Use Cases
Practical examples and workflows for clinical terminology lookup.
These examples show how to use Clinical Terminology MCP servers in real-world scenarios with your AI assistant.
Clinical Documentation
Finding diagnosis codes
“Find the ICD-10 code for type 2 diabetes with diabetic nephropathy”
Your AI assistant will use icd10_search to find matching codes, returning results like:
- E11.21 — Type 2 diabetes mellitus with diabetic nephropathy
Looking up procedure codes
“What is SNOMED code 80146002?”
Your AI assistant uses snomed_lookup to return:
- 80146002 | Appendectomy (procedure) |
Drug Information
Searching for medications
“Search RxNorm for metformin extended release”
Your AI assistant uses rxnorm_search to find:
- Metformin Hydrochloride 500 MG Extended Release Oral Tablet
- Metformin Hydrochloride 750 MG Extended Release Oral Tablet
- And other matching formulations
Drug classification
“What is the ATC classification for metformin?”
Your AI assistant uses rxnorm_atc to return the Anatomical Therapeutic Chemical classification:
- A10BA — Biguanides (oral blood glucose lowering drugs)
Laboratory Values
Finding lab codes
“What’s the LOINC code for hemoglobin A1c?”
Your AI assistant uses loinc_search to find:
- 4548-4 — Hemoglobin A1c/Hemoglobin.total in Blood
Understanding lab panels
“What tests are included in a basic metabolic panel?”
Your AI assistant can search LOINC and SNOMED to identify:
- Glucose, BUN, Creatinine, Sodium, Potassium, Chloride, CO2
Terminology Mapping
Cross-terminology lookup
“Map SNOMED code 73211009 (Diabetes mellitus) to ICD-10”
Your AI assistant uses umls_crosswalk to find:
- SNOMED 73211009 → ICD-10-CM E11.9 (Type 2 diabetes mellitus without complications)
Concept relationships
“What are the subtypes of hypertension in SNOMED?”
Your AI assistant uses snomed_hierarchy to explore:
- Essential hypertension
- Secondary hypertension
- Malignant hypertension
- Pulmonary hypertension
- And other descendants
Unit Conversion
Converting clinical measurements
“Convert 98.6 degrees Fahrenheit to Celsius”
Your AI assistant uses ucum_convert to return:
- 37.0 °C
Validating unit expressions
“Is mg/dL a valid UCUM unit?”
Your AI assistant uses ucum_validate to confirm the unit is valid according to UCUM specification.
ECL Queries (Advanced)
SNOMED Expression Constraint Language (ECL) enables powerful queries.
Finding all descendants
“Find all types of pneumonia in SNOMED using ECL”
Your AI assistant uses snomed_ecl with the query << 233604007 (descendants of Pneumonia):
- Bacterial pneumonia
- Viral pneumonia
- Aspiration pneumonia
- Hospital-acquired pneumonia
- And hundreds more specific types
Finding related concepts
“Find all procedures that involve the heart”
Your AI assistant uses snomed_ecl with a query like:
<< 71388002 : 363704007 = << 80891009
(Procedures with a finding site of heart structure)
Healthcare Application Development
Building decision support
Developers can use these servers to:
- Validate diagnosis codes entered by users
- Suggest related or more specific codes
- Map between different coding systems
- Convert units for lab value comparisons
EHR integration prompts
Example prompts for EHR-related tasks:
“Given this patient’s conditions [list], suggest relevant SNOMED codes”
“Convert these lab results to standard LOINC codes for interoperability”
“Map this insurance claim’s ICD-10 codes to SNOMED for clinical use”
Tips for Effective Queries
- Be specific: “Type 2 diabetes” works better than just “diabetes”
- Use standard terms: Clinical terminology searches work best with standard medical language
- Specify the code system: If you need a specific terminology (ICD-10 vs SNOMED), say so
- Request details: Ask for code, description, and relationships when needed
- Combine tools: Your AI assistant can use multiple terminology servers in one response
Limitations
- Search results depend on the underlying terminology service
- Some code systems require valid credentials
- ECL queries are SNOMED-specific
- Not all mappings exist between terminologies