Clinical coding sits at the heart of how the NHS records, shares, and makes sense of patient information. Every diagnosis documented, every procedure performed, and every condition managed generates data. Without a consistent, structured way of capturing that data, it becomes difficult to coordinate care, fund services accurately, or identify patterns across populations. For clinical administrators working within NHS systems, understanding the coding standards that underpin these processes shapes how records are maintained, how data flows between organisations, and how clinical documentation translates into actionable information.

What are clinical coding standards and why do they matter?

Clinical coding is the process of converting clinical information (diagnoses, symptoms, procedures, findings) into standardised, machine-readable codes. These codes allow healthcare organisations to communicate consistently across different systems, settings, and countries.

In the NHS, clinical coding underpins several critical functions:

• Patient safety: Accurate coded records ensure that clinicians across care settings have access to consistent, structured information about a patient's history

• NHS funding and commissioning: Coded inpatient episodes determine Healthcare Resource Group (HRG) classifications, which directly affect hospital tariffs and funding allocations

• National statistics and research: Coded data feeds into public health surveillance, mortality reporting, and epidemiological research

• System interoperability: Standardised codes allow medical record systems to exchange structured data reliably across primary, secondary, and community care

Without consistent coding, data quality across NHS datasets deteriorates rapidly, making it harder to identify care gaps, allocate resources, or support population health analytics.

The three main clinical coding standards used in the UK

Three coding systems are most frequently encountered in UK healthcare settings: SNOMED CT, ICD-10, and Read codes. Each plays a distinct role:

• SNOMED CT is the mandated clinical terminology for recording information at the point of care across NHS medical record systems

• ICD-10 is the international classification system applied retrospectively to inpatient episodes for administrative, statistical, and funding purposes

• Read codes are a legacy system, now deprecated, but still embedded in decades of historical GP records

Understanding how these systems relate to one another, and where each applies, is essential for clinical administrators managing records, data flows, and coding governance.

SNOMED CT: the UK's primary clinical terminology standard

What SNOMED CT is and how it works

SNOMED CT (Systematized Nomenclature of Medicine – Clinical Terms) is the most comprehensive clinical terminology system available globally. It provides a structured, computable way of recording clinical findings, diagnoses, procedures, and observations, so that information can be understood consistently by both humans and computer systems.

Unlike older flat-list coding systems, SNOMED CT is built around a concept-based architecture. Each clinical concept receives a unique numeric identifier, a preferred term, synonyms, and a set of defined relationships to other concepts. This hierarchical and relational structure lets the system represent clinical meaning with a level of precision that simpler systems cannot match.

For example, a concept such as 'type 2 diabetes mellitus with peripheral neuropathy' is not just a label. Within SNOMED CT, it carries explicit relationships to its parent concepts (diabetes mellitus, peripheral neuropathy), so systems can reason over that data automatically.

Where SNOMED CT is used in the NHS

NHS England mandated SNOMED CT as the required terminology for all clinical documentation in primary care medical record systems, with an implementation deadline of April 2018. The mandate was subsequently extended to secondary care, mental health, and community care settings, with a deadline of April 2020.

This is confirmed in the NHS Standards Directory entry for SNOMED CT, which lists it as an active, mandated national information standard under the Health and Social Care Act 2012.

The UK maintains several national extensions to the core SNOMED CT release, managed by NHS England:

• UK Clinical Extension: Covers clinical concepts specific to the NHS

• UK Drug Extension: Integrates with the Dictionary of Medicines and Devices (dm+d)

• UK Pathology Extension: Supports laboratory and pathology reporting

These extensions ensure that SNOMED CT reflects the specific clinical and administrative context of NHS practice, rather than relying solely on the international release.

In Scotland, Public Health Scotland has confirmed that SNOMED CT will replace Read codes as the national clinical terminology for both primary and secondary care, though the transition timeline differs from England's.

