Health Economics and Outcomes Research Pipeline on OMOP CDM • omopHeor

omopHeor is an R-based analytical ecosystem developed by IOMED for Real-World Evidence (RWE) and Health Economics and Outcomes Research (HEOR). It streamlines the end-to-end process of Healthcare Resource Utilization (HCRU) analysis, direct medical costing, and Cost-Effectiveness Analysis (CEA) directly on observational healthcare data structured in the OMOP Common Data Model (CDM).

Key Features

OMOP CDM Native: Connects directly to OMOP CDM databases via CDMConnector and omopgenerics.
Modular Monorepo Architecture: Decouples lightweight HCRU extraction (CohortUtilisation), direct costing (CohortCosts), and HEOR modeling (CohortEconomics) under a unified metapackage (omopHeor).
3-Layer Care Utilization Suite: Episode constructors, in-database column enrichers, and standardized GT / Flextable reporting.
Direct Medical Cost Extraction: Polymorphic OMOP COST table linkage across visits, drugs, and procedures with automated zero-fill handling.
End-to-End Decision Science: Propensity score adjustment (Cyclops), Markov state-transition models, and CEA visualizations (BCEA).
Instant Prototyping: Includes mockOmopHeor() providing an in-memory synthetic DuckDB OMOP CDM database.

Getting Started

Prerequisites

R (>= 4.1.0)
pak package manager (install.packages("pak"))

Installation

# Install the complete omopHeor metapackage (recommended)
pak::pkg_install("iomedhealth/omopHeor")

# Or install individual standalone domain packages
pak::pkg_install("iomedhealth/omopHeor/packages/CohortUtilisation")
pak::pkg_install("iomedhealth/omopHeor/packages/CohortCosts")
pak::pkg_install("iomedhealth/omopHeor/packages/CohortEconomics")

Quick Start 1: In-Database Cohort Utilization & Cost Enrichment

Enrich study cohorts in-database across configurable temporal windows (e.g., baseline [-365, -1], follow-up [0, 365], or full follow-up [0, Inf]) and produce publication tables:

library(omopHeor)
library(dplyr)

# 0. Connect to CDM (built-in synthetic mock database)
cdm <- mockOmopHeor()

# 1. Enrich cohort with visits, prescriptions, and direct costs in-database
cdm$study_enriched <- cdm$target_cohort |>
  addVisits(
    window = list(baseline = c(-365, -1), followup = c(0, 365)),
    settings = c("inpatient", "outpatient", "emergency"),
    stratifySpecialty = TRUE,
    readmissions = TRUE
  ) |>
  addPrescriptions(
    window = list(followup = c(0, 365)),
    daysSupply = TRUE,
    pdc = TRUE
  ) |>
  addCosts(
    window = list(followup = c(0, 365)),
    costField = "total_paid",
    name = "study_enriched"
  )

# 2. Summarise utilization and costs into standardised results
util_summary <- summariseUtilization(cdm$study_enriched)
cost_summary <- summariseCosts(cdm$study_enriched)

# 3. Format publication tables (GT, Flextable, or Tibble)
tableUtilization(util_summary)
tableCosts(cost_summary)

Quick Start 2: 6-Stage HEOR Causal & Decision-Analytic Simulation

Run the complete causal inference, Markov state-transition modeling, and economic simulation pipeline:

library(omopHeor)

# 0. Setup Connection
cdm <- mockOmopHeor()

# 1-6. Run the End-to-End Pipeline
study <- init(
  cdm = cdm,
  target_cohort = "target_cohort",
  comparator_cohort = "comparator_cohort",
  outcome_cohort = "outcome_cohort"
) |>
  summarise_baseline() |>
  extract_hcru() |>
  fit_ps() |>
  adjust_ps() |>
  compile_trajectories() |>
  simulate_economics(time_horizon = 10, n_samples = 100) |>
  run_cea()

# Decision-Analytic Visualizations & Summary
plot_ceac(study)
plot_plane(study)
table_summary(study)

Outputs

Cost-Effectiveness Plane Cost-Effectiveness Acceptability Curve

Documentation & Guides

For in-depth architecture, tutorials, and technical specifications, explore the documentation articles:

The omopHeor Ecosystem & Modular Architecture: Comprehensive breakdown of CohortUtilisation, CohortCosts, CohortEconomics, technology stack, and package design.
Introduction to HEOR for OMOP Users: Conceptual guide translating OMOP CDM vocabulary to Health Economics (ICER, QALY, Markov health states).
Cohort Utilization & Cost Enrichment: Step-by-step hands-on tutorial for the 3-layer in-database cohort enrichers.
HCRU & Direct Cost Extraction Logic: Deep dive into OMOP COST polymorphic linkage, schema checks, and zero-fill fallback rules.

Contributing

We welcome contributions! Please review .github/CONTRIBUTING.md and AGENTS.md for coding standards, styling conventions, and test procedures.

License

This project is licensed under the MIT License - see the LICENSE file for details.

omopHeor HEOR & HCRU on OMOP