View Submission

I have been interested since medical school in how we can sum or compare the total effect of analgesia across different opioid medications. This project gave me the opportunity to start at the beginning: describing the measurement bias in total opioid burden before trying to improve the method.

The mathematics was initially intimidating, but developing a proof to explain the bias was one of the most rewarding parts of the project.

Aims and Background

Published conversion tables have offered guidance on converting between opioids for the last 20 years. More recently, total opioid burden (TOB) has emerged as a quantitative outcome in clinical trials. Although these measures allow researchers to combine different opioid medications into a single outcome, the conversion values can also introduce measurement bias when comparing TOB between groups.

During my PhD Year 1 secondment, I described and demonstrated this bias. The project combined simulation studies with an exploratory literature search and a systematic review protocol, providing both a demonstration of the problem and a pathway to understand how it appears across the perioperative literature.

Work

Describing the bias

The project began with simulation studies to explore how opioid conversion values could affect comparisons between groups. Following feedback from colleagues, I pursued a mathematical proof to describe the bias rather than relying on simulations alone. The proof allowed the simulations to quantify and demonstrate the problem more clearly.

This was my first experience of using mathematics to communicate an important but nuanced clinical-methodological point. It was also an opportunity to work in a new field while keeping the findings connected to my areas of interest and expertise.

Understanding the literature

The literature work was deliberately focused on a small and closed topic. The exploratory search provides a snapshot of how TOB is currently used, while the systematic review protocol sets out a route to investigate the issue more formally.

Building this workstream from the underlying problem, rather than starting with a later-stage improvement, was an important decision. I hope that describing the bias first will capture the attention of invested readers and lay the groundwork for the follow-up project.

Learning and Reflections

This project was my first experience of the intersection between clinical studies and methodology and statistics. I enjoyed the opportunity to move into a new field while ensuring that the findings and implications remained clinically relevant.

Another significant learning point was how to build a research narrative and pipeline of work. My initial idea was a later-phase improvement to the methodological problem. With encouragement from colleagues, I started at the beginning with the problem itself. The concept of building projects that complement and lead into each other will serve me well when planning my larger PhD.

Future Work

Total opioid burden is likely to feature heavily in my current and future studies. By describing the methodological issues and setting out a reporting guideline, I am well placed to use these outcomes in a scientifically robust way.

This project has also opened up a follow-on project, which I currently plan as a summer mini-project, exploring new ways to develop opioid conversion values. The systematic review protocol will also be advertised to incoming pharmacy students as a potential master’s student project that I could supervise.

This project has given me a stronger foundation for using total opioid burden, while also showing me how a carefully defined methodological problem can become the first step in a longer programme of research.