Visualizing Prehospital Cardiac Arrest Data
- 1 minDuring the summer of 2016, I interned as a data analyst with Dr. Leonard Weiss at the University of Pittsburgh’s Department of Emergency Medicine. Dr. Weiss’ research is primarily in the field of prehospital cardiac arrest treatment. Using a large dataset of patient care reports from local paramedics and emergency medical technicians (EMTs), I identified and investigated metrics capable of enhancing understanding of emergency medical services’ responses to cardiac arrests using traditional hypothesis testing, time series analysis, and autoregressive integrated moving average models. After identifying key metrics and trends, I presented the results presented to physicians, faculty, and researchers via an interactive application built using R and Shiny. This approach allowed audience members to easily explore the visualized conclusions and trends themselves, rather than inundate members with static bar charts.
While the interactive Shiny application contains protected health information and therefore cannot be shared publically, below are screenshots of the application using notional data. Please note that all instances and trends presented in the screenshots below are based on synthesized data and are not reflective of real patient care or EMS operations.
Header image source: PlusPNG.com.
Tom Goode
Data Scientist & EMS Researcher