Direct comparison
SNOMED vs ICD-10: Clinical Coding Differences
Compare SNOMED CT and ICD-10 architectures, point-of-care EHR documentation vs statistical classification, NLM cross-mapping, and how LOINC relates.
Written and maintained by CASRAI Editorial Board
Last updated
Ask CASRAI · free to try
Ask about SNOMED vs ICD-10: Clinical Coding Differences
Ask your first 2 questions free below. Subscribers get 150 a day for $29 a month.
An AI assistant specialized in research administration. It cites the sources behind every answer, labels web answers and says when it can't answer.
Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.
Works on this site and inside Claude, Cursor and the AI tools you already use.
Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays free to read, with no account and no card.
How do SNOMED CT, ICD-10 compare side by side?
The table below compares SNOMED CT, ICD-10 across 10 procurement-relevant dimensions, from fundamental nature through source.
Side-by-side comparison
| Dimension | SNOMED CT | ICD-10 |
|---|---|---|
| Fundamental nature | Multiaxial reference terminology and ontology designed to capture detailed, unambiguous patient clinical data at the point of care. | Statistical classification system designed to aggregate diseases and health conditions into mutually exclusive categories for reporting and billing. |
| Architecture and data model | Poly-hierarchical ontology built on formal Description Logic (OWL-compliant) with concept IDs (SCTID), descriptions, and semantic relationships. | Mono-hierarchical tabular classification organized into 22 chapters with rigid alphanumeric code structures (A00–Z99) and single parentage. |
| Granularity and scope | Over 350,000 active concepts and >1 million descriptions covering diagnoses, clinical findings, surgical procedures, body anatomy, and organisms. | ~14,400 base WHO categories (expanded to ~70,000 codes in US ICD-10-CM) focused strictly on diseases, injuries, and external causes of morbidity. |
| Primary operational purpose | Direct clinical EHR documentation, problem lists, electronic care plans, and real-time clinical decision support (CDS) alerting rules. | Administrative health insurance billing, reimbursement claims, hospital discharge case-mix index, and national mortality tracking. |
| Post-coordination capability | Built-in compositional grammar enabling clinicians to combine concepts (e.g., finding site, laterality, severity) dynamically at data entry. | Strictly pre-coordinated; codes are fixed and static, with limited pairing via dagger (†) and asterisk (*) etiology-manifestation rules. |
| Relationship to LOINC | Complementary; LOINC specifies the observation or test question (e.g., serum potassium), while SNOMED CT represents findings and result values. | Indirect; ICD-10 classifies final clinical conditions resulting from tests, but does not represent laboratory observation codes or test orders. |
| Maintenance and governance | Owned and governed internationally by SNOMED International (IHTSDO), with domestic editions maintained by national release centers (e.g., US NLM). | Owned and governed by the World Health Organization (WHO), with national clinical modifications managed by domestic agencies (e.g., CDC NCHS/CMS). |
| Cross-mapping mechanisms | Source of standard clinical maps; US NLM produces rule-based SNOMED CT to ICD-10-CM maps (Refset 6011000124106) using the I-MAGIC algorithm. | Target of clinical mapping; mapping from ICD-10 back to SNOMED CT is asymmetric (one-to-many) and often loses granular clinical specificity. |
| When to pick each option | Select for EHR documentation, point-of-care problem lists, CDS logic, and standardized phenotype cohorts in OMOP Common Data Model registries. | Select for statutory administrative billing, insurance claims processing (e.g., HIPAA transactions), and public health mortality reporting. |
| Source | SNOMED International (What is SNOMED CT?, verified 2026-11-07); US NLM (SNOMED CT to ICD-10-CM Map Refset, verified 2026-11-07). | World Health Organization (ICD Classification Standards, verified 2026-11-07); CDC NCHS / CMS (ICD-10-CM Guidelines, verified 2026-11-07). |
Common questions
Common questions about SNOMED CT vs ICD-10
What is the primary difference between SNOMED CT and ICD-10?
+
The primary difference is between a clinical reference terminology and an administrative classification. SNOMED CT provides high clinical granularity with over 350,000 poly-hierarchical concepts designed for clinicians entering data at the point of care in EHRs. ICD-10 is a classification system designed to aggregate disease data into broad, mutually exclusive categories for statistical reporting, epidemiology, and insurance reimbursement.
How does LOINC relate to SNOMED CT and ICD-10?
+
LOINC (Logical Observation Identifiers Names and Codes) and SNOMED CT serve complementary functions in health data exchange: LOINC identifies the 'question' or observation (such as a laboratory test order or vital sign measurement), whereas SNOMED CT represents the 'answer' or clinical finding (such as a specific pathogen or diagnosis). In August 2023, SNOMED International and Regenstrief Institute formalized a partnership linking LOINC and SNOMED CT. ICD-10 does not represent laboratory observation orders, but classifies the resulting diagnostic condition.
Can SNOMED CT codes be automatically mapped to ICD-10?
+
Yes. The US National Library of Medicine (NLM), in collaboration with SNOMED International and WHO, maintains the official SNOMED CT to ICD-10-CM Map (Refset 6011000124106). The interactive I-MAGIC (Interactive Map-Assisted Generation of ICD Codes) algorithm uses clinical heuristics to translate granular SNOMED CT concept entries into compliant administrative ICD-10 billing codes.
