Why SNOMED CT and dm+d changes matter fundamentally for Digital Health, MedTech, and Biotech companies
Between 2025 and 2026, the NHS will quietly implement one of the most consequential digital infrastructure changes of the decade. It is a change that will not attract headlines, does not alter clinical guidelines, and will barely be noticed by clinicians or patients. Yet for companies developing digital health platforms, medical devices, diagnostics, artificial intelligence systems, and biotechnology solutions, its implications are profound. The NHS is rewriting the language of medicines for the next decade. If your product speaks the old language, it may still run, but it will increasingly be misunderstood.
This is not a policy reform.
It is not a reimbursement decision.
It is not a formulary update.
Instead, it is a structural change to how medicines are represented, modelled, and interpreted across NHS information systems. In effect, the NHS is rewriting the internal language that computers use to understand medicines — and that language underpins prescribing systems, decision support, analytics, real-world evidence, and AI.
SNOMED CT provides a comprehensive, computer-readable clinical vocabulary used across NHS digital records.¹ The UK edition includes national extensions, such as the UK Drug Extension.² The NHS Dictionary of Medicines and Devices (dm+d) is the mandated NHS standard for uniquely identifying medicines and devices, updated weekly and used in data exchange across systems.
At Odelle Technology, we view this moment as a structural inflection point for companies operating in, or planning to enter, the UK healthcare system.
What exactly is happening?
NHS England is updating the SNOMED CT UK Drug Extension so that it fully aligns with the international drug model maintained by SNOMED International.
In practical terms, this means that the UK’s medicines terminology, historically shaped by local requirements and pragmatic shortcuts, is being realigned to match the international SNOMED model for national drug extensions. This is not a cosmetic change to names or codes. It is a redesign of the underlying data model.
Put simply:
The NHS is changing the internal “grammar” that computers use to understand medicines.
This grammar determines how digital systems interpret:
- what constitutes a medicine
- which active ingredients it contains
- who supplies or manufactures it
- how it is formulated and packaged
- and how it relates to other medicines across the care pathway
For decades, the UK’s Dictionary of Medicines and Devices (dm+d) has functioned effectively within the NHS. However, its internal structure has diverged from the international SNOMED model, creating growing challenges around interoperability, analytics, scalability, and international alignment. Those challenges are now being addressed decisively and permanently.
When is this happening?
The changes are being introduced in phases across 2025 and 2026, reflecting the NHS’s preference for stability, testing, and controlled transition:
- Phase 1 – completed during 2025
- Phase 2 – goes live on 18 February 2026
- Phases 3 and 4 – tested during 2026 and planned to go live together later in the year
This staged approach is deliberate. Rather than rushing change into live systems, NHS England is prioritising data integrity, downstream system stability, and safe migration.
NHS England is updating the SNOMED CT UK Drug Extension to align with the SNOMED International drug model, with updates scheduled in phases during 2025–2026. Terminology service documentation describes content and modelling changes to improve alignment and reduce variation between dm+d and the SNOMED CT UK Drug Extension
What do Phases 2, 3, and 4 actually mean in human terms?
Medicines exist at multiple levels of abstraction, from conceptual definitions to the exact pack dispensed to a patient. The NHS is standardising all of these layers.
Phase 2: Cleaning the “middle layer” (February 2026)
Phase 2 focuses on restructuring the abstract, conceptual layer of medicines. These are not branded products or physical packs, but the standardised definitions that sit between high-level drug concepts and specific products.
This work is best understood as foundational plumbing:
- removing legacy groupings and shortcuts
- standardising relationships between concepts
- aligning UK structures with international logic
Most clinicians and patients will never notice this change. However, digital systems, analytics platforms, and decision engines depend on this layer being coherent and stable. Without it, higher-level functionality becomes fragile.
Phase 3: Actual Medicinal Products (AMPs)
Phase 3 moves from abstraction to reality. Actual Medicinal Products (AMPs) represent specific, real-world medicines, for example:
Paracetamol 500 mg tablet, manufactured by Company X
In this phase, NHS England is standardising how:
- active ingredients
- suppliers and manufacturers
- strengths and formulations
- and product relationships
are represented in machine-readable form.
This has direct implications for:
- prescribing and medicines optimisation logic
- clinical decision support
- safety alerts and contraindication checks
- attribution of outcomes and adverse events
In short, Phase 3 defines how digital systems understand what medicine was actually used.
Phase 4: Actual Medicinal Product Packs (AMPPs)
Phase 4 goes one step further, addressing what is physically dispensed to patients.
Actual Medicinal Product Packs (AMPPs) represent concrete items such as:
Paracetamol 500 mg tablets, pack of 32
This phase standardises:
- pack size and units
- pack contents
- relationships between products and packs
These details matter far beyond pharmacy inventory. They affect:
- reimbursement and payment logic
- supply chain integration
- real-world evidence generation
- adherence and persistence analysis
Phases 3 and 4 are being released together to avoid temporary inconsistencies and to minimise system risk.
Why this matters for Digital Health companies
If your digital health product references medicines in any way, whether through prescribing data, pathway analysis, adherence support, deprescribing tools, or AI trained on EHR data, then this change determines whether your system will continue to interpret medicines correctly from 2026 onwards.
