OmopConstructor

1. Overview
2. Key Features
3. Installation
4. Usage Example
5. Main Functions

The OmopConstructor R package provides functionality to construct and standardize derived tables in the Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM), such as observation_period, drug_era, and condition_era.

---

Overview

In OMOP CDM architectures, certain derived tables represent consolidated chronological intervals synthesized from raw clinical encounter tables. For instance, the observation_period table defines the continuous time span during which a patient’s healthcare interactions are reliably captured by the database.

OmopConstructor automates the derivation and recalculation of these tables directly against database connections, allowing researchers and data engineers to establish custom observation boundaries, collapse gaps, and enforce study-specific observation rules.

---

Key Features

• Automated Observation Period Construction: Synthesizes observation periods from clinical activity across visits, conditions, drugs, procedures, and measurements.
• Customizable Persistence & Gaps: Configurable allowable gap days (collapseDays, persistenceDays) to bridge intermittent observation breaks.
• Demographic & Mortality Censoring: Automatically caps observation periods at recorded death dates, maximum plausible age (e.g., censorAge = 120), or extraction date ceilings.
• Multi-DBMS Compatibility: Validated across PostgreSQL, Snowflake, Amazon Redshift, Microsoft SQL Server, and DuckDB.
• Containerized ETL: Includes Docker and CLI workflows for reproducible, headless batch construction.

---

Installation

Install OmopConstructor from CRAN or GitHub:

# From CRAN
install.packages("OmopConstructor")

# Development version from GitHub
# install.packages("pak")
pak::pkg_install("ohdsi/OmopConstructor")

---

Usage Example

The following example demonstrates constructing a tailored observation_period table in an OMOP CDM reference using DuckDB:

library(CDMConnector)
library(OmopConstructor)
library(OmopSketch)
library(dplyr)

# 0. Connect to OMOP CDM database
con <- DBI::dbConnect(duckdb::duckdb(), eunomiaDir("GiBleed"))
cdm <- cdmFromCon(con, cdmSchema = "main", writeSchema = "main")

# 1. Build standardized observation periods
cdm <- buildObservationPeriod(
  cdm = cdm,
  collapseDays = 30,
  persistenceDays = 30,
  dateRange = as.Date(c("1950-01-01", "2025-12-31")),
  censorAge = 120
)

# 2. Inspect generated observation_period table
cdm$observation_period |>
  glimpse()

# 3. Summarise and validate observation periods with OmopSketch
obs_summary <- summariseObservationPeriod(
  observationPeriod = cdm$observation_period
)

tableObservationPeriod(
  result = obs_summary,
  type = "gt"
)

---

Main Functions

Function	Purpose
buildObservationPeriod()	Constructs or recalculates the observation_period table from clinical event tables with customizable collapse rules, date ceilings, and censoring criteria.