Introduction
Like many people, I have a running Note on my iPhone that highlights all my big ideas. The type of ideas you come up with while fighting insomnia on a Monday night. Naturally, many of ideas have to do with applying data science in healthcare. Throughout my Master’s coursework in the past year, I’ve had the opportunity to reflect on the combination of what I’m learning and my background as an analyst in healthcare system operations. The result is a bunch of cool ideas that I would love to work on in my future career. In this post, I will introduce a few of these ideas with the aim to more fully elaborate on each one in a future post.
An Outline
The ideas I will touch on include the following :
- Patient Segmentation
- Automated Process Mining
- Structured Causal Inference Processes
- N of 1 Trial Methodology
- Patient Similarity Algorithms
- Flexible Healthcare Machine Learning Platform
All of these ideas fall under what might be considered a Learning Health System. As defined by the Institute of Medicine in the link above, a Learning Health System is a “system in which science, informatics, incentives, and culture are aligned for continuous improvement and innovation, with best practices seamlessly embedded in the delivery process and new knowledge captured as an integral by-product of the delivery experience”. In simple terms, the system should learn from every patient that interacts with the system, and then disseminate the resulting knowledge to care providers. The above ideas focus more on the learning side of things, but I will also touch on how to operationalize these tools for use in care delivery and analysis.
Conclusions
While everything listed above is merely an idea, they are all possible and have important implications for improving healthcare delivery and outcomes. In the specific posts, I hope to specify how these ideas can be implemented and explain their benefits. Healthcare systems, payers, and product creators will all be interested in these techniques, and I’m sure some of them are already being worked on. As always, feel free to reach out and connect if interested!