Diagnostics of a codelist of measurement codes within a cohort

Usage

summariseCohortMeasurementUse(
  cohort,
  codes = NULL,
  timing = "during",
  byConcept = TRUE,
  byYear = FALSE,
  bySex = FALSE,
  ageGroup = NULL,
  dateRange = as.Date(c(NA, NA)),
  estimates = list(measurement_summary = c("min", "q25", "median", "q75", "max",
    "density"), measurement_value_as_number = c("min", "q01", "q05", "q25", "median",
    "q75", "q95", "q99", "max", "count_missing", "percentage_missing", "density"),
    measurement_value_as_concept = c("count", "percentage")),
  histogram = NULL,
  checks = c("measurement_summary", "measurement_value_as_number",
    "measurement_value_as_concept")
)

Arguments

cohort

A cohort in which to perform the diagnostics of the measurement codes provided.

codes

A codelist of measurement/observation codes for which to perform diagnostics. If NULL it uses the codelist used to create each of the cohorts.

timing

Three options: 1) "any" if the interest is on measurement recorded any time, 2) "during", if interested in measurements while the subject is in the cohort (or in observation if cohort = NULL), and 3) "cohort_start_date" for measurements occurring at cohort start date (or at "observation_period_start_date if cohort = NULL).

byConcept

TRUE or FALSE. If TRUE code use will be summarised by concept.

byYear

TRUE or FALSE. If TRUE code use will be summarised by year.

bySex

TRUE or FALSE. If TRUE code use will be summarised by sex.

ageGroup

If not NULL, a list of ageGroup vectors of length two.

dateRange

Two dates. The first indicating the earliest measurement date and the second indicating the latest possible measurement date.

estimates

A named list indicating, for each measurement diagnostics check, which estimates to retrieve. The names of the list should correspond to the diagnostics checks, and each list element should be a character vector specifying the estimates to compute.

Allowed estimates are those supported by the `summariseResult()` function in the **PatientProfiles** package. If omitted, all available estimates for each check will be returned.

histogram

Named list where names point to checks for which to get estimates for a histogram, and elements are numeric vectors indicating the bind-width. See function examples. Histogram only available for "measurement_summary" and "measurement_value_as_number".

checks

Diagnostics to run. Options are: "measurement_summary", "measurement_value_as_number", and "measurement_value_as_concept".

Value

A summarised result

Examples

# \donttest{
library(MeasurementDiagnostics)

cdm <- mockMeasurementDiagnostics()

result <- summariseCohortMeasurementUse(
  codes = list("test_codelist" = c(3001467L, 45875977L)),
  cohort = cdm$my_cohort, timing = "cohort_start_date"
)
#> → Getting measurement records based on 2 concepts.
#> → Subsetting records to the subjects and timing of interest.
#> → Getting time between records per person.
#> Summarising timings
#> ℹ The following estimates will be computed:
#> • time: min, q25, median, q75, max, density
#> ! Table is collected to memory as not all requested estimates are supported on
#>   the database side
#> → Start summary of data, at 2026-01-15 11:41:12.259733
#> ✔ Summary finished, at 2026-01-15 11:41:12.424236
#> → Getting measurements per subject.
#> Summarising subjects
#> ℹ The following estimates will be computed:
#> • measurements_per_subject: min, q25, median, q75, max, density
#> ! Table is collected to memory as not all requested estimates are supported on
#>   the database side
#> → Start summary of data, at 2026-01-15 11:41:12.932845
#> ✔ Summary finished, at 2026-01-15 11:41:13.043038
#> ! 2 duplicated rows eliminated.
#> → Summarising results - value as number.
#> Summarising value as number
#> ℹ The following estimates will be computed:
#> • value_as_number: min, q01, q05, q25, median, q75, q95, q99, max,
#>   count_missing, percentage_missing, density
#> ! Table is collected to memory as not all requested estimates are supported on
#>   the database side
#> → Start summary of data, at 2026-01-15 11:41:14.250465
#> ✔ Summary finished, at 2026-01-15 11:41:14.502911
#> → Summarising results - value as concept.
#> Summarising value as number
#> ℹ The following estimates will be computed:
#> • value_as_concept_id: count, percentage
#> → Start summary of data, at 2026-01-15 11:41:15.109098
#> ✔ Summary finished, at 2026-01-15 11:41:15.26809
#> → Binding all diagnostic results.

result <- summariseCohortMeasurementUse(
  codes = list("test_codelist" = c(3001467L, 45875977L)),
  cohort = cdm$my_cohort, timing = "cohort_start_date",
  histogram = list(
    "time" = list(
      '0 to 100' = c(0, 100), '110 to 200' = c(110, 200),
      '210 to 300' = c(210, 300), '310 to Inf' = c(310, Inf)
    ),
    "measurements_per_subject" = list(
      '0 to 10' = c(0, 10), '11 to 20' = c(11, 20), '21 to 30' = c(21, 30),
      '31 to Inf' = c(31, Inf)
    ),
    "value_as_number" =  list(
      '0 to 5' = c(0, 5), '6 to 10' = c(6, 10), '11 to 15' = c(11, 15),
      '>15' = c(16, Inf)
    )
  )
)
#> → Getting measurement records based on 2 concepts.
#> → Subsetting records to the subjects and timing of interest.
#> → Getting time between records per person.
#> Summarising timings
#> ℹ The following estimates will be computed:
#> • time: min, q25, median, q75, max, density, count
#> • time_band: count
#> ! Table is collected to memory as not all requested estimates are supported on
#>   the database side
#> → Start summary of data, at 2026-01-15 11:41:18.492308
#> ✔ Summary finished, at 2026-01-15 11:41:18.678832
#> → Getting measurements per subject.
#> Summarising subjects
#> ℹ The following estimates will be computed:
#> • measurements_per_subject: min, q25, median, q75, max, density, count
#> • measurements_per_subject_band: count
#> ! Table is collected to memory as not all requested estimates are supported on
#>   the database side
#> → Start summary of data, at 2026-01-15 11:41:19.427564
#> ✔ Summary finished, at 2026-01-15 11:41:19.551477
#> ! 2 duplicated rows eliminated.
#> → Summarising results - value as number.
#> Summarising value as number
#> ℹ The following estimates will be computed:
#> • value_as_number: min, q01, q05, q25, median, q75, q95, q99, max,
#>   count_missing, percentage_missing, density
#> • value_as_number_band: count
#> ! Table is collected to memory as not all requested estimates are supported on
#>   the database side
#> → Start summary of data, at 2026-01-15 11:41:21.006592
#> ✔ Summary finished, at 2026-01-15 11:41:21.286875
#> → Summarising results - value as concept.
#> Summarising value as number
#> ℹ The following estimates will be computed:
#> • value_as_concept_id: count, percentage
#> → Start summary of data, at 2026-01-15 11:41:21.896325
#> ✔ Summary finished, at 2026-01-15 11:41:22.066225
#> → Binding all diagnostic results.

CDMConnector::cdmDisconnect(cdm = cdm)
# }