Abstract Out-of-hospital cardiac arrest (OHCA) affects over 350,000 Americans annually and survival rates are very low. For every 1-minute delay in achieving return of effective heart function after collapse, the chance of survival drops by 10%. Bystanders can aid in the emergency treatment of OHCA victims by performing chest compressions and by using automated external defibrillators (AEDs). However, current bystander use of static AEDs is very low and defibrillation is primarily administered by first responders and emergency medical services (EMS) whose median arrival time (8 minutes) is too late to save most OHCA patients. Using a drone to deliver AEDs to OHCA victims within 3 to 5 minutes of the 911 call is an exciting new concept that is based on current technical capabilities of drones. Early work with simulation models has demonstrated the potential of a strategically designed drone network to deliver an AED to an OHCA substantially more rapidly than EMS can achieve. However, these early simulations assumed complete effectiveness of AED use when delivered to an OHCA scene without considering the bystander variables. It is well-known that bystanders may hesitate to perform CPR and to apply an AED, and that select demographic and neighborhood factors (age, sex, race/gender, education) may be predictive of such treatment variability. The time it takes a bystander to extract an AED and apply it successfully in OHCA may critically impact overall survival gains from timely drone AED delivery. An accurate understanding of potential treatment effectiveness should account for expected bystander performance. The overarching aim of this application is to utilize data science and simulation research to estimate end-user performance and treatment-effectiveness of a drone network accounting for community, first responder, and EMS performance. Aim 1 will determine the optimal placement of drone stations to ensure timely AED arrival in high-OHCA risk geographic areas (within 3 to 5 minutes) across North Carolina. Aim 2 will define and determine the association of community phenotypic clusters on OHCA treatment patterns in high-incidence NC communities. Aim 3 will use simulated drone AED OHCA scenarios to define drone-AED-bystander treatment intervals among community phenotypic clusters (e.g., minority, rural, low education, elderly) in high-OHCA risk NC neighborhoods. Results from Aims 2 and 3 will be used to refine our optimization model (Aim 1) to estimate treatment effectiveness and efficiency. The proposed work will be carried out under the direct supervision of Dr. Starks mentorship team: mentor (Dr. Daniel Mark), co-mentor (Dr. Christopher Granger), and her advisory team (Drs. Billy Williams and Graham Nichol). This K23 application with the support and guidance of her mentorship team and advisory committee will position Dr. Starks to eventually lead independent NIH funded studies focused on community treatment of OHCA, including developing/testing i...