Create cohorts measurement based cohorts — measurementCohort • CohortConstructor

measurementCohort() creates cohorts based on patient records contained in the measurement table. This function extends the conceptCohort() as it allows for measurement values associated with the records to be specified.

If valueAsConcept and valueAsNumber are NULL then no requirements on of the values associated with measurement records and using measurementCohort() will lead to the same result as using conceptCohort() (so long as all concepts are from the measurement domain).
If one of valueAsConcept and valueAsNumber is not NULL then records will be required to have values that satisfy the requirement specified.
If both valueAsConcept and valueAsNumber are not NULL, records will be required to have values that fulfill either of the requirements

Usage

measurementCohort(
  cdm,
  conceptSet,
  name,
  valueAsConcept = NULL,
  valueAsNumber = NULL,
  table = c("measurement", "observation"),
  inObservation = TRUE
)

Arguments

cdm: A cdm reference.
conceptSet: A conceptSet, which can either be a codelist or a conceptSetExpression.
name: Name of the new cohort table created in the cdm object.
valueAsConcept: A vector of cohort IDs used to filter measurements. Only measurements with these values in the value_as_concept_id column of the measurement table will be included. If NULL all entries independent of their value as concept will be considered.
valueAsNumber: A list indicating the range of values and the unit they correspond to, as follows: list("unit_concept_id" = c(rangeValue1, rangeValue2)). If no name is supplied in the list, no requirement on unit concept id will be applied. If NULL, all entries independent of their value as number will be included.
table: Name of OMOP tables to search for records of the concepts provided. Options are "measurement" and/or "observation".
inObservation: If TRUE, only records in observation will be used. If FALSE, records before the start of observation period will be considered, with startdate the start of observation.

Value

A cohort table

Examples

# \donttest{
library(CohortConstructor)
cdm <- mockCohortConstructor(con = NULL)
cdm$concept <- cdm$concept |>
  dplyr::union_all(
    dplyr::tibble(
      concept_id = c(4326744, 4298393, 45770407, 8876, 4124457),
      concept_name = c("Blood pressure", "Systemic blood pressure",
                       "Baseline blood pressure", "millimeter mercury column",
                       "Normal range"),
      domain_id = "Measurement",
      vocabulary_id = c("SNOMED", "SNOMED", "SNOMED", "UCUM", "SNOMED"),
      standard_concept = "S",
      concept_class_id = c("Observable Entity", "Observable Entity",
                           "Observable Entity", "Unit", "Qualifier Value"),
      concept_code = NA,
      valid_start_date = NA,
      valid_end_date = NA,
      invalid_reason = NA
    )
  )
cdm$measurement <- dplyr::tibble(
  measurement_id = 1:4,
  person_id = c(1, 1, 2, 3),
  measurement_concept_id = c(4326744, 4298393, 4298393, 45770407),
  measurement_date = as.Date(c("2000-07-01", "2000-12-11", "2002-09-08",
  "2015-02-19")),
  measurement_type_concept_id = NA,
  value_as_number = c(100, 125, NA, NA),
  value_as_concept_id = c(0, 0, 0, 4124457),
  unit_concept_id = c(8876, 8876, 0, 0)
)
cdm <- CDMConnector::copyCdmTo(
  con = DBI::dbConnect(duckdb::duckdb()),
  cdm = cdm, schema = "main")

cdm$cohort <- measurementCohort(
  cdm = cdm,
  name = "cohort",
  conceptSet = list("normal_blood_pressure" = c(4326744, 4298393, 45770407)),
  valueAsConcept = c(4124457),
  valueAsNumber = list("8876" = c(70, 120)),
  inObservation = TRUE
)
#> Warning: ! `codelist` casted to integers.
#> Warning: ✖ Domain observation (3 concepts) excluded because table observation is not
#>   present in the cdm.
#> ℹ Subsetting table measurement using 3 concepts with domain: measurement.
#> ℹ Combining tables.
#> ℹ Applying measurement requirements.
#> ℹ Getting records in observation.
#> ℹ Creating cohort attributes.
#> ✔ Cohort cohort created.

cdm$cohort
#> # Source:   table<cohort> [?? x 4]
#> # Database: DuckDB v1.2.1 [unknown@Linux 6.8.0-1021-azure:R 4.4.3/:memory:]
#> # ℹ 4 variables: cohort_definition_id <int>, subject_id <dbl>,
#> #   cohort_start_date <date>, cohort_end_date <date>

# You can also create multiple measurement cohorts, and include records
# outside the observation period.

cdm$cohort2 <- measurementCohort(
  cdm = cdm,
  name = "cohort2",
  conceptSet = list("normal_blood_pressure" = c(4326744, 4298393, 45770407),
                  "high_blood_pressure" = c(4326744, 4298393, 45770407)),
  valueAsConcept = c(4124457),
  valueAsNumber = list("8876" = c(70, 120),
                       "8876" = c(121, 200)),
  inObservation = FALSE
)
#> Warning: ! `codelist` casted to integers.
#> Warning: ✖ Domain observation (6 concepts) excluded because table observation is not
#>   present in the cdm.
#> ℹ Subsetting table measurement using 6 concepts with domain: measurement.
#> ℹ Combining tables.
#> ℹ Applying measurement requirements.
#> ℹ Getting records in observation.
#> Warning: Missing values are always removed in SQL aggregation functions.
#> Use `na.rm = TRUE` to silence this warning
#> This warning is displayed once every 8 hours.
#> ℹ Creating cohort attributes.
#> ✔ Cohort cohort2 created.

cdm$cohort2
#> # Source:   table<cohort2> [?? x 4]
#> # Database: DuckDB v1.2.1 [unknown@Linux 6.8.0-1021-azure:R 4.4.3/:memory:]
#>   cohort_definition_id subject_id cohort_start_date cohort_end_date
#>                  <int>      <dbl> <date>            <date>         
#> 1                    1          1 2009-10-09        2009-10-09     
#> 2                    2          1 2009-10-09        2009-10-09     

# }