SNOMED CT (Systematized Nomenclature of Medicine - Clinical Terms)

One-sentence definition: SNOMED CT is a comprehensive, multilingual clinical terminology maintained by SNOMED International that provides standardized concept identifiers for diseases, clinical findings, procedures, and anatomy, enabling consistent clinical documentation and computable health data exchange across systems.

Full Definition

SNOMED CT is the most clinically expressive terminology in widespread healthcare use. Where coding systems like ICD-10 classify diseases for billing and epidemiology, SNOMED CT is designed for clinical documentation: it encodes not just diagnoses but findings, symptoms, observable entities, body structures, substances, procedures, and situations — the full vocabulary of clinical practice.

Each concept in SNOMED CT is identified by a permanent numeric concept ID (e.g., 73211009 for “diabetes mellitus”) and defined by its relationships to other concepts. These relationships form a directed acyclic graph: diabetes mellitus is a kind of metabolic disease, which is a kind of disorder, and so on. This hierarchical structure makes SNOMED CT uniquely suited for subsumption queries — finding all patients with any kind of diabetes, including subtypes not yet in existence when the query was written.

SNOMED CT is developed and maintained by SNOMED International, a non-profit member-driven organisation. Member countries (including the US, UK, Australia, Canada, and over 40 others) pay licensing fees and receive the full release. The US extension is managed by the National Library of Medicine. The UK edition is managed by NHS England.

For how SNOMED CT is used in clinical data mapping, when to choose it over LOINC or ICD-10, and how it appears in FHIR value set bindings, see Healthcare Data Semantics.

Context and Usage

Where This Term Appears

SNOMED CT is used across the clinical data layer of healthcare systems:

• FHIR resources: Condition.code, Observation.valueCodeableConcept, Procedure.code, AllergyIntolerance.code, ClinicalImpression.finding.itemCodeableConcept
• EHR clinical documentation: problem lists, allergy records, procedure notes, clinical assessments
• Decision support rules: CDS Hooks logic, quality measures, care gap analyses
• Value set definitions: FHIR ValueSets bound to clinical elements often include SNOMED CT codes
• US Core profiles: US Core Condition requires SNOMED CT (or ICD-10-CM) for Condition.code
• UK Core profiles: NHS England mandates SNOMED CT for clinical concepts throughout the UK Core IG

Common Usage Examples

In conversation: “The problem list uses SNOMED CT — we map from our local codes to SCTIDs at the point of entry so everything downstream is computable.”

In documentation: “The Condition.code element must use a SNOMED CT code from the Clinical Findings hierarchy or an ICD-10-CM code.”

In technical contexts:

"code": {
  "coding": [
    {
      "system": "http://snomed.info/sct",
      "code": "44054006",
      "display": "Diabetes mellitus type 2"
    }
  ]
}

Why SNOMED CT Exists

Early electronic health records used institution-specific or system-specific codes for clinical documentation. A “diabetes” entry in one EHR had no reliable connection to a “diabetes” entry in another — even within the same health system. Aggregation across patients, decision support, and quality measurement required custom translation tables for every pairing of systems.

SNOMED CT (and its predecessor SNOMED) was developed to provide a universal clinical vocabulary that encodes meaning precisely enough to support clinical reasoning. The key distinction: SNOMED CT doesn’t just assign a code to a concept — it defines the concept through its relationships, so that a system can reason over the terminology without additional domain knowledge. “Type 2 diabetes mellitus is a subtype of diabetes mellitus” is encoded in the terminology itself.

Key Characteristics

Concept-based structure

Every SNOMED CT concept has a globally unique, permanent concept ID (a long integer), a preferred term (human-readable name), synonyms, and a set of defining relationships. The concept ID is what software stores and exchanges; the preferred term is for display.

A single clinical concept may have dozens of synonyms in one language and translations in another. The concept ID remains stable across all of them, which is what enables multilingual interoperability.

Hierarchical relationships

SNOMED CT uses IS-A relationships to build a hierarchy (the primary hierarchy). 44054006 (Diabetes mellitus type 2) IS-A 73211009 (Diabetes mellitus) IS-A 154769004 (Metabolic disease) IS-A 64572001 (Disease).

This hierarchy powers subsumption: querying for all patients with any condition that IS-A “Diabetes mellitus” finds all subtypes automatically — gestational diabetes, maturity onset diabetes of the young, unspecified diabetes, and any future subtypes added to the terminology — without changing the query.

