Cohort

A Cohort defines a study population: entry EventFrame → include / exclude criteria → attrition → DataFrame.

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))
    .include(other_frame, window=(-365, 0))
    .exclude(db.diagnosis.matching("C", match="starts_with"), window=(-365, 0))
)

print(cohort.attrition())
final = cohort.to_df()

Building blocks

Method	Meaning
`.entry(frame, index="first")`	Starting population + index date (once)
`.include(source, window=…)`	Require criterion / EventFrame
`.exclude(source, window=…)`	Remove patients who meet criterion
`.attrition()`	Remaining / dropped counts per step
`.spec()` / `eb.from_spec(...)`	Serialize / restore
`.to_df()` / `.to_sql()`	Run / inspect

Criteria helpers

Helper	Meaning
`db.demographic.age_at_index(lo, hi)`	Age at index
`db.enrollment.covering_index(days_before, days_after)`	Continuous enrollment around index
`db.demographic.sex(...)` / `.race(...)`	Demographics
`db.death.recorded(window=…)`	Death recorded
`include(frame, window=…)`	Any EventFrame as a patient set

window=(before, after) is days relative to the cohort entry index date.

Getting started

Data model

Guides

API reference

Building blocks

Criteria helpers

​Building blocks

​Criteria helpers

​Related

Building blocks

Criteria helpers

Related