How SNOMED CT codes are structured

The architecture of SNOMED CT is built around three core components:

• Concepts: Each unique clinical idea receives a permanent numeric identifier (for example, 44054006 for 'diabetes mellitus type 2')

• Descriptions: Each concept has a preferred term and may have multiple synonyms, so clinicians can record using natural language while the system maps to the underlying concept

• Relationships: Concepts connect through defined relationships (such as 'is a', 'finding site', or 'associated morphology'), creating a semantic network rather than a simple list

This structure makes SNOMED CT significantly more expressive than older flat-list systems such as Read codes. It also supports cross-mapping to other coding systems, including ICD-10, so that data recorded in SNOMED CT can be translated for statistical and administrative purposes without requiring separate manual coding.

SNOMED CT and clinical decision support

The relational structure of SNOMED CT is what makes it particularly valuable for modern clinical systems. Because concepts connect through defined relationships, medical record platforms can use SNOMED CT data to:

• Trigger automated clinical alerts (for example, sepsis screening based on coded observations)

• Support population health analytics by querying across related concepts

• Provide structured, computable input data for AI-assisted clinical decision support

A published pipeline for automated heart failure diagnosis using SNOMED CT demonstrated how structured clinical notes, when coded consistently, can be processed by machine learning models to support diagnostic accuracy. This illustrates the downstream value of high-quality coded data at the point of care.

Research into hybrid frameworks for clinical concept extraction from discharge summaries has also shown that standardised terminologies like SNOMED CT provide the structured foundation that natural language processing (NLP) tools require to extract and classify clinical information reliably.

ICD-10: the international standard for disease classification

What ICD-10 is and who maintains it

ICD-10 (International Classification of Diseases, 10th Revision) is maintained by the World Health Organization (WHO) and is used globally to categorise diseases, conditions, injuries, and causes of death. It uses alphanumeric codes, structured as a letter followed by up to four digits, to classify clinical conditions for statistical and administrative purposes.

ICD-10 is distinct from SNOMED CT in both purpose and scope. SNOMED CT captures clinical meaning at the point of care with high granularity. ICD-10 groups conditions into categories suitable for epidemiological analysis, mortality reporting, and resource allocation.

How ICD-10 is used in UK secondary care

In NHS hospitals, trained clinical coders apply ICD-10 codes to inpatient episodes after discharge. This retrospective coding process draws on clinical documentation (discharge summaries, operative notes, and clinical letters) to assign codes representing the principal diagnosis, secondary diagnoses, and complications.

These codes serve several specific functions in secondary care:

• HRG classification: ICD-10 codes, combined with OPCS-4 procedure codes, determine the Healthcare Resource Group assigned to each episode, which directly drives payment under the NHS national tariff

• National statistics: Coded episode data feeds into Hospital Episode Statistics (HES) and national mortality data

• Commissioning and planning: Aggregated coded data informs NHS commissioning decisions and service planning

As NHS England's guidance on ICD-10 and OPCS-4 makes clear, the roles of SNOMED CT and ICD-10 are complementary rather than competing. SNOMED CT supports direct care recording, while ICD-10 and OPCS-4 serve epidemiological, statistical, and health management functions.

The National Clinical Coding Standards for ICD-10 and OPCS-4 are updated annually, with the version applicable from 1 April 2025 now available electronically via the NHS Classifications Browser. The printed reference books have been discontinued.

ICD-10 vs ICD-11: what the transition means for the UK

The WHO released ICD-11 in 2019, and it came into effect internationally in January 2022. ICD-11 offers a substantially modernised structure, including better alignment with SNOMED CT, improved granularity, and a digital-first design.

The UK is currently planning its transition from ICD-10 to ICD-11. For NHS coding teams, this will eventually require updates to coding workflows, training, and the data infrastructure that processes and stores coded episode data. Existing ICD-10-coded historical data will need to remain accessible and interpretable alongside any new ICD-11 coded records, creating a period of parallel operation with significant implications for clinical administrators managing data governance.

