Summarise treatments

Introduction

After creating a study cohort, for example of some specific condition of interest, we may be interested in describing the treatments received by the individuals within it. Here we show how such a summary can be obtained.

Create mock table

We will use mock data contained in the package throughout the vignette. Let’s modify cohort tables cohort1 and cohort2 in our mock dataset, so the first table includes 3 cohorts of health conditions (our study cohorts), while the second contains three are of treatments they could receive.

library(DrugUtilisation)
#> Error in get(paste0(generic, ".", class), envir = get_method_env()) : 
#>   object 'type_sum.accel' not found

cdm <- mockDrugUtilisation(numberIndividual = 200)

new_cohort_set <- settings(cdm$cohort1) |>
  dplyr::arrange(cohort_definition_id) |>
  dplyr::mutate(cohort_name = c("asthma", "bronchitis", "pneumonia"))

cdm$cohort1 <- cdm$cohort1 |>
  omopgenerics::newCohortTable(cohortSetRef = new_cohort_set)

new_cohort_set <- settings(cdm$cohort2) |>
  dplyr::arrange(cohort_definition_id) |>
  dplyr::mutate(cohort_name = c("albuterol", "fluticasone", "montelukast"))

cdm$cohort2 <- cdm$cohort2 |>
  omopgenerics::newCohortTable(cohortSetRef = new_cohort_set)

Notice that cohort1 is a cohort table with three cohorts representing three different conditions:

settings(cdm$cohort1)
#> # A tibble: 3 × 2
#>   cohort_definition_id cohort_name
#>                  <int> <chr>      
#> 1                    1 asthma     
#> 2                    2 bronchitis 
#> 3                    3 pneumonia

And cohort2 is a cohort table with three different treatment cohorts:

settings(cdm$cohort2)
#> # A tibble: 3 × 2
#>   cohort_definition_id cohort_name
#>                  <int> <chr>      
#> 1                    1 albuterol  
#> 2                    2 fluticasone
#> 3                    3 montelukast

Summarise treatment

The summariseTreatment() function produces a summary of the treatment received by our study cohorts. There are three mandatory arguments:

1. cohort: cohort from the cdm object.
2. treatmentCohortName: name of the treatment cohort’s table.
3. window: list of the windows where to summarise the treatments.

See an example of its usage below, where we use summariseTreatment() to summarise treatments defined in cohort2 in the cohorts defined in cohort1.

summariseTreatment(
  cohort = cdm$cohort1,
  treatmentCohortName = c("cohort2"),
  window = list(c(0, 0), c(1, 30))
)

strata parameter

We can also stratify our cohort and calculate the estimates within each strata group by using the strata parameter.

cdm[["cohort1"]] <- cdm[["cohort1"]] |>
  PatientProfiles::addSex() |>
  PatientProfiles::addAge(ageGroup = list("<40" = c(0, 39), ">=40" = c(40, 150)))

results <- summariseTreatment(
  cohort = cdm$cohort1,
  treatmentCohortName = c("cohort2"),
  window = list(c(0, 0)),
  treatmentCohortId = 1,
  strata = list("sex", "age_group")
)

Notice that we have also used the treatmentCohortId parameter to specify that we only want to explore albuterol across the cohorts defined in cohort1.

Visualise results

The package includes table and plot functions to help visualise the results from summariseTreatment().

Tables

The tableTreatment() function generates a table in gt, flextable, or tibble format from the summarised_result produced by summariseTreatment(). This function has customisation options to format the table according to user preferences.

tableTreatment(result = results)

								cdm_name
								DUS MOCK
censor_date	cohort_table_name	index_date	treatment_cohort_name	sex	age_group	Treatment	estimate_name	cohort_name
								asthma	bronchitis	pneumonia
Medication on index date
NA	temp	cohort_start_date	cohort2	overall	overall	albuterol	N (%)	5 (7.6 %)	6 (11.8 %)	7 (8.4 %)
NA	temp	cohort_start_date	cohort2	overall	overall	untreated	N (%)	61 (92.4 %)	45 (88.2 %)	76 (91.6 %)
NA	temp	cohort_start_date	cohort2	overall	overall	not in observation	N (%)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)
NA	temp	cohort_start_date	cohort2	Female	overall	albuterol	N (%)	2 (6.9 %)	3 (11.1 %)	2 (5.1 %)
NA	temp	cohort_start_date	cohort2	Female	overall	untreated	N (%)	27 (93.1 %)	24 (88.9 %)	37 (94.9 %)
NA	temp	cohort_start_date	cohort2	Female	overall	not in observation	N (%)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)
NA	temp	cohort_start_date	cohort2	Male	overall	albuterol	N (%)	3 (8.1 %)	3 (12.5 %)	5 (11.4 %)
NA	temp	cohort_start_date	cohort2	Male	overall	untreated	N (%)	34 (91.9 %)	21 (87.5 %)	39 (88.6 %)
NA	temp	cohort_start_date	cohort2	Male	overall	not in observation	N (%)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)
NA	temp	cohort_start_date	cohort2	overall	<40	albuterol	N (%)	5 (9.8 %)	6 (13.6 %)	4 (6.7 %)
NA	temp	cohort_start_date	cohort2	overall	<40	untreated	N (%)	46 (90.2 %)	38 (86.4 %)	56 (93.3 %)
NA	temp	cohort_start_date	cohort2	overall	<40	not in observation	N (%)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)
NA	temp	cohort_start_date	cohort2	overall	>=40	albuterol	N (%)	0 (0.0 %)	0 (0.0 %)	3 (13.0 %)
NA	temp	cohort_start_date	cohort2	overall	>=40	untreated	N (%)	15 (100.0 %)	7 (100.0 %)	20 (87.0 %)
NA	temp	cohort_start_date	cohort2	overall	>=40	not in observation	N (%)	0 (0.0 %)	0 (0.0 %)	0 (0.0 %)

Plots

The plotTreatment() function creates a bar plot showing the percentage of treated and untreated in each cohort, stratum, and time-window. This function offers customization options for colors, faceting, and handling of strata.

plotTreatment(
  result = results,
  facet =  sex + age_group ~ window_name + cohort_name,
  colour = "variable_level"
)