DrugUtilisation contains functions to instantiate and characterise drug cohorts in data mapped to the OMOP Common Data Model. The package supports:
Creation of drug cohorts
Identification of indications for those in a drug cohort
Summarising drug utilisation among a cohort in terms of duration, quantity, and dose
Description of treatment adherence based on proportion of patients covered
Detailing treatment restart and switching after an initial treatment discontinuation
First, we need to create a cdm reference for the data we´ll be using. Here we generate an example with simulated data, but to see how you would set this up for your database please consult the CDMConnector package connection examples.
library(DrugUtilisation)
library(CDMConnector)
library(omopgenerics)
library(dplyr)
cdm <- mockDrugUtilisation(numberIndividual = 100)To generate the cohort of acetaminophen users we will use
generateIngredientCohortSet, concatenating any records with
fewer than 7 days between them. We then filter our cohort records to
only include the first record per person and require that they have at
least 30 days observation in the database prior to their drug start
date.
cdm <- generateIngredientCohortSet(
cdm = cdm,
name = "dus_cohort",
ingredient = "acetaminophen",
gapEra = 7
)
#> Warning: ! `codelist` contains numeric values, they are casted to integers.
cdm$dus_cohort |>
requireIsFirstDrugEntry() |>
requireObservationBeforeDrug(days = 30)
#> # Source: table<main.dus_cohort> [?? x 4]
#> # Database: DuckDB v1.1.0 [root@Darwin 24.0.0:R 4.4.1/:memory:]
#> cohort_definition_id subject_id cohort_start_date cohort_end_date
#> <int> <int> <date> <date>
#> 1 1 1 2021-08-22 2022-01-20
#> 2 1 2 2003-04-08 2006-03-14
#> 3 1 4 1971-04-20 1971-08-01
#> 4 1 5 2010-10-12 2017-10-23
#> 5 1 6 2019-06-04 2019-11-06
#> 6 1 7 2011-11-30 2012-03-22
#> 7 1 8 1993-03-18 1996-08-09
#> 8 1 10 2009-08-21 2010-02-21
#> 9 1 11 2022-03-27 2022-07-13
#> 10 1 14 2010-12-27 2012-10-26
#> # ℹ more rowsNow we´ve created our cohort we could first summarise the indications of the cohort. These indications will always be cohorts, so we first need to create them. Here we create two indication cohorts, one for headache and the other for influenza.
indications <- list(headache = 378253, influenza = 4266367)
cdm <- generateConceptCohortSet(cdm,
conceptSet = indications,
name = "indications_cohort"
)
#> Warning: ! 3 casted column in indications_cohort (cohort_attrition) as do not match
#> expected column type:
#> • `reason_id` from numeric to integer
#> • `excluded_records` from numeric to integer
#> • `excluded_subjects` from numeric to integer
#> Warning: ! 1 casted column in indications_cohort (cohort_codelist) as do not match
#> expected column type:
#> • `concept_id` from numeric to integerWe can summarise the indication results using the
summariseIndication function:
indication_summary <- cdm$dus_cohort |>
summariseIndication(
indicationCohortName = "indications_cohort",
unknownIndicationTable = "condition_occurrence",
indicationWindow = list(c(-30, 0))
)
#> Getting specified indications
#> Creating indication summary variables
#> Getting unknown indications
#> Summarising indication results
indication_summary |> glimpse()
#> Rows: 12
#> Columns: 13
#> $ result_id <int> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1
#> $ cdm_name <chr> "DUS MOCK", "DUS MOCK", "DUS MOCK", "DUS MOCK", "DUS …
#> $ group_name <chr> "cohort_name", "cohort_name", "cohort_name", "cohort_…
#> $ group_level <chr> "161_acetaminophen", "161_acetaminophen", "161_acetam…
#> $ strata_name <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ strata_level <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ variable_name <chr> "number records", "number subjects", "Indication from…
#> $ variable_level <chr> NA, NA, "headache", "headache", "influenza", "influen…
#> $ estimate_name <chr> "count", "count", "count", "percentage", "count", "pe…
#> $ estimate_type <chr> "integer", "integer", "integer", "percentage", "integ…
#> $ estimate_value <chr> "61", "61", "1", "1.63934426229508", "0", "0", "0", "…
#> $ additional_name <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ additional_level <chr> "overall", "overall", "overall", "overall", "overall"…We can quickly obtain a summary of drug utilisation among our cohort, with various measures calculated for a provided ingredient concept (in this case the concept for acetaminophen).
drug_utilisation_summary <- cdm$dus_cohort |>
summariseDrugUtilisation(
ingredientConceptId = 1125315,
gapEra = 7
)
#> Warning: ! `codelist` contains numeric values, they are casted to integers.
drug_utilisation_summary |> glimpse()
#> Rows: 58
#> Columns: 13
#> $ result_id <int> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,…
#> $ cdm_name <chr> "DUS MOCK", "DUS MOCK", "DUS MOCK", "DUS MOCK", "DUS …
#> $ group_name <chr> "cohort_name", "cohort_name", "cohort_name", "cohort_…
#> $ group_level <chr> "161_acetaminophen", "161_acetaminophen", "161_acetam…
#> $ strata_name <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ strata_level <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ variable_name <chr> "number records", "number subjects", "number exposure…
#> $ variable_level <chr> NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, NA, N…
#> $ estimate_name <chr> "count", "count", "q25", "median", "q75", "mean", "sd…
#> $ estimate_type <chr> "integer", "integer", "integer", "integer", "integer"…
#> $ estimate_value <chr> "61", "61", "1", "1", "1", "1.22950819672131", "0.528…
#> $ additional_name <chr> "overall", "overall", "concept_set", "concept_set", "…
#> $ additional_level <chr> "overall", "overall", "ingredient_1125315_descendants…
table(drug_utilisation_summary$variable_name)
#>
#> cumulative dose cumulative quantity exposed time initial daily dose
#> 7 7 7 7
#> initial quantity number eras number exposures number records
#> 7 7 7 1
#> number subjects time to exposure
#> 1 7Now we can combine our results and suppress any counts less than 5 so that they are ready to be shared.
results <- bind(
indication_summary,
drug_utilisation_summary
) |>
suppress(minCellCount = 5)
results |> glimpse()
#> Rows: 70
#> Columns: 13
#> $ result_id <int> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2,…
#> $ cdm_name <chr> "DUS MOCK", "DUS MOCK", "DUS MOCK", "DUS MOCK", "DUS …
#> $ group_name <chr> "cohort_name", "cohort_name", "cohort_name", "cohort_…
#> $ group_level <chr> "161_acetaminophen", "161_acetaminophen", "161_acetam…
#> $ strata_name <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ strata_level <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ variable_name <chr> "number records", "number subjects", "Indication from…
#> $ variable_level <chr> NA, NA, "headache", "headache", "influenza", "influen…
#> $ estimate_name <chr> "count", "count", "count", "percentage", "count", "pe…
#> $ estimate_type <chr> "integer", "integer", "integer", "percentage", "integ…
#> $ estimate_value <chr> "61", "61", NA, NA, "0", "0", "0", "0", NA, NA, "57",…
#> $ additional_name <chr> "overall", "overall", "overall", "overall", "overall"…
#> $ additional_level <chr> "overall", "overall", "overall", "overall", "overall"…There are many more drug-related analyses that we could have done with this acetaminophen cohort using the DrugUtilisation package. Please see the package website for more details.