Read codes: a legacy system still present in UK records

What Read codes are and where they came from

Read codes were developed in the 1980s by Dr James Read, a Loughborough GP, and became the standard clinical coding system for UK primary care over the following decades. Two main versions exist: Read Version 2 (Read v2) and Clinical Terms Version 3 (CTV3), also known as Read Version 3.

Both versions were deeply embedded in the major GP clinical systems used across England, including EMIS and Vision, and recorded diagnoses, symptoms, medications, investigations, and administrative events across millions of patient records.

Why Read codes are considered a legacy system

Read codes have several structural limitations that ultimately led to their deprecation:

• Hierarchical inconsistencies: The classification structure of Read codes was not consistently applied, making it difficult to query data reliably across systems

• Limited expressivity: Read codes use a five-character alphanumeric format that constrains the granularity and nuance of clinical recording compared to SNOMED CT's concept-based model

• Lack of international interoperability: Read codes are a UK-specific system with no international equivalent, limiting their utility for cross-border data exchange or research

These limitations are why NHS England mandated migration to SNOMED CT, with Read codes formally deprecated in primary care from approximately 2018, while secondary care systems were required to implement SNOMED CT by March 2020.

Why Read codes still matter today

Despite their deprecation, Read codes remain clinically relevant for one important reason: decades of patient records held in GP systems contain Read-coded data. For clinical administrators, this has practical consequences:

• Longitudinal care records for patients registered with a GP for many years will contain a mixture of Read-coded historical data and more recent SNOMED CT-coded entries

• Accurate mapping between Read codes and SNOMED CT concepts is essential to ensure that historical data remains interpretable and can be used safely in clinical decision-making

• Research datasets and audit tools that draw on historical primary care data must account for the Read-to-SNOMED transition to avoid misclassification

The UK LLC dataset documentation confirms that Read codes were deprecated in primary care around 2018 and in secondary care by March 2020, but notes their continued presence in historical NHS England datasets. This is a practical reality that administrators working with longitudinal data will encounter regularly.

Research into semantic mapping challenges when translating local terminologies to SNOMED CT highlights that this kind of migration is not straightforward. Translation and modelling decisions introduce risks, as concepts that appear equivalent may carry subtly different clinical meanings. This underscores the importance of validated mapping tools and governance oversight during any terminology transition.

How SNOMED CT, ICD-10, and Read codes work together

These three systems are not mutually exclusive. In practice, NHS data flows depend on all three being understood in relation to one another.

The general model works as follows:

• At the point of care, clinicians record using SNOMED CT within their medical record system. This captures clinical meaning with high granularity and supports direct patient care

• After discharge or episode completion, trained clinical coders translate the clinical documentation into ICD-10 (and OPCS-4) codes for administrative, statistical, and funding purposes

• For historical records, Read-coded data from GP systems continues to exist alongside SNOMED CT-coded records, with crossmaps providing the bridge between systems

SNOMED International maintains crossmaps between SNOMED CT and ICD-10, ICD-O, and MedDRA. These semi-automated maps, validated by both WHO and SNOMED International, allow SNOMED CT-recorded clinical data to be translated into ICD-10 classifications for reporting purposes, reducing the need for entirely separate manual coding processes in some contexts.

Crossmapping is not always a perfect one-to-one translation. SNOMED CT's greater granularity means that a single SNOMED CT concept may map to a broader ICD-10 category, and some clinical nuance is inevitably lost in translation. This is an acknowledged limitation of any crossmapping approach, and coding governance frameworks need to account for it.

Clinical coding and AI: how automation is changing the landscape

Artificial intelligence is beginning to change how clinical codes are generated and applied, with implications for both coding accuracy and documentation burden.

