Acute Large Vessel Occlusion Stroke

Abstract

Mechanical thrombectomy (MT) is the preferred treatment for large vessel occlusion stroke (LVO; ie, occlusion of the basilar, middle cerebral, and internal carotid arteries), and neurological outcome improves with earlier treatment. MT requires substantial infrastructure and expertise and is primarily offered at comprehensive stroke centers (CSCs). Patients with LVO who present to facilities without MT require interfacility transfer, yet transfers are associated with delay and diminished neurological outcome. We sought to examine the time spent at transferring EDs to identify opportunities for improvement. This was a retrospective observational study conducted at our CSC with a large catchment area. Included patients were transferred from an outside ED between January 1, 2012 and December 31, 2017, and received MT after arrival. Clinical data and time intervals were collected from the medical record. Time intervals included ED Door-to-CT, CT-to-CSC Alert, CSC Alert-to-ED Departure, Total ED Length of Stay (LOS), and ED Arrival-to-OR Arrival. We estimated differences in medians for each time interval with 95% confidence intervals. Multivariable logistic regression modeling was used to examine the association of time spent in the ED with good 90-day neurologic outcome via modified Rankin Score (mRS; good ≤2, poor >3) while controlling for potential confounders (NIH stroke scale, Age, tPA administration). We identified 111 patients who met inclusion. 54 (49%) of patients were female. Outcomes at 90 days were available for 95 (86%) patients. Among patients with outcomes, 59 (62%) of patients had a poor 90-day neurologic outcome, while 36 (38%) had a good neurologic outcome. Median total ED LOS was 99 (interquartile range [IQR] 71-139) minutes, and median ED Arrival-to-OR Arrival was 239 (IQR 187-304) minutes. The largest proportion of ED time was CT-to-CSC Alert which comprised 44% of ED LOS. Component intervals were similar by neurological outcome except CT-to-CSC Alert (Figure). Patients who had good outcomes experienced a 13-minute shorter CT-to-CSC Alert (35 vs 48 minutes). However, this difference was not significant in multivariable modeling. Only initial NIH Stroke scale was associated with good 90-day neurological outcome (adjusted odds ratio 0.84, 95% CI 0.77, 0.92, p<0.0001). Among patients with LVO transferred for MT, the largest proportion of time at transferring EDs occurred between CT and the CSC Alert suggesting potential diagnostic and coordination delays. While overall time spent at transferring EDs for patients with LVO was not associated with neurologic outcomes, this study only includes patients who received MT. Future investigations should focus on the time between CT and stroke alert as potentially modifiable, the generalizability of results and interventions to other sites, and populations who did not receive MT.