For my MSc dissertation, I wanted to look at variations in clinical practice in hospital medicine. Building off my time working in acute medicine in Dublin, I was very aware that individual practice varied hugely, despite both national and hospital guidelines, let alone the huge swathes of research and other medical resources.

I decided to do a realist systematic review, inspired by the work of Chris Bonell and others. I liked that this approach combined the rigour and replicability of the traditional systematic review, with the rich description and colour of the realist paradigm. This was particularly important for this project, as I wanted to synthesise evidence on complex social interventions. I wanted to go beyond a simple pooled estimate of effect size and figure out what worked in what circumstances, and why.

I found that interventions that worked used strategies that educated, persuaded, or prompted clinical decision-makers. Interventions that didn’t work failed to anticipate and address subsequent barriers to implementation.

There were two big learning points for me from this work. The first was the lack of high-quality published evaluations, with the vast majority of papers being simple uncontrolled before and after studies. Additionally, nearly all papers had very poor description of their interventions and context. They often failed to thorougly describe the problem at hand, previous attempts at solving the problem, and other contextual factors. They also rarely detailed the theorised mechanism at play and how this interacted with the context to produce the desired outcome. This made it difficult to generalise the results to other settings.

This project deepened an interest in evaluation for me, and has led me to think more about the quality and depth of evaluations, particularly in the healthcare space. I was delighted to receive a distinction mark for this project, and hoping to publish the work in the near future. If you’d like to read the project or chat about this, please get in touch!