Skip to contents

Introduction To DrugExposureDiagnostics

Source: vignettes/IntroductionToDrugExposureDiagnostics.Rmd
IntroductionToDrugExposureDiagnostics.Rmd
library(DrugExposureDiagnostics)
library(CDMConnector)
#> Error in get(paste0(generic, ".", class), envir = get_method_env()) : 
#>   object 'type_sum.accel' not found
library(dplyr)
library(DT)

First, connect to the database. Examples of how to connect your database using CDMConnector can be found here: https://darwin-eu.github.io/CDMConnector/articles/a04_DBI_connection_examples.html
Here we use the internal mock database.

cdm <- mockDrugExposure()
#> Note: method with signature 'DBIConnection#Id' chosen for function 'dbExistsTable',
#>  target signature 'duckdb_connection#Id'.
#>  "duckdb_connection#ANY" would also be valid
#> ! cdm name not specified and could not be inferred from the cdm source table
#> Warning: ! 1 column in cdm_source do not match expected column type:
#> • `cdm_version_concept_id` is character but expected integer

Drug(s) of interest

In the DrugExposureDiagnostics package, all the diagnostics are conducted on ingredient level, and if requested, additionally on drug level. We will use the ingredient “acetaminophen” as an example.
Here is a brief look at this ingredient.

Property Value
Concept Name acetaminophen
Domain ID Drug
Concept Class ID Ingredient
Vocabulary ID RxNorm
Concept ID 1125315
Concept code 161
Validity Valid
Concept Standard
Valid start 01-Jan-1970
Valid end 31-Dec-2099

Running the diagnostics

We can run all available checks at the same time using the ´executeChecks()´ function. This will return a list which contains the results of each check.
Multiple ingredients can be added in the vector of ´ingredients´.
If the parameter ´outputFolder´ is specified, the results will be saved to disk. The created outputs (CSV files) will be zipped and saved to a file ´filename´.zip This zip file serves as the input for the corresponding Shiny Application. If the parameter ´outputFolder´ is not specified and you want to save results to disk, it is needed to call ´writeResultToDisk´ (see below). If ´outputFolder´ is not specified, results are only stored in memory.

all_checks <- executeChecks(cdm,
  ingredients = c(1125315),
  subsetToConceptId = NULL,
  checks = c(
    "missing", "exposureDuration", "type", "route", "sourceConcept", "daysSupply",
    "verbatimEndDate", "dose", "sig", "quantity", "diagnosticsSummary"
  ),
  minCellCount = 5,
  sample = 10000,
  tablePrefix = NULL,
  earliestStartDate = "2010-01-01",
  verbose = FALSE,
  byConcept = TRUE,
  outputFolder = "output_folder",
  filename = "your_database"
)
#> population after earliestStartDate smaller than sample, sampling ignored
#> ℹ The following estimates will be computed:
#> • daily_dose: count_missing, percentage_missing, mean, sd, q05, q25, median,
#>   q75, q95, min, max
#> ! Table is collected to memory as not all requested estimates are supported on
#>   the database side
#> → Start summary of data, at 2024-12-23 09:20:21.692082
#> 
#> ✔ Summary finished, at 2024-12-23 09:20:21.983488

The cdm is the database reference of the OMOP CDM using the CDMConnector package.
The ingredients is a list of ingredients of interests, by default it is 1125315 for acetaminophen. The subsetToConceptId vector of concept IDs of the ingredients to filter. If a concept ID is positive it will be included, a negative one will be excluded. If NULL, all concept IDs for an ingredient will be considered. checks allows to select the checks to be executed, by default the missing values, the exposure duration and the quantity checks will be run. The minCellCount is minimum number of events to report, numbers lower than this will be obscured. sample is the number of samples, by default, 10.000 drug record samples will be used.
The tablePrefix is an optional value for database tables that will be created during executeChecks. earliestStartDate is the earliest data from which drug records will be included. verbose is a parameter that enables the printing of messages to the console. byConcept is a boolean that determines if only overall results should be returned or also by drug concept. outputFolder is the folder to write to. If NULL, results will not be written to disk, but just saved in memory. databaseId is the database identifier. This is an optional parameter used to create the name of a temporary directory to store results. fileName is the output file name. If NULL, it will be set to the databaseId.

The check “missing” outputs the missingValuesOverall if byConcept = FALSE and additionally missingValuesbyConcept if byConcept = TRUE
The check “exposureDuration” outputs the drugExposureDurationOverall if byConcept = FALSE and additionally drugExposureDurationbyConcept if byConcept = TRUE
The check “type” outputs the drugTypesOverall if byConcept = FALSE and additionally drugTypesbyConcept if byConcept = TRUE
The check “route” outputs the drugRoutesOverall if byConcept = FALSE and additionally drugRoutesbyConcept if byConcept = TRUE
The check “sourceConcept” outputs the sourceConceptsOverall (which is actually the drug concept level. It disregards the byConcept argument because only the drug concept level makes sense here)
The check “daysSupply” outputs the drugDaysSupplyOverall if byConcept = FALSE and additionally drugDaysSupplybyConcept if byConcept = TRUE
The check “verbatimEndDate” outputs the drugVerbatimEndDate if byConcept = FALSE and additionally drugVerbatimEndDatebyConcept if byConcept = TRUE
The check “dose” outputs the drugDoseOverall (output only on ingredient level. It disregards the byConcept argument because only the drug concept level makes sense here)
The check “sig” outputs the drugSig if byConcept = FALSE and additionally drugSigbyConcept if byConcept = TRUE
The check “quantity” outputs the drugQuantity if byConcept = FALSE and additionally drugQuantitybyConcept if byConcept = TRUE
The check “diagnosticsSummary” outputs a summary called diagnosticsSummary