AI medical assistants and ambient voice technology are now capable of processing clinical documentation (including consultation transcripts and discharge summaries) and suggesting or generating structured clinical codes. This has several potential benefits for NHS workflows:

• Reduced documentation burden: Clinicians spend less time manually selecting codes within medical record interfaces when AI tools can propose appropriate SNOMED CT concepts based on the clinical narrative

• Improved coding consistency: Automated suggestions based on structured terminologies can reduce the variability in how similar conditions are coded across different clinicians or settings

• Faster episode coding in secondary care: AI-assisted tools may support clinical coders in identifying relevant ICD-10 codes from discharge documentation more efficiently

Research into NLP-based approaches to diagnostic code variability has shown that natural language processing can improve consistency in clinical coding by identifying cases that manual coding workflows might miss or classify inconsistently, particularly in complex, multi-condition patient populations.

AI-assisted coding does have limitations. Automated systems depend on the quality and completeness of the underlying clinical documentation. If the source text is ambiguous or incomplete, the resulting code suggestions will reflect those gaps. Human oversight, whether from clinicians selecting codes at the point of care or trained coders reviewing AI-generated suggestions, remains essential to maintain coding accuracy and clinical governance standards.

Who is responsible for clinical coding in the NHS?

Responsibility for clinical coding in the NHS is divided between two distinct groups, each operating in a different setting with different tools and standards.

In primary care, clinicians (GPs, nurses, and other registered practitioners) record coded data directly within their medical record system at the point of care, using SNOMED CT. The accuracy of this coded data depends on clinician training, system configuration, and the quality of the templates and workflows available within the system.

In secondary care, dedicated clinical coding professionals translate inpatient episode documentation into ICD-10 and OPCS-4 codes after the patient has been discharged. Clinical coders in NHS hospitals are typically trained to national standards, and many hold qualifications from the Institute of Health Records and Information Management (IHRIM). Their work directly affects hospital income, national statistics, and the integrity of NHS datasets.

The governance implications of each model differ. In primary care, coding accuracy is embedded in clinical workflow and is difficult to audit retrospectively without structured data quality programmes. In secondary care, coding is a distinct professional function with defined quality standards, audit processes, and national benchmarking.

As medical record systems become more sophisticated and AI-assisted coding tools become more prevalent, the boundary between these two models is likely to evolve. Clinical coders will increasingly work alongside automated tools rather than solely from paper or PDF documentation.

Key organisations and resources governing clinical coding in the UK

Several organisations play a role in setting, maintaining, and governing clinical coding standards across the NHS:

• NHS England is responsible for mandating and overseeing the implementation of SNOMED CT across NHS providers in England, and for publishing the National Clinical Coding Standards for ICD-10 and OPCS-4

• NHS Digital (now integrated into NHS England) manages the UK SNOMED CT releases, including the UK Clinical Extension, UK Drug Extension, and UK Pathology Extension, and provides the NHS Classifications Browser

• SNOMED International is the non-profit organisation that owns and maintains the core SNOMED CT release, and publishes the crossmaps to ICD-10 and other classification systems

• The World Health Organization (WHO) maintains ICD-10 and ICD-11, and coordinates their international implementation

For clinical administrators seeking further technical detail, the following are authoritative starting points:

• The NHS England's guidance on clinical coding and SNOMED CT provides definitions and standards for NHS data items, including coded clinical data

• The NHS Standards Directory entry for SNOMED CT sets out the legal basis, implementation requirements, and current status of the mandate

• The UK LLC coding guidance offers a clear, structured overview of all coding systems present in NHS England datasets, useful for administrators working with linked or research-grade data

• ISD Scotland's terminology services page covers the coding landscape specific to Scottish NHS settings

Frequently asked questions

▶ What is clinical coding and why does it matter in the NHS?

Clinical coding converts clinical information (diagnoses, symptoms, procedures, and findings) into standardised, machine-readable codes. In the NHS, accurate coding underpins patient safety, hospital funding, national statistics, and the ability of medical record systems to exchange structured data reliably across primary, secondary, and community care.

▶ What are the three main clinical coding standards used in the UK?

