Community wide real-time stroke code capture to better identify changes in health care utilization during COVID-19

Introduction: Many healthcare systems reported a decline in stroke admissions in the early months of the COVID-19 pandemic We used real-time hospital admission data from Stroke Receiving Centers (SRCs) across San Diego County to quantify changes in stroke patients accessing healthcare with the onset of the COVID-19 pandemic Rather than waiting for months-delayed discharge data, real-time stroke code data was used to understand the impact on healthcare utilization which may better inform mitigation strategies to encourage accessing care for acute stroke Methods: We analyzed the total number of patients presenting to any of the 18 San Diego County SRCs for which a stroke code was activated between January 1, 2019 and July 31, 2020;and separated the times into: pre-pandemic (PP) as January 2019 thru February 2020, early-pandemic (EP) as March and April 2020, and mid-pandemic (MP) as May-July 2020 Patients arriving via emergency medical services or private transport were included A public messaging campaign regarding the safety of accessing care for acute stroke started in early May 2020 Results: A total of 14,028 stroke codes were initiated between January 2019 and July 2020 An average of 43 2 stroke codes were activated per stroke center per month (range=39 6 to 46 7 activations per stroke center per month) during PP, 30 6 during EP and 37 7 during MP (p= 019) Overall, 30% fewer stroke code activations occurred during EP compared to the same months in the PP (p= 012) Mid-pandemic, there were 14 6% fewer stroke code activations compared to the same months pre-pandemic (p= 095) Conclusion: Stroke code activations decreased by 30% across San Diego County SRCs in the EPperiod compared to the previous year It is unclear if this is primarily due to decreased healthcareutilization at the start of the COVID-19 pandemic or if there were changes in stroke incidence MPshowed stroke code activations increased compared to EP This may be partially due to the publicmessaging campaign initiated after an analysis of PP to EP stroke code activations We willcontinue to analyze stroke code data to better understand the impact of public messagingcampaigns and determine when activations have returned to PP levels