Here, we see the files created:

names(all_checks)
#>  [1] "conceptSummary"                "missingValuesOverall"         
#>  [3] "missingValuesByConcept"        "drugExposureDurationOverall"  
#>  [5] "drugExposureDurationByConcept" "drugTypesOverall"             
#>  [7] "drugTypesByConcept"            "drugRoutesOverall"            
#>  [9] "drugRoutesByConcept"           "drugSourceConceptsOverall"    
#> [11] "drugDaysSupply"                "drugDaysSupplyByConcept"      
#> [13] "drugVerbatimEndDate"           "drugVerbatimEndDateByConcept" 
#> [15] "drugDose"                      "drugSig"                      
#> [17] "drugSigByConcept"              "drugQuantity"                 
#> [19] "drugQuantityByConcept"         "diagnosticsSummary"           
#> [21] "metadata"

Assessing the drug concepts (of interest)

The first item shown, the conceptSummary is not the output from a check but always output. This table is used within the function as an input for all the requested checks.
It contains information from the CONCEPT table and DRUG_STRENGTH table.
The conceptSummary has several use:
It can be quickly seen whether a certain drug is mapped on ingredient level. Thereby, the drug_concept_id and ingredient_concept_id are identical.
Furthermore, the conceptSummary can be used to make decisions about further refine the concept list. For example, though the argument subsetToConceptId in the executeChecks() function, the selected concepts can be requested or eliminated.
The conceptSummary in the case of acetaminophen with subsetToConceptId = NULL, the following drug concepts are part of the diagnostics investigation.

datatable(all_checks$conceptSummary,
  rownames = FALSE
)
Column Description
drug_concept_id ID of the drug concept.
drug Name of the drug concept.
ingredient_concept_id Concept ID of ingredient.
ingredient Name of drug ingredient for which the checks have been performed.
n_records Number of records for drug concept.
domain_id From the CONCEPT table. A foreign key to the domain table the drug concept belongs to.
vocabulary_id From the CONCEPT table. A foreign key to the vocabulary table indicating from which source the drug concept has been adapted.
concept_class_id From the CONCEPT table. The concept class of the drug concept (e.g.‘Ingredient’ or ‘Clinical Drug’).
standard_concept From the CONCEPT table. A flag indicating whether the drug concept is standard (‘S’), a classification (‘C’), or a non-standard Source (NULL).
concept_code From the CONCEPT table. The concept code represents the identifier of the Concept in the source vocabulary.
valid_start_date From the CONCEPT table. The date when the drug concept was first recorded
valid_end_date From the CONCEPT table. The date when the drug concept became invalid because it was deleted or superseded (updated) by a new concept.
invalid_reason From the CONCEPT table. Reason the Concept was invalidated. Possible values are D (deleted), U (replaced with an update) or NULL when valid_end_date has the default value.
amount_value From DRUG_STRENGTH table. The numeric value or the amount of active ingredient contained within the drug product.
amount_unit_concept_id From DRUG_STRENGTH table. The Concept representing the Unit of measure for the amount of active ingredient contained within the drug product.
numerator_value From DRUG_STRENGTH table. The concentration of the active ingredient contained within the drug product.
numerator_unit_concept_id From DRUG_STRENGTH table. The Concept representing the Unit of measure for the concentration of active ingredient.
numerator_unit From the CONCEPT table. The concept name associated with the numerator_unit_concept_id.
denominator_value From DRUG_STRENGTH table. The amount of total liquid (or other divisible product, such as ointment, gel, spray, etc.).
denominator_unit_concept_id From DRUG_STRENGTH table. The Concept representing the denominator unit for the concentration of active ingredient.
denominator_unit From the CONCEPT table. The concept name associated with the denominator_unit_concept_id
box_size The number of units of Clinical Branded Drug or Quantified Clinical or Branded Drug contained in a box as dispensed to the patient.
amount_unit From DRUG_STRENGTH table.
dose_form From CONCEPT_RELATIONSHIP table. The RxNorm dose form associated with the drug concept.
result_obscured TRUE if count has been suppressed due to being below the minimum cell count, otherwise FALSE.

Saving the diagnostic output

After running the checks, we can write the created outputs (CSV files) into a zip file to disk using the writeResultToDisk() function. Note: this is only needed if the outputFolder argument is not added to executeChecks! The zip created from using the writeResultToDisk() function serves as the input for the corresponding Shiny Application.
Please replace the “your_database” with the name of your database. Create your designated “output_folder” at your desired location and put the path if not the working directory.

writeResultToDisk(all_checks,
  databaseId = "your_database",
  outputFolder = "output_folder"
)

View results in the Shiny app & make it available for publishing 

viewResults(
  dataFolder = file.path(getwd(), "output_folder"),
  makePublishable = TRUE,
  publishDir = file.path(getwd(), "MyStudyResultsExplorer"),
  overwritePublishDir = TRUE
)