Effects of Predictive Algorithms in Opioid Pain Management
Bauman, Penn, School of Engineering and Applied Science, University of Virginia
Vrugtman, Rosanne, Department of Computer Science, University of Virginia
Graham, Daniel, Department of Computer Science, University of Virginia
Earle, Joshua, Department of Engineering and Society, Virginia Tech
The experience of advanced cancer comes with many struggles, but one of the most constant is the task of pain management.
Behavioral and Environmental Sensing and Intervention for Cancer (BESI-C) is a join research project between the University of Virginia School of Nursing and School of Engineering and Applied Science, which aims to use predictive algorithms to simplify the process of managing cancer pain at home.
The final goal of the project is to create a system which can monitor cancer patients day-to-day activities with noninvasive sensors and use that data to predict pain events.
Predictions will be made by a machine learning algorithm which can analyze patients data in real time.
With these predictions, the system can automate the timing of pain medication, reducing the burden of pain management for patients and their caregivers.
In order to create this predictive algorithm and allow patients to use it, monitoring systems must be created to collect data about patients and their environments.
A prototype of this system is currently being developed and tested by the BESI-C project.
This system consists of cloud infrastructure to store and analyze data as well as devices which are placed in patient homes, including wall-mounted sensor packs, smart watches, a devices to allow the system internet access.
This system allows data to be collected from patients and their homes continually and automatically.
The technical portion of this thesis describes in detail how the wall-mounted sensors work and the other technologies which support them work.
If completed and put into use outside of academic studies, the BESI-C system would be unlike and existing medical product or device.
This makes it difficult to predict possible social side effects or ethical issues that might appear during its use.
To predict, and hopefully prevent, some of these issues before they appear, I have considered ethical issues generated by research into systems similar to the BESI-C system.
To consider the effects of BESI-C system's constant monitoring, I researched self tracking systems.
To compare BESI-C system to existing solutions to support patients dealing with advanced cancer, I consider research into ethical issues surrounding inpatient care.
Finally, to predict ethical issues that might arise from the BESI-C project's use of machine learning, I considered research into algorithm bias.
With the ethical concerns generated by these different fields of research, I am able to predict multiple possible ethical issues which could arise from the BESI-C system.
The ethical concerns generated by these different fields of research predict multiple possible ethical issues which could arise from the BESI-C system.
Knowing potential issues in advance, I can recommend strategies the BESI-C project should focus on to minimize or avoid their impacts, creating a more ethical and more helpful product.
BS (Bachelor of Science)
All rights reserved (no additional license for public reuse)