Modeling and spatial assessment of a bystander notification system for out of hospital cardiac arrests in Pittsburgh, PA

Background: Out-of-hospital cardiac arrest (OHCA) is a substantial public health burden. Rapid cardiopulmonary resuscitation (CPR) or defibrillation with an automated external defibrillator (AED) can increase OHCA survival. Bystander notification systems (BNS) aim to facilitate this. PulsePoint, is a BNS that directs trained responders to OHCAs to initiate CPR. This project was conducted to assess the viability of PulsePoint if it were deployed in Pittsburgh, PA.\ud \ud Methods: Simulations incorporating actual AEDs, OHCAs, and population characteristics were run to assess bystander deployment strategies. Weighting factors to calculate number of bystanders in each census tract were: fixed random points, population, population density, shape area, cardiac arrest incidence, cardiovascular disease mortality (CVD) and age over 65. Maps were created for each factor calculating total distance and time for a layperson to travel to an AED and then an OHCA. Repeated simulations produced statistical summaries for distance and time. Using the estimator with the lowest distance, spatial analysis was conducted to determine the difference in response time between deployed bystanders and emergency medical services, and then estimate then change in one-month survival outcomes using PulsePoint. \ud \ud Results: CVD and Fixed Random Points produced the lowest total distance. Among 92% of the census tracts, bystanders arrived prior to EMS. The survival outcome model predicts that 17.2% (Median = 17.1%, Standard Error = 3.7%) of victims will survive when assessed by layperson and 7.9% (Median = 7.9%, Standard Error = 0.179) with EMS. \ud \ud Conclusion: Bystander notification systems with the use of PulsePoint show potential to increase survival after an OHCA. \ud