Examples - EncodeBox

Runnable patterns for FDA Sentinel CDM with the sample database and the unified query API.

Prerequisites

cd /path/to/encodebox
pixi install -e dev
pixi run db-up

Run examples:

python examples/simple_example.py

Three key steps

1

Register connector

import encodebox as eb

eb.register_connector("connectors/postgres_local_sentinel/connector.yml")

2

Connect → Database

db = eb.connect(
    "postgres",
    host="localhost",
    port=5433,
    database="encodebox_sample",
    username="postgres",
    password="postgres",
    connector="postgres_local_sentinel",
)

3

Query

diabetes = (
    db.diagnosis
    .matching("E11", code_type="10", match="starts_with")
)
df = diabetes.to_df()

Common queries

# Diagnoses
db.diagnosis.matching("E11", code_type="10", match="starts_with")

# Procedures
db.procedure.matching("45378", code_type="C4", match="exactly")

# Medications (dispensing)
db.medication.matching("00002-7510", code_type="ND", match="starts_with")

# Multiple code sets (UNION → code_set)
db.diagnosis.matching(
    [["E11"], ["I10"]],
    code_type="10",
    match=["starts_with", "exactly"],
)

Occurrence and gap

index = (
    db.diagnosis
    .matching("E11", code_type="10", match="starts_with")
    .occurring(at_least=2, gap_days=30)
    .with_index_date()
    .first_per_patient()
)

Cohort study

cohort = (
    eb.Cohort("t2dm", database=db)
    .entry(index, index="first")
    .include(db.demographic.age_at_index(18, 89))
    .include(db.enrollment.covering_index(days_before=365, days_after=0))
    .exclude(db.diagnosis.matching("C", match="starts_with"), window=(-365, 0))
)
print(cohort.attrition())
df = cohort.to_df()

See also: Walkthrough, Query patterns.