Health Data in the Wild
Health data is messy, particularly when it is collected for care rather than research. It is full of inconsistencies, missing data, and bias. As someone who has worked with electronic health records for over a decade but also been on the other end of it as a GP, I have a particular perspective on what health data is actually like. I have written a lot about clinical coding, codelists, working with messy data, data pipelines, and the bias that arises when data is collected for care rather than research. These posts are about what it is actually like to work with health data in the wild.
What Do We Mean by Bias in Health Data Research?
A reference guide to what bias means in health data research and the distinct types you meet in electronic health records and other health datasets.
How to Create a Codelist
A practical guide to creating codelists from scratch for health data research.
What is a Codelist?
What are codelists and why do they matter? An accessible introduction to coding systems, clinical nuance, and research reproducibility.
Clinic to Code to Care
This blog came out of a talk Steph Jones and I gave at Women in Data and AI in October 2025. It explores the journey of information from a patient in clinic to how that information is coded for research and ultimately ends up informing statistical and machine learning models that can help improve patient care.
SNOMED and friends
This blog provides an introduction to SNOMED codes and how they are used in routine care in the UK. It also covers some of the quirks of SNOMED and the challenges of using it in research.
An Introduction to Electronic Health Records
A quick primer on what is an electronic health record and how it is used in clinical practice and research. This post is UK focussed but the principles are the same in many other countries.