Patient-Level Indices from the OMOP Common Data Model

Provides tools to derive standardised, reproducible patient-level indices and covariates from Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) databases. Functions calculate comorbidity and frailty scores, including the Charlson Comorbidity Index, Electronic Frailty Index, and Hospital Frailty Risk Score, as well as body mass index, polypharmacy, ethnicity, location, and socioeconomic status measures.

OmopIndices

The goal of OmopIndices is to enable standardised and reproducible derivation of clinically and epidemiologically relevant patient-level indices and covariates directly from OMOP CDM database instances.

Ecosystem

OmopIndices is part of the ecosystem of packages defined by omopgenerics. For more details on the ecosystem you can read the Tidy R programming with the OMOP Common Data Model book.

Tested sources

Source	Driver	CDM reference	Status
Local R data frame	N/A	omopgenerics::cdmFromTables()
In-memory DuckDB database	duckdb	CDMConnector::cdmFromCon()

Installation

You can install the CRAN version of OmopIndices as:

install.packages("OmopIndices")

Or you can install the development version of OmopIndices from GitHub with:

# install.packages("pak")
pak::pkg_install("OHDSI/OmopIndices")

Main functionality

OmopIndices adds patient-level measures to an existing cdm_table. Each function returns the input table with one or more columns added, so functions can be composed with the pipe operator. The input table must contain person_id or subject_id and, for index-based measures, a Date column that identifies the index date.

The index functions use OMOP concepts from the supplied conceptSet to find records in the requested window. By default, the package uses internal concept sets that can be inspected using getIndexCodelist(). A concept set can also be supplied as a codelist, codelist_with_details, or concept_set_expression object.

Examples

The examples below use the GiBleed database bundled with the omock package. Each group starts from the same mock cohort, and the examples can be adapted to any compatible OMOP CDM table. For index-based measures, the default index date is cohort_start_date; use indexDate to select another Date column. Windows are expressed in days relative to the index date, so c(-365, 0) means the year before and including the index date.

library(omock)
library(duckdb)
library(OmopIndices)
library(dplyr)
library(CohortConstructor)

cdm <- mockCdmFromDataset(datasetName = "GiBleed", source = "duckdb")

# Let's create a simple sinusitis cohort for the examples
cdm$cohort <- conceptCohort(
  cdm = cdm,
  conceptSet = list(sinusitis = c(257012L, 4283893L, 4294548L, 40481087L)),
  name = "cohort"
)

Comorbidity and frailty

The comorbidity and frailty functions add a numeric score. The Charlson and updated Charlson functions can optionally include age adjustment, and the frailty scores can also add a categorical classification of the score. The package provides default categories for the Electronic Frailty Index and Hospital Frailty Risk Score; these can be overridden with categories.

comorbidity <- cdm$cohort |>
  addCharlsonIndex(ageAdjusted = TRUE) |>
  addUpdatedCharlsonIndex(ageAdjusted = TRUE) |>
  addElectronicFrailtyIndex() |>
  addHospitalFrailtyRiskScore(
    categories = list(
      low = c(0, 5),
      intermediate = c(5, 15),
      high = c(15, Inf)
    )
  )

comorbidity |>
  select(subject_id, cohort_start_date, charlson, updated_charlson, efi,
         efi_categories, hfrs, hfrs_categories) |>
  glimpse()
#> Rows: ??
#> Columns: 8
#> Database: DuckDB 1.5.5 [root@Darwin 25.6.0:R 4.4.1//private/var/folders/pl/k11lm9710hlgl02nvzx4z9wr0000gp/T/Rtmp5pfEVs/file15da6d47cdf9.duckdb]
#> $ subject_id        <int> 2859, 5165, 1250, 3241, 1381, 4403, 1945, 5309, 4521…
#> $ cohort_start_date <date> 1982-07-04, 2010-11-09, 1994-01-09, 2002-08-02, 201…
#> $ charlson          <dbl> 1, 0, 0, 2, 2, 4, 0, 0, 5, 0, 1, 3, 0, 0, 0, 0, 1, 0…
#> $ updated_charlson  <dbl> 0, 0, 0, 0, 1, 3, 0, 0, 6, 0, 0, 2, 0, 0, 0, 0, 0, 0…
#> $ efi               <dbl> 0.08333333, 0.11111111, 0.11111111, 0.11111111, 0.11…
#> $ efi_categories    <chr> "fit", "fit", "fit", "fit", "fit", "fit", "fit", "fi…
#> $ hfrs              <dbl> 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.5, 0.0, 11.1, 0.0, 0…
#> $ hfrs_categories   <chr> "low", "low", "low", "low", "low", "low", "low", "lo…

Each index uses an internal concept set by default. To apply a study-specific definition, supply a named conceptSet containing the required codelists. The available internal codelists can be inspected with getIndexCodelist().

