The goal of DrugExposureDiagnostics is to summarise ingredient specific drug exposure data in the OMOP CDM.
You can install the DrugExposureDiagnostics from CRAN like this:
or install the development version:
In order to use the example data from the eunomia dataset, you need to set an environment variable. You can do it like this:
{r} usethis::edit_r_environ()
Let´s look at the ingredient acetaminophen (https://athena.ohdsi.org/search-terms/terms/1125315).
We can run all the checks available in ´DrugExposureDiagnostics´ using the ´executeChecks´ function.
all_checks <- executeChecks(cdm = cdm,
ingredients = 1125315,
checks = c("missing", "exposureDuration", "type", "route", "sourceConcept", "daysSupply", "verbatimEndDate",
"dose", "sig", "quantity", "histogram", "diagnosticsSummary"))
#> population after earliestStartDate smaller than sample
#> Joining with `by = join_by(ingredient_concept_id)`
#> Joining with `by = join_by(ingredient_concept_id)`
The output is a list which contains the following set of tibbles:
names(all_checks)
#> [1] "conceptSummary" "missingValuesOverall"
#> [3] "missingValuesByConcept" "drugExposureDurationOverall"
#> [5] "drugExposureDurationByConcept" "drugTypesOverall"
#> [7] "drugTypesByConcept" "drugRoutesOverall"
#> [9] "drugRoutesByConcept" "drugSourceConceptsOverall"
#> [11] "drugSourceConceptsByConcept" "drugDaysSupply"
#> [13] "drugDaysSupplyByConcept" "drugVerbatimEndDate"
#> [15] "drugVerbatimEndDateByConcept" "drugDose"
#> [17] "drugDoseByConcept" "drugSig"
#> [19] "drugSigByConcept" "drugQuantity"
#> [21] "drugQuantityByConcept" "drugDaysSupplyHistogram"
#> [23] "drugQuantityHistogram" "drugDurationHistogram"
#> [25] "diagnosticsSummary"
The first item contains information on the concept ids that are used in the database for a given ingredient.
glimpse(all_checks$conceptSummary)
#> Rows: 2
#> Columns: 26
#> Rowwise:
#> $ drug_concept_id <int> 40162522, 1127078
#> $ drug <chr> "Acetaminophen 325 MG / Hydrocodone Bitart…
#> $ ingredient_concept_id <dbl> 1125315, 1125315
#> $ ingredient <chr> "Acetaminophen", "Acetaminophen"
#> $ n_records <int> 312, 2158
#> $ n_patients <int> 312, 1428
#> $ domain_id <chr> "Drug", "Drug"
#> $ vocabulary_id <chr> "RxNorm", "RxNorm"
#> $ concept_class_id <chr> "Clinical Drug", "Clinical Drug"
#> $ standard_concept <chr> "S", "S"
#> $ concept_code <chr> "857005", "282464"
#> $ valid_start_date <date> 2009-08-02, 1970-01-01
#> $ valid_end_date <date> 2099-12-31, 2099-12-31
#> $ invalid_reason <chr> NA, NA
#> $ amount_value <dbl> 1, 1
#> $ amount_unit_concept_id <dbl> 1, 1
#> $ numerator_value <dbl> 1, 1
#> $ numerator_unit_concept_id <dbl> 1, 1
#> $ numerator_unit <chr> NA, NA
#> $ denominator_value <dbl> 1, 1
#> $ denominator_unit_concept_id <dbl> 1, 1
#> $ denominator_unit <chr> NA, NA
#> $ box_size <int> NA, NA
#> $ amount_unit <chr> NA, NA
#> $ dose_form <chr> NA, NA
#> $ result_obscured <lgl> FALSE, FALSE
all_checks$conceptSummary %>%
select("drug_concept_id", "drug")
#> # A tibble: 2 × 2
#> # Rowwise:
#> drug_concept_id drug
#> <int> <chr>
#> 1 40162522 Acetaminophen 325 MG / Hydrocodone Bitartrate 7.5 MG Oral Tab…
#> 2 1127078 Acetaminophen 160 MG Oral Tablet
Other tibbles then contain information from the various checks performed.
