omopgenerics

  1. Overview
    1. Key Features:
  2. Core Data Classes
    1. 1. <cdm_reference>
    2. 2. <cohort_table>
    3. 3. <summarised_result>
    4. 4. <codelist> & <concept_set>
  3. Installation
  4. Example Usage
  5. Core Generic Functions Reference

The omopgenerics package provides the fundamental data structures, S3 classes, validation systems, and generic methods that ensure seamless interoperability across the entire OHDSI and DARWIN-EU R software ecosystem.

Overview

In modular analytical pipelines, different packages must agree on standard data structures. omopgenerics acts as the shared foundation across all packages (such as CohortConstructor, PatientProfiles, CohortCharacteristics, and visOmopResults), defining standardized classes for databases, cohort tables, and study results.

Key Features:

  • Formal S3 Class Definitions: Standardized objects for CDM references, cohort tables, and analytical results.
  • Unified Result Specification (<summarised_result>): Standard 13-column schema for all epidemiological estimates, counts, and summaries.
  • Rigorous Class Validation: Automated checks verifying table column names, data types, and OMOP integrity rules.
  • Export & Import Pipelines: Serializes analysis results into standardized CSV archives for federated data sharing.

Core Data Classes

1. <cdm_reference>

Represents an active connection to an OMOP CDM database containing database metadata and pointers to <cdm_table> objects.

2. <cohort_table>

A specialized database table containing patient cohorts. Every <cohort_table> must contain at least four mandatory columns:

  • cohort_definition_id (integer): Unique numeric identifier for the cohort definition.
  • subject_id (integer): Unique identifier of the patient (person_id).
  • cohort_start_date (date): Date of cohort entry (index date).
  • cohort_end_date (date): Date of cohort exit.

Cohort tables carry attached metadata accessible via settings(), attrition(), and cohortCount().

3. <summarised_result>

The universal output format for OHDSI analytics packages, containing a standardized 13-column structure: cdm_name, group_name, group_level, strata_name, strata_level, variable_name, variable_level, estimate_name, estimate_type, estimate_value, additional_name, additional_level, package_name, package_version.

4. <codelist> & <concept_set>

Named lists of standard concept IDs representing clinical definitions, validated and convertible across formats.

Installation 

install.packages("omopgenerics")

Example Usage 

library(omopgenerics)
library(dplyr)

# Inspect cohort metadata
settings(cdm$my_cohort)
attrition(cdm$my_cohort)
cohortCount(cdm$my_cohort)

# Export summarised results for federated study sharing
exportSummarisedResult(
  result = my_analysis_results,
  path = "study_results.csv"
)

# Import results from study partner
partner_results <- importSummarisedResult(
  path = "study_results.csv"
)

Core Generic Functions Reference

Function Purpose
cohortCount(cohort) Returns record and subject counts for each cohort in a <cohort_table>.
attrition(cohort) Retrieves the step-by-step attrition table detailing inclusions and exclusions.
settings(cohort) Returns the metadata and configuration parameters used to construct the cohort.
validateCohortTable(cohort) Verifies that a table meets all formal requirements of an OMOP cohort table.
validateSummarisedResult(result) Validates that an analysis output strictly adheres to the 13-column schema.
splitGroup(), splitStrata() Splits grouped and stratified key-value columns into separate tidy columns.
pivotEstimates() Reshapes long <summarised_result> estimate columns into a wide tidy data frame.
exportSummarisedResult(result, path) Exports validated results to standard compressed or CSV files.
importSummarisedResult(path) Imports and validates results generated by other network study sites.