Improving Out-of-hospital Cardiac Arrest Outcomes and Automated External Defibrillator Placement Guidelines through Optimization and Predictive Modeling

Out-of-hospital cardiac arrest (OHCA) is an often fatal, time-sensitive emergency event. OHCA survival decreases significantly for every minute without treatment. Defibrillation, an effective means to resuscitate OHCA victims, can be safely delivered by untrained bystanders using an automated external defibrillator (AED). The dissemination of AEDs in public areas for bystander use has been shown to reduce the time to defibrillation and triple OHCA survival. However, bystander AED use and OHCA survival remain low despite the increasing amount of resources dedicated to AED placement. Identifying the limitations of existing AED network design and exploring new strategies for AED deployment are crucial to ensure efforts and resources spent on AED placements directly translate to improved OHCA outcomes. To address these questions, this thesis focuses on increasing the understanding of spatiotemporal OHCA risk and developing novel tools to improve the assessment and design of AED networks, using optimization and machine learning techniques, with the ultimate goal of increasing bystander AED use and OHCA survival. Specifically, we analyzed AED placement guidelines, existing AED networks, bystander response records, and real OHCA patient data from Toronto, Canada, and Copenhagen, Denmark and identified limited temporal AED accessibility as a key factor responsible for low AED use rates. Motivated by this finding, we developed a spatiotemporal AED placement optimization model that maximizes spatial and temporal AED accessibility. We showed our model reverses the effects of limited temporal AED accessibility in both Toronto and Copenhagen, two vastly different cities with distinct geographies, populations, and EMS systems. This demonstrated the effectiveness and generalizability of our optimization approach. To validate the clinical value of our optimization model, we developed a novel in silico trial framework that computationally evaluates AED networks in a real-life deployment situation. Using this framework, we conducted two in silico trials and established the superiority of optimization approaches to current practices and established placement guidelines, which represent the gold standard of evidence-based AED placement recommendations, in improving estimated OHCA outcomes. In conclusion, our findings identify the importance of temporal AED accessibility, data-driven AED placement optimization models, and centralized decision making in improving AED use and OHCA survival.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.