Application of Data Science to Paramedic Data

Description

Paramedic data has significant potential for research. Paramedics see many patients every year and collect a wide variety of crucial data at each encounter. This data is rarely used for good reason: it’s messy and hard to work with. But like theunderdog character in a classic movie, with a little bit of work and a lot of understanding, paramedic data has significant potential to change the world of medical research. Paramedics throughout the world are involved in research every day, but most of this research uses purpose-built data structures and never takes advantage of the existing data that paramedics create as part of their everyday work. Through a project-based approach grounded in developing a better understanding of the opioid crisis, this thesis will examine the quantity and structure of the existing paramedic data, the complexities of its current design, the steps necessary to access it, and the processes necessary to clean existing data to a point where it can be easily modelled. Once we have our dataset, we will explore the challenges of choosing key metrics by examining the effectiveness of metrics currently employed to monitor the opioid crisis and the influences public health programs and changing policies have had on these metrics. Next, we will explore the temporal distributions of opioid and other intoxicant use with an eye to providing data to support public health in their harm reduction efforts. And lastly, we will look at the effect of fixed- and floating-point temporal influences on intoxicant-related calls with an eye to how these temporal points can affect call volumes. By using this exploration of the opioid crisis, this thesis will show that with a more thorough understanding of what paramedic data is, what data points are available, and the processes needed to transform it, paramedic data has the potential to greatly expand the limits of health care data science into a more precise and more all-encompassing discipline.
Author Keywords: Ambulance, Data Science, Opioid, Overdose, Paramedic, Pre-hospital