Clinical codesets sit at the heart of modern healthcare. They determine who is identified, what is measured and even how decisions are made. Yet despite their centrality, the skills required to achieve high‑quality clinical codesets remain poorly understood and are inconsistently taught.
As healthcare systems become increasingly data-driven, the need for structured training and professional development in codeset creation is becoming more important. There is a clear gap between the growing reliance on coded data and the level of expertise required to design, validate and maintain high-quality codesets.
The hidden skillset behind every good codeset
Writing a clinical codeset is not an administrative task; it is a multidisciplinary craft that draws on the following:
- clinical expertise to define the intent and understand the nuances of disease presentation
- terminology knowledge to navigate SNOMED CT, dm+d, and other coding systems
- health informatics insight to understand how data is captured, stored and queried
- analytical thinking to translate clinical concepts into precise, reproducible definitions
- governance awareness to ensure transparency, version control and auditability
This combination is rare - and without training, remains inaccessible to many who produce or use codesets in their daily work. It is most effective when paired with strong attention to detail and a willingness to recognise and learn from errors.
Why training and mentorship matters
The NHS is advancing integrated care, shared analytics platforms, system‑level decision‑making and widespread AI adoption. In this context, a poor codeset could have far‑reaching consequences.
Training and development is crucial for:
- patient safety; mis-specified and inappropriately positioned codesets could miss high‑risk patients or generate false positives (and so burden clinical teams)
- research quality - cohort definitions shape findings. If the codeset is flawed, the evidence base becomes distorted
- supplier credibility - digital products built on weak codesets undermine trust and create variation
- interoperability - consistent, well‑constructed codesets are the glue that holds multi‑supplier ecosystems together
- workforce sustainability - without good training and mentoring, organisations might rely on a very small handful of experts (a fragile and unsustainable model)
- realising the benefits of AI – accurate, well‑defined codesets enable algorithms to identify the right patients, support clinical decisions, and drive meaningful innovation
System-wide capability gap
Despite their importance, codeset skills are not yet consistently embedded within NHS workforce development plans, national digital capability frameworks or research institution training programmes. As a result, teams often develop expertise through experience, rather than through structured support, best practice or formal clinical oversight. Recognising codeset development as a professional competency could help strengthen this area, supported by mentorship, sustained investment and organisational recognition to build and maintain a skilled workforce.
What good training and mentorship looks like
A modern training and education programme for codeset development would ideally include:
- foundations of clinical terminology - understanding SNOMED CT hierarchies, dm+d structures, ICD-10 classification etc., along with coding rules and coding behaviour across primary, secondary and other care sectors
- principles of good codeset design - purpose definition, inclusion/exclusion logic, metadata creation, and alignment with clinical practice (includes understanding the context they are to be deployed in)
- practical skills in data exploration - how to inspect source data, identify anomalies and validate ;assumptions
- clinical oversight and governance - how to work with clinicians, document decisions and maintain exacting version control
- ethical and clinical safety considerations - recognising the downstream impact of errors and the importance of transparency
- tools for managing codesets – tools and platforms to support the creation, curation, validation and maintenance of codesets
- maintenance and lifecycle management - how to update codesets in response to new releases, guideline changes and new evidence
Professional development
The ideal future would be one where codeset development is more firmly established as a professional discipline. Steps that might help include:
- developing structured training pathways
- recognising codeset development as specialism
- embedding it in digital workforce strategies
- developing consistent standards and supporting suppliers to adopt them
- encouraging academia to teach codeset methodology amongst its clinical research community
- creating communities of practice/mentorship networks
Final thoughts
High-quality clinical codesets are the bedrock of safe, modern healthcare. They do not happen by accident, but are the product of skilled people who are supported by training and the right working culture.
PRIMIS' long tradition of training and supporting GP Clinical Advisers provides a strong foundation for future development. The team lead our codeset development work, helping to ensure clinical relevance, robustness and practical applicability.
Kerry Oliver, PRIMIS