Drug Exposure

Epic equivalent: Medication orders / MAR (Medication Administration Record) / Prescription history / Pharmacy data

The drug_exposure table captures every medication event — inpatient administrations, outpatient prescriptions, pharmacy dispenses, and self-reported medications. In Epic, this data is scattered across medication orders, the MAR, prescription history, and pharmacy modules. In OMOP, it all lands here.

Each row is a single medication exposure event. Medications are coded in RxNorm (mapped from NDC, GPI, or local codes).

Epic-to-OMOP Field Mapping

Field reference (click to expand)

OMOP Field	Epic Equivalent	What It Captures
drug_exposure_id	Medication record ID	Unique identifier for the exposure event
person_id	Patient ID / MRN	Links to the patient
drug_concept_id	RxNorm standard concept	Standardized drug (ingredient, clinical drug, or branded). Always use for analysis
drug_exposure_start_date	Order/admin start date	When the medication exposure began
drug_exposure_end_date	Order/admin end date	When it ended. Often null — may need to infer from days_supply
drug_type_concept_id	Data provenance	Whether from prescription, dispensing, administration, etc.
stop_reason	Discontinuation reason	Why the medication was stopped (if documented)
refills	Number of refills	Refill count authorized
quantity	Quantity dispensed	Amount dispensed or administered
days_supply	Days supply	Intended duration of the supply
route_concept_id	Route of administration	Oral, IV, subcutaneous, etc.
sig	Prescription instructions	Free-text dosing instructions
visit_occurrence_id	Linked encounter	Visit during which the drug was ordered/given
drug_source_value	NDC / local code	Original code from the source system

What to Watch For

Common pitfalls

Many rows per medication per patient

Each refill, each MAR administration, each separate prescription generates a row. Use drug_era to collapse into continuous exposure periods.

Null end dates are common

drug_exposure_end_date is often missing. You may need to calculate it as start_date + days_supply.

Route and sig are often sparse

Not all source systems capture route_concept_id or sig consistently.

Ingredient-level analysis

For drug class analyses (e.g., "all ACE inhibitors"), query at the RxNorm ingredient level using concept_ancestor.

Research Patterns

Question	Tables Involved
ACE inhibitor use among hypertension patients	drug_exposure (ACE ingredient descendants) + condition_occurrence (hypertension)
Duration of endocrine therapy in breast cancer	drug_exposure + condition_occurrence (breast cancer)
Statin prescribing equity across racial groups	drug_exposure + person.race_concept_id
Antibiotics given in the ER	drug_exposure + visit_occurrence filtered by ER
Chemotherapy cycle detection	drug_exposure (chemo drugs) grouped by person_id + date intervals