OmopSketch

  1. Overview
    1. Key Features:
  2. Installation
  3. Getting Started
  4. Core API Reference
    1. Database Snapshot & Overview
    2. Observation Period & Temporal Coverage
    3. Clinical Table Quality & Concept Profiling

The OmopSketch package provides fast, automated tools for high-level characterization and exploratory data analysis of OMOP Common Data Model (CDM) databases.

Overview

Before launching an observational study, researchers must understand the shape, size, historical coverage, and data quality of their OMOP database. OmopSketch generates rapid summaries (“sketches”) of database demographics, table volumes, observation periods, and concept usage patterns.

Key Features:

  • Instant Snapshot: Generates an executive summary of vocabulary versions, patient counts, and table sizes.
  • Observation Period Dynamics: Evaluates longitudinal coverage, continuous observation trends, and seasonal gaps.
  • Clinical Table Profiling: Analyzes record density, missingness, and mapping proportions across OMOP clinical domains.
  • Interactive Shiny Explorer: Launches a self-contained Shiny dashboard for visual exploration of database characteristics.

Installation 

# Install from CRAN
install.packages("OmopSketch")

# Or development version from GitHub
# pak::pkg_install("ohdsi/OmopSketch")

Getting Started 

library(OmopSketch)
library(CDMConnector)

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

# 1. Summarize database snapshot
snapshot <- summariseOmopSnapshot(cdm)
tableOmopSnapshot(snapshot)

# 2. Analyze observation period coverage
obs_summary <- summariseObservationPeriod(cdm$observation_period)
plotObservationPeriod(obs_summary)

# 3. Analyze clinical records in condition_occurrence
condition_profile <- summariseClinicalRecords(
  cdm = cdm,
  omopTableName = "condition_occurrence"
)

Core API Reference

Database Snapshot & Overview

Function Description
summariseOmopSnapshot(cdm) Computes database metadata, vocabulary release, total patient counts, and table row counts.
tableOmopSnapshot(result) Renders a formatted publication-ready table of the database snapshot.

Observation Period & Temporal Coverage

Function Description
summariseObservationPeriod(observationPeriod) Analyzes observation duration, gaps, and records per person.
summariseInObservation(observationPeriod, interval) Tracks population in-observation trends over years, quarters, or months.
plotObservationPeriod(result) Plots distributions of observation time.
plotInObservation(result) Generates temporal trend curves of actively observed patients.

Clinical Table Quality & Concept Profiling

Function Description
summariseMissingData(cdm, omopTableName) Evaluates column completeness and zero/unmapped concept ID frequencies.
summariseClinicalRecords(cdm, omopTableName) Computes records per patient, standard vs non-standard concept mappings, and domain alignment.
summariseRecordCount(cdm, omopTableName, interval) Tracks longitudinal event recording trends across calendar intervals.
summariseConceptIdCounts(cdm, omopTableName) Computes record and patient counts per individual clinical concept ID.
tableTopConceptCounts(result, top) Displays the most frequently recorded concept IDs in a clinical table.