Search for medical concepts across OHDSI standardized vocabularies by name, synonym, or clinical term. Returns matching concepts with IDs, names, vocabulary, domain, and standard status. Use this when you need to find the OMOP concept ID for a medical term. Examples: 'type 2 diabetes', 'metformin 500mg', 'systolic blood pressure', 'HbA1c'.

Search for medical concepts using natural language with neural embeddings. Unlike keyword search, semantic search understands clinical meaning — 'heart attack' finds 'Myocardial infarction', 'high blood sugar' finds 'Hyperglycemia'. Returns concepts ranked by similarity score. Use this when the user describes symptoms, conditions, or treatments in everyday language rather than exact medical terminology.

Find medical concepts similar to a reference concept, name, or natural language query. Supports three algorithms: 'semantic' (neural embeddings — best for meaning), 'lexical' (text matching — best for typos), 'hybrid' (combined — default). Provide exactly ONE of: concept_id, concept_name, or query. Use this to explore related concepts, find alternative codes, or build phenotype concept sets. Tip: For drug vocabularies like RxNorm, use drug class names ('ACE inhibitors', 'beta blockers', 'antihypertensives') rather than symptom descriptions ('medications for high blood pressure') — the embedding model aligns better with clinical terminology than lay language.

Get detailed information about a specific OMOP concept by its numeric concept_id. Returns the concept name, vocabulary, domain, concept class, standard status, valid dates, and synonyms. Use this when you already have a concept_id and need its details.

Look up an OMOP concept using a vocabulary-specific code and vocabulary ID. Both parameters are required to avoid ambiguity - the same code can exist in multiple vocabularies (e.g., 'E11' exists in both ICD10CM and ICD10). If multiple concepts share the same code within a vocabulary, all matches are returned - prefer the one with standard_concept='S'.

Get a comprehensive view of a medical concept in one call: detailed info, ancestors/descendants hierarchy, and cross-vocabulary mappings. Use this instead of calling get_concept + get_hierarchy + map_concept separately. Ideal for understanding what a concept is, where it sits in the classification tree, and how it maps to other coding systems.

Find mappings FROM a source concept TO equivalent concepts in other vocabularies. The concept_id you provide is always the SOURCE — results show what it maps TO. Returns cross-vocabulary mappings with relationship types and mapping quality. If no mappings exist, the response explicitly states 'No mappings found' with mapped=false in JSON — never returns ambiguous empty results. Example: provide a SNOMED concept_id and filter by target_vocabularies='ICD10CM' to get the ICD-10 equivalent.

Navigate the vocabulary hierarchy for a concept. Use direction='up' for ancestors (broader terms like 'Diabetes mellitus' → 'Metabolic disease'), direction='down' for descendants (narrower terms, essential for building concept sets in phenotype definitions), or direction='both' for full hierarchical context. Results are capped at max_results nodes (default 500). If truncated, the response will indicate total available count so you can narrow with vocabulary_ids or reduce max_levels.

List all available medical vocabularies in the OMOP standardized vocabulary system with concept counts and metadata. Use this to understand what terminology systems are available (SNOMED CT, ICD-10-CM, RxNorm, LOINC, etc.) and their scope.

Resolve a FHIR coded value (system URI + code) to its OMOP standard concept and CDM target table. Supports text-only input via semantic search fallback and optional Phoebe recommendations. Examples: system='http://snomed.info/sct' code='44054006' for Type 2 diabetes, or display='heart attack' for text-only resolution.

Resolve a FHIR CodeableConcept with multiple codings. Picks the best match per OHDSI vocabulary preference (SNOMED > RxNorm > LOINC > CVX > ICD-10). Falls back to the text field via semantic search if no coding resolves.