In the meantime, you can read a summary of my research on my website: tbeaney.github.io/MMclustering... or access my thesis here 📚 : doi.org/10.25560/116...

Over the next week, I’ll share the key findings from each chapter of my thesis. But my main takeaway: clustering in healthcare should be a tool to a well-defined end, and not an end in itself.

My thesis explores clustering, a method which groups similar diseases or people to identify common patterns in the population. I also use methods that consider the order in which conditions develop, not just the conditions someone has at a given time.

Addressing this is difficult because MLTC is a broad term encompassing many diseases and many people. In fact, there are tens of thousands of unique combinations of diseases in the population. How can we simplify this complexity?

Globally, more people are living with multiple long-term conditions (MLTC). But health systems are designed around single diseases, meaning that people with MLTC often need to see multiple specialists, creating care which is hard to navigate, inefficient and fragmented.