Recode medication variables for cycles 1-2 (wide format)

Recodes medication variables from cycles 1-2 wide-format data (one row per respondent, up to 80 ATC/MHR columns), and merges into the main cycle dataset. Wraps recodeflow::rec_with_table() and converts factor outputs to numeric.

Usage

recode_meds_cycles1to2(
  data