Many platforms contain assumptions hard-coded around the historical dm+d structure. As the underlying model changes, those assumptions can lead to:
- silent logic failures
- incorrect alerts or recommendations
- misclassified analytics
- delays or failures in NHS integration
Silent errors are the most dangerous kind, because they undermine trust without triggering obvious faults.
dm+d concepts use SNOMED CT concept IDs and are included in the UK Drug Extension, enabling consistent reference across systems.⁴ By aligning the drug extension model closer to the international SNOMED framework, the NHS aims to improve semantic stability and inter-system interoperability.
Why this matters for MedTech companies, even if you are “not a drug”
Many medical devices interact with medicines indirectly but critically. Devices may:
- trigger medication changes
- rely on medication context for safety or stratification
- generate evidence linked to treatment decisions
Examples include:
- sepsis diagnostics prompting antibiotic escalation
- cardiology devices linked to anticoagulation pathways
- oncology technologies linked to systemic therapies
From an NHS perspective, if your device references medicines, medicines coding becomes part of your regulatory and implementation surface. The changes to SNOMED CT and dm+d define what “correct” looks like going forward.
Why this matters for Biotech and diagnostics companies
For biotech and diagnostics companies, the impact is subtle — and therefore easy to underestimate.
Clinical, economic, and HTA claims often depend on:
- treatment switches
- dose escalation or de-escalation
- persistence and adherence
- downstream healthcare utilisation and costs
Under the new model:
- AMP identifies the actual medicinal product used
- AMPP identifies the exact pack dispensed
That distinction directly affects:
- real-world evidence quality
- HTA submissions and NICE engagement
- post-market surveillance and lifecycle evidence
Weak linkage between diagnostics, treatments, and outcomes translates directly into weaker evidence.
This is also an AI and data-science issue
AI and machine-learning models trained on NHS data depend on:
- stable identifiers
- consistent hierarchies
- predictable relationships
The SNOMED CT and dm+d alignment removes UK-specific shortcuts and introduces semantic stability over time. This enables longitudinal analysis, cross-system learning, and safer deployment of AI at scale.
For AI-driven healthcare companies, this change marks the difference between robust, scalable models and brittle systems that fail at the edges.
Commercial reality: this will influence NHS buying decisions
Increasingly, NHS buyers, integrators, and digital assurance teams ask:
- “Is this SNOMED-compliant?”
- “Does it use dm+d correctly?”
- “Will it survive terminology updates?”
From 2026 onwards, a response that relies on the old model will carry risk. Alignment will increasingly be interpreted as:
- lower integration risk
- lower long-term maintenance cost
- higher strategic value
The real risk of ignoring this
These changes will not cause immediate system failures. Instead, they manifest as:
- subtle data corruption
- inconsistent analytics
- pilots that stall without clear explanation
- gradual erosion of NHS confidence
These are precisely the failures that are hardest to diagnose — and easiest for NHS organisations to walk away from.
References
1. NHS England Changes to dm+d and SNOMED CT UK Drug Extension
NHS England explains why and how the dm+d dictionary and the SNOMED CT UK Drug Extension are being aligned — clarifying differences in content and release cycles and planned enhancements such as ingredient modelling and trade family groups. https://digital.nhs.uk/services/terminology-and-classifications/uk-medicines-terminology-futures/changes-to-digital-terminologies
Provides evidence of specific changes, alignment efforts, and interoperability motives behind terminology updates.
2. NHS England Updating the SNOMED CT UK Drug Extension Model
Official NHS England page describing plans to align the UK Drug Extension model with the international SNOMED model over multiple phases in 2025–2026. https://isd.digital.nhs.uk/trud/users/guest/filters/0/categories/26
Supports statements about modelling changes and phased rollout, clarifying timeline and technical goals.
3. NHS Digital dm+d (Dictionary of Medicines and Devices)
The NHS standard for uniquely identifying medicines and devices, required for electronic exchange in NHS systems. https://digital.nhs.uk/services/terminology-and-classifications/dm-d
Authoritative definition and explanation of dm+d’s status as a national information standard — useful in background sections.
4. NHS BSA SNOMED CT and dm+d Relationship
Explains that dm+d data must be used in electronic medicine data exchange and discusses links with SNOMED CT. https://www.nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/nhs-dictionary-medicines-and-devices-dmd/snomed-ct-and-dmd
Supports claims about practical usage requirements in NHS systems and the linkage between dm+d and SNOMED CT.
5. NHS Digital – SNOMED CT Clinical Terminology Standard
Official SNOMED CT overview as an information standard for clinical terminology in NHS England. https://www.england.nhs.uk/digitaltechnology/digital-primary-care/snomed-ct/
Good for explaining what SNOMED CT is, how it’s used, and why shared clinical language matters.
6. SNOMED CT – International Standards Context
SNOMED CT is maintained by SNOMED International as a global clinical terminology and forms the basis of UK clinical code sets. https://www.snomed.org/what-is-snomed-ct
Useful for international context, interoperability, and justification for aligning UK drug extension with the international model.
7. NHS Digital – Release Structure of SNOMED CT UK Editions
Details that the UK edition includes the International Edition plus UK Drug Extension content in Release Format 2. https://isd.digital.nhs.uk/trud/users/guest/filters/0/categories/26
Helpful for showing the technical packaging of SNOMED CT data and how extensions are integrated.
8. Simplifier UK Core Implementation Guide (dm+d within SNOMED)
Shows that all dm+d concepts are included in the SNOMED CT UK Drug Extension, demonstrating close integration. https://simplifier.net/guide/ukcoreimplementationguideformedicines/NHSDictionaryofMedicinesandDevicesdmd?version=current