The three systems most commonly encountered in UK healthcare are SNOMED CT (Systematized Nomenclature of Medicine – Clinical Terms), ICD-10 (International Classification of Diseases, 10th Revision), and Read codes. SNOMED CT is the mandated standard for recording information at the point of care. ICD-10 is applied retrospectively to inpatient episodes for administrative and funding purposes. Read codes are a legacy system, now deprecated, but still present in decades of historical GP records.

▶ Is SNOMED CT mandatory across the NHS?

Yes. NHS England mandated SNOMED CT as the required terminology for all clinical documentation in primary care medical record systems, with an implementation deadline of April 2018. The mandate was extended to secondary care, mental health, and community care settings, with a deadline of April 2020. It is listed as an active, mandated national information standard under the Health and Social Care Act 2012.

▶ How does ICD-10 differ from SNOMED CT?

SNOMED CT captures clinical meaning at the point of care with high granularity, supporting direct patient care and clinical decision support. ICD-10 groups conditions into broader categories suited to epidemiological analysis, mortality reporting, and resource allocation. In NHS hospitals, trained clinical coders apply ICD-10 codes retrospectively after a patient is discharged, drawing on discharge summaries and clinical letters. The two systems are complementary rather than competing.

▶ Why are Read codes still relevant if they've been deprecated?

Read codes were formally deprecated in primary care from around 2018 and in secondary care by March 2020, but decades of patient records held in GP systems still contain Read-coded data. Patients registered with a GP for many years will have longitudinal records mixing Read-coded historical entries with more recent SNOMED CT-coded ones. Accurate mapping between the two systems is essential for safe clinical decision-making and reliable research datasets.

▶ How do SNOMED CT, ICD-10, and Read codes work together in practice?

Clinicians record using SNOMED CT within their medical record system at the point of care. After discharge, trained clinical coders translate that documentation into ICD-10 and OPCS-4 (Office of Population Censuses and Surveys Classification of Interventions and Procedures, version 4) codes for administrative and funding purposes. For historical records, Read-coded data sits alongside SNOMED CT entries, with crossmaps providing the bridge between systems. SNOMED International maintains crossmaps between SNOMED CT and ICD-10 to support this translation.

▶ How is AI changing clinical coding in the NHS?

AI medical assistants and ambient voice technology can now process clinical documentation, including consultation transcripts and discharge summaries, and suggest or generate structured clinical codes. This can reduce the documentation burden on clinicians, improve coding consistency, and help clinical coders identify relevant ICD-10 codes from discharge documentation more efficiently. Human oversight remains essential, as automated systems depend on the quality and completeness of the underlying clinical documentation.

▶ Who is responsible for clinical coding in primary and secondary care?

In primary care, clinicians (GPs, nurses, and other registered practitioners) record coded data directly within their medical record system at the point of care using SNOMED CT. In secondary care, dedicated clinical coding professionals translate inpatient episode documentation into ICD-10 and OPCS-4 codes after discharge. Many NHS hospital coders hold qualifications from the Institute of Health Records and Information Management (IHRIM), and their work directly affects hospital income and national statistics.

▶ What does the transition from ICD-10 to ICD-11 mean for NHS coding teams?

The World Health Organization released ICD-11 in 2019, and it came into effect internationally in January 2022. The UK is currently planning its transition from ICD-10 to ICD-11. For NHS coding teams, this will eventually require updates to coding workflows, training, and data infrastructure. Existing ICD-10-coded historical data will need to remain accessible alongside new ICD-11-coded records, creating a period of parallel operation with significant implications for clinical administrators managing data governance.

▶ Where can clinical administrators find authoritative guidance on NHS coding standards?

NHS England publishes guidance on clinical coding and SNOMED CT, covering definitions and standards for NHS data items. The NHS Standards Directory entry for SNOMED CT sets out the legal basis and current mandate status. The UK LLC coding guidance provides a structured overview of all coding systems present in NHS England datasets. For Scotland, ISD Scotland's terminology services page covers the coding landscape specific to Scottish NHS settings.