For example, we can see a summary of missingness for the ingredient-related records in the drug exposure table, both overall and by concept.
all_checks$missingValuesOverall
#> # A tibble: 18 × 9
#> # Rowwise: ingredient_concept_id, ingredient
#> ingredient_concept_id ingredient variable n_records n_sample
#> <dbl> <chr> <chr> <int> <dbl>
#> 1 1125315 Acetaminophen n_missing_drug_exposu… 59 10000
#> 2 1125315 Acetaminophen n_missing_drug_exposu… 59 10000
#> 3 1125315 Acetaminophen n_missing_verbatim_en… 59 10000
#> 4 1125315 Acetaminophen n_missing_drug_type_c… 59 10000
#> 5 1125315 Acetaminophen n_missing_stop_reason 59 10000
#> 6 1125315 Acetaminophen n_missing_refills 59 10000
#> 7 1125315 Acetaminophen n_missing_quantity 59 10000
#> 8 1125315 Acetaminophen n_missing_days_supply 59 10000
#> 9 1125315 Acetaminophen n_missing_sig 59 10000
#> 10 1125315 Acetaminophen n_missing_route_conce… 59 10000
#> 11 1125315 Acetaminophen n_missing_lot_number 59 10000
#> 12 1125315 Acetaminophen n_missing_provider_id 59 10000
#> 13 1125315 Acetaminophen n_missing_visit_occur… 59 10000
#> 14 1125315 Acetaminophen n_missing_visit_detai… 59 10000
#> 15 1125315 Acetaminophen n_missing_drug_source… 59 10000
#> 16 1125315 Acetaminophen n_missing_drug_source… 59 10000
#> 17 1125315 Acetaminophen n_missing_route_sourc… 59 10000
#> 18 1125315 Acetaminophen n_missing_dose_unit_s… 59 10000
#> # ℹ 4 more variables: n_records_not_missing_value <dbl>,
#> # n_records_missing_value <dbl>, proportion_records_missing_value <dbl>,
#> # result_obscured <lgl>
all_checks$missingValuesByConcept
#> # A tibble: 18 × 11
#> # Rowwise: drug_concept_id, drug, ingredient_concept_id, ingredient
#> drug_concept_id drug ingredient_concept_id ingredient variable n_records
#> <int> <chr> <dbl> <chr> <chr> <int>
#> 1 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 2 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 3 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 4 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 5 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 6 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 7 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 8 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 9 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 10 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 11 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 12 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 13 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 14 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 15 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 16 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 17 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> 18 40162522 Acetamin… 1125315 Acetamino… n_missi… 59
#> # ℹ 5 more variables: n_sample <dbl>, n_records_not_missing_value <dbl>,
#> # n_records_missing_value <dbl>, proportion_records_missing_value <dbl>,
#> # result_obscured <lgl>
Or we can also see a summary of drug exposure duration (drug_exposure_end_date - drug_exposure_end_date + 1), again overall or by concept.
all_checks$drugExposureDurationOverall
#> # A tibble: 1 × 17
#> # Rowwise: ingredient_concept_id
#> ingredient_concept_id ingredient n_records n_sample n_non_negative_days
#> <dbl> <chr> <int> <dbl> <int>
#> 1 1125315 Acetaminophen 59 10000 59
#> # ℹ 12 more variables: n_negative_days <int>, proportion_negative_days <dbl>,
#> # minimum_drug_exposure_days <dbl>, q05_drug_exposure_days <dbl>,
#> # q10_drug_exposure_days <dbl>, q25_drug_exposure_days <dbl>,
#> # median_drug_exposure_days <dbl>, q75_drug_exposure_days <dbl>,
#> # q90_drug_exposure_days <dbl>, q95_drug_exposure_days <dbl>,
#> # maximum_drug_exposure_days <dbl>, result_obscured <lgl>
all_checks$drugExposureDurationByConcept
#> # A tibble: 1 × 19
#> # Rowwise: drug_concept_id, drug, ingredient_concept_id
#> drug_concept_id drug ingredient_concept_id ingredient n_records n_sample
#> <int> <chr> <dbl> <chr> <int> <dbl>
#> 1 40162522 Acetamino… 1125315 Acetamino… 59 10000
#> # ℹ 13 more variables: n_non_negative_days <int>, n_negative_days <int>,
#> # proportion_negative_days <dbl>, minimum_drug_exposure_days <dbl>,
#> # q05_drug_exposure_days <dbl>, q10_drug_exposure_days <dbl>,
#> # q25_drug_exposure_days <dbl>, median_drug_exposure_days <dbl>,
#> # q75_drug_exposure_days <dbl>, q90_drug_exposure_days <dbl>,
#> # q95_drug_exposure_days <dbl>, maximum_drug_exposure_days <dbl>,
#> # result_obscured <lgl>
For further information on the checks performed please see the package vignettes.
After running the checks we can write the CSVs to disk using the writeResultToDisk
function.