Abstract TP209: Provider-Activated Stroke Code Process Results in Higher Yield and Shorter Treatment Times

Background: Many stroke centers empower nurses to activate acute stroke protocols to expedite stroke care. Alternative models rely on provider assessment prior to activation but risk delayed treatment. We sought to compare yield and treatment times of provider vs. nurse activated acute stroke protocols. Methods: We performed a retrospective cross-sectional analysis of de-identified stroke code activations from 2023-2024 at two medical centers in an affiliated system of care. One thrombectomy-capable center utilizes a provider-activated stroke code protocol (provider cohort) and one comprehensive center a nurse-activated (nurse cohort). Stroke code activations data is collected as part of routine quality efforts. Both hospitals participate in Get With the Guidelines™-Stroke. The same group of residents serves as “stroke code responders” at the two hospitals. We assessed yield of each protocol, defined as the percentage of stroke code activations per total number of stroke discharges. We calculated median times of stroke center metrics for each cohort as door to: MD/provider, stroke team, CT, needle, puncture, and device and performed unpaired t-tests to assess differences between cohorts. Results: There were 1,623 stroke code activations per 808 discharges in the provider cohort compared to 5,750 activations per 1,250 discharges in the nurse cohort. Stroke code yield was 49.8% for the provider cohort vs. 21.7% for the nurse cohort. There were 64 thrombolysis cases and 97 thrombectomy cases in the provider cohort and respectively 40 and 140 in the nurse cohort. Median door to needle time was 33 min in the provider cohort vs. 37 in the nurse cohort (p=0.0083). Process metrics were significantly better in the provider cohort, with a door to MD of 0 vs. 6 min (p=0.0000), door to stroke of 2 vs. 7 min (p=0.0000) and door to CT time of 18 vs. 20 min (p=0.0000). Door to puncture was better in the nurse cohort (92 vs. 98 min, p=0.0000) but there was no difference in door to device (116 nurse vs. 117.5 min provider, p= 0.8802). Conclusions: In this analysis of provider vs. nurse activated stroke protocols, yield and treatment times were significantly better in the provider cohort. Additional assessment prior to stroke code activation may save limited hospital resources, such as radiology and “neurology stroke responders.” Additionally, it may minimize delays with alert fatigue and activation of stroke codes for unstable patients, while providing more autonomy to ED teams.