OHDSI CDM v5.4 Architecture
OMOP Data Dictionary
Our dataset is fully normalized to the OHDSI OMOP Common Data Model (CDM) version 5.4. Below is a high-level overview of the available clinical domains and their underlying vocabularies.
1. Clinical Data Domains
| OMOP Table | Description | Primary Vocabulary |
|---|---|---|
| person | Patient demographics including year of birth, gender, race, and ethnicity. | Standard |
| condition_occurrence | Records of diagnoses, signs, and symptoms documented in the clinical record. | SNOMED CT |
| drug_exposure | Records of all drugs administered, prescribed, or dispensed. | RxNorm |
| measurement | Structured laboratory tests, vital signs, and quantitative clinical findings. | LOINC |
| procedure_occurrence | Records of activities or processes ordered or carried out on a patient. | SNOMED / CPT4 |
| observation | Clinical facts that do not fall into standard domains (e.g. physical activity, surveys). | Standard / LOINC |
| note | Unstructured clinical free-text (de-identified) such as discharge summaries. | Standard |
2. Health System Data
| OMOP Table | Description |
|---|---|
| location | Physical location of care sites (aggregated to region/state for privacy). |
| care_site | Type of facility (e.g., Inpatient, Outpatient Clinic, Pharmacy). |
| provider | Anonymized healthcare provider specialty classifications. |
Full Schema Documentation
Because we rigorously follow the OMOP v5.4 standard, you can query our dataset using any standard OHDSI tools. For column-level constraints and types, please refer to the official specifications.
View Official OHDSI Schema