Abstract 313: Visualizing Out-Of-Hospital Cardiac Arrest in the Young: A Network Analysis of the Cardiac Arrest Registry for Enhanced Survival in Chicago
Introduction: The interactions of various variables on out-of-hospital cardiac arrest (OHCA) in the young (1-35 years old) outcomes are complex. Network models have emerged as a way to abstract complex systems and gain insights into relational patterns among observed variables. Hypothesis: Network analysis helps provide qualitative and quantitative insights into how various variables interact with each other and affect outcomes in OHCA in the young. Methods: A mixed graphical network analysis was performed using variables collected by CARES. The network allows the visualization and quantification of each unique interaction between two variables that cannot be explained away by other variables in the data set. The strength of the underlying interaction is proportional to the thickness of the connections (edges) between the variables (nodes). We used the mgm package in R. Results: Figure 1 shows the network of the OHCA in the young cases in Chicago from 2013 to 2017. There are apparent clusters. Sustained return of spontaneous circulation and hypothermia are strongly correlated with survival and neurological outcomes. This cluster is in turn connected to the rest of the network by survival to emergency room. The interaction between any two variables can also be quantified. For example, American Indians cases occur more often in disadvantaged locations when compared to Whites (OR 4.5). The network also predicts how much one node can be explained by adjacent nodes. Only 20% of survival to emergency room is explained by its adjacent nodes. The remaining 80% is attributed to variables not represented in this network. This suggests that interventions to improve this node is difficult unless further data is available. Conclusion: Network analysis provides both a qualitative and quantitative evaluation of the complex system governing OHCA in the young. The networks predictive capability could help in identifying the most effective interventions to improve outcomes.