Compositional grammar

SNOMED CT includes a formal grammar for composing post-coordinated expressions — combining concepts to represent clinical situations not captured by a single concept. For example, “fracture of left femur” can be composed from “fracture” + “femur” + “left laterality.” Most healthcare deployments use pre-coordinated concepts (single concept IDs) rather than compositions, which simplifies implementation at the cost of some expressivity.

Practical Examples

Diagnosis coding

A patient with heart failure recorded in FHIR using SNOMED CT:

{
  "resourceType": "Condition",
  "code": {
    "coding": [
      {
        "system": "http://snomed.info/sct",
        "code": "84114007",
        "display": "Heart failure"
      }
    ]
  },
  "clinicalStatus": {
    "coding": [
      {
        "system": "http://terminology.hl7.org/CodeSystem/condition-clinical",
        "code": "active"
      }
    ]
  }
}

Coded observation values

SNOMED CT is the preferred system for coded result values in Observations — where LOINC codes the type of observation, SNOMED CT codes the result:

"valueCodeableConcept": {
  "coding": [
    {
      "system": "http://snomed.info/sct",
      "code": "10828004",
      "display": "Positive"
    }
  ]
}

Technical Considerations

Licensing and distribution

SNOMED CT licensing depends on jurisdiction. In SNOMED International member countries (US, UK, Australia, EU members, and others), use within healthcare is free for authorized organizations. The US receives the US edition via the NLM; UK via NHS England. In non-member countries, licensing requires a direct agreement with SNOMED International.

Unlike LOINC (completely free worldwide) or ICD-10-CM (US government publication, free), SNOMED CT licensing is the most complex of the major clinical terminologies. Vendors distributing products containing SNOMED CT codes need a sublicense.

Version management

SNOMED CT releases twice yearly (January and July for the International edition; member editions have their own schedules). Concepts are never deleted — they are retired and replaced, with explicit mappings to the successor. This ensures historical data remains interpretable.

The International edition and national editions (US, UK, AU) are separate releases that can contain country-specific content. Most FHIR value sets reference specific versions; validators should be configured with the edition and version applicable to the deployment context.

Relationship to Other Terms

Related Terms

• LOINC — the observation type coding system; pairs with SNOMED CT for result values
• FHIR — SNOMED CT is used throughout FHIR resources for clinical concepts
• ICD-10 — diagnosis classification for billing; often used alongside SNOMED CT in the same system

Contrasting Terms

• SNOMED CT vs ICD-10: ICD-10 is designed for administrative reporting and billing — it classifies diagnoses into groupings suitable for mortality statistics and reimbursement. SNOMED CT is designed for clinical documentation — it encodes clinical findings with the precision needed for decision support and care coordination. Many EHRs maintain both: SNOMED CT on the problem list, ICD-10 on the claim.

• SNOMED CT vs LOINC: LOINC identifies the type of observation (what test was performed). SNOMED CT identifies clinical findings, diagnoses, and result values (what was found). They are complementary and often appear together in the same FHIR resource: Observation.code uses LOINC; Observation.valueCodeableConcept may use SNOMED CT.

Common Misconceptions

Misconception 1: SNOMED CT is free everywhere

• Incorrect belief: SNOMED CT, like LOINC, is free to use in any product.
• Reality: SNOMED CT is free within member countries for qualified use, but requires a license agreement in non-member countries, and vendors embedding it in distributed software need sublicenses regardless of jurisdiction.
• Why it matters: A health IT vendor distributing software internationally needs to understand the licensing obligations before embedding SNOMED CT codes.

Misconception 2: SNOMED CT replaces ICD

• Incorrect belief: As clinical systems adopt SNOMED CT, ICD codes become unnecessary.
• Reality: They serve different functions and coexist. ICD codes are required for claims, public health reporting, and regulatory submissions. SNOMED CT serves clinical documentation and decision support. Map between them where needed; don’t expect one to replace the other.
• Why it matters: Integration designs that assume SNOMED CT eliminates the need for ICD-10 in billing workflows will fail at the payer interface.

Cross-References

Related Glossary Terms

• LOINC — universal observation coding system; pairs with SNOMED CT
• FHIR — the data exchange standard that uses SNOMED CT throughout clinical resources

---

Last reviewed: January 19, 2026 Definition authority: SNOMED International Content status: Canonical reference