Abstract Body: Introduction:
Inpatient stroke alerts requiring an in-person neurologist to assess and evaluate patients for thrombolytic therapy may contribute to delays in evaluation and treatment. A new inpatient stroke alert process was developed to improve assessment and treatment of stroke. Timeliness of stroke evaluation and treatment was assessed before and after change in processes.

Methods:
A retrospective study on patients with stroke alerts called between January 2023 and July 2024 at a 375-bed regional hospital. Patients without complete data were excluded. A new inhouse stroke alert process was implemented in two phases. Phase one (March 2023) utilized telestroke instead of in-person neurology and phase two (November 2023) refined the process by replacing phone-call paging with secure-text communication. The pre-rollout periods involved interdepartmental and interdisciplinary planning, education and communication about appending the Rapid Response Team process in the facility. Primary outcome examined the time from alert to CT using a quantile (median) regression analysis. Secondary analysis evaluated the proportion of alert to CT within 20-minute target using chi-square analysis. Subgroup analysis evaluated paging time from phone-call to secure-text page using a quantile (median) regression.

Results:
A total of 135 patients were identified and after exclusions, 130 were included. Mean age was 72.75 ± 12.46 years, 65 patients (50%) were female and median NIHSS was 6.5 [interquartile range (IQR) 2 – 15]. Stroke alerts increased from before phase one to after phase one, 5.36±1.49 versus 7.24±2.95, but difference was not statistically significant, p = 0.30. One patient was treated with IV thrombolysis before phase one compared to five after with median alert to needle (ATN) time of 84 versus 81 minutes. One case (20%) after implementation of phase one achieved ATN time of less than 60 minutes. Alert to CT scan decreased by 18 minutes [95%CI 10.98, 25,02] after phase one, p < 0.001. The proportion of stroke alerts achieving CT target times increased after phase one, 13% versus 60%, OR = 9.75 [95% CI, 2.10, 45.24], p=0.001. Paging time decreased by 2 minutes [95%CI 0.86, 3.14] after phase two compared to prior, p=0.001.

Conclusions:
The new in-house process resulted in faster imaging, communication and improved assessment times. Continued process improvement is needed to decrease alert to needle times and further improve patient outcomes.