getIndexCodelist("charlson")
#> 
#> - aids (11 codes)
#> - any_malignancy (716 codes)
#> - cerebrovascular_disease (129 codes)
#> - chronic_pulmonary_disease (92 codes)
#> - congestive_heart_failure (41 codes)
#> - connective_tissue_disease (60 codes)
#> along with 11 more codelists

Clinical covariates

addBMI() selects a BMI measurement from a time window, while addPolypharmacyCount() calculates the maximum number of simultaneous drug ingredients in its window. Here BMI is taken from the last measurement in the preceding year and polypharmacy is assessed over the preceding 30 days.

clinical <- cdm$cohort |>
  addBMI(window = c(-365, 0), order = "last") |>
  addPolypharmacyCount(window = c(-30, 0))

clinical |>
  select(subject_id, cohort_start_date, bmi, polypharmacy_count) |>
  glimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB 1.5.5 [root@Darwin 25.6.0:R 4.4.1//private/var/folders/pl/k11lm9710hlgl02nvzx4z9wr0000gp/T/Rtmp5pfEVs/file15da6d47cdf9.duckdb]
#> $ subject_id         <int> 2859, 5165, 3241, 1381, 4403, 1242, 3083, 410, 5276…
#> $ cohort_start_date  <date> 1982-07-04, 2010-11-09, 2002-08-02, 2010-08-08, 19…
#> $ bmi                <int> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA,…
#> $ polypharmacy_count <int> 2, 2, 2, 2, 2, 2, 2, 2, 3, 2, 2, 2, 3, 2, 1, 2, 2, …

Demographics

addEthnicity() uses available ethnicity and race fields from the OMOP person table in sequence. addLocation() first uses location_id and then falls back to the location associated with care_site_id. Missing values can be replaced with study-specific labels.

demographics <- cdm$cohort |>
  addEthnicity(missingEthnicityValue = "Unknown") |>
  addLocation(
    from = c("location_id", "care_site_id"),
    locationSource = "location_source_value",
    missingLocationValue = "Unknown"
  )

demographics |>
  select(subject_id, cohort_start_date, ethnicity, location) |>
  glimpse()
#> Rows: ??
#> Columns: 4
#> Database: DuckDB 1.5.5 [root@Darwin 25.6.0:R 4.4.1//private/var/folders/pl/k11lm9710hlgl02nvzx4z9wr0000gp/T/Rtmp5pfEVs/file15da6d47cdf9.duckdb]
#> $ subject_id        <int> 2859, 5165, 3241, 1381, 4403, 1242, 3083, 410, 5276,…
#> $ cohort_start_date <date> 1982-07-04, 2010-11-09, 2002-08-02, 2010-08-08, 199…
#> $ ethnicity         <chr> "Unknown", "Unknown", "Unknown", "Unknown", "Unknown…
#> $ location          <chr> "Unknown", "Unknown", "Unknown", "Unknown", "Unknown…

Socioeconomic status

Socioeconomic status can be derived from Townsend deprivation scores, the Index of Multiple Deprivation (IMD), or a prioritised combination of both. The following example retains the two source-specific outputs and also creates a combined value that prefers IMD and falls back to Townsend.

socioeconomic <- cdm$cohort |>
  addTownsend(nameStyle = "townsend") |>
  addIndexOfMultipleDeprivation(nameStyle = "imd") |>
  addSocioEconomicStatus(
    from = c("imd", "townsend"),
    nameStyle = "socio_economic_status"
  )

socioeconomic |>
  select(subject_id, cohort_start_date, townsend, imd,
         socio_economic_status) |>
  glimpse()
#> Rows: ??
#> Columns: 5
#> Database: DuckDB 1.5.5 [root@Darwin 25.6.0:R 4.4.1//private/var/folders/pl/k11lm9710hlgl02nvzx4z9wr0000gp/T/Rtmp5pfEVs/file15da6d47cdf9.duckdb]
#> $ subject_id            <int> 2859, 5165, 3241, 1381, 4403, 1242, 3083, 410, 5…
#> $ cohort_start_date     <date> 1982-07-04, 2010-11-09, 2002-08-02, 2010-08-08,…
#> $ townsend              <chr> "Missing", "Missing", "Missing", "Missing", "Mis…
#> $ imd                   <chr> "Missing", "Missing", "Missing", "Missing", "Mis…
#> $ socio_economic_status <chr> "Missing", "Missing", "Missing", "Missing", "Mis…

The default codelists and output names are suitable for exploratory use, but for a reproducible analysis you should specify the concept set, date window, selection order, and output name required by the study protocol.

Function reference

See the reference site for complete argument descriptions, accepted input types, and examples for every exported function. The package is tested with local data frames and DuckDB in continuous integration. It is designed to work with database-backed OMOP CDM sources supported by omopgenerics and CDMConnector; verify any additional backend in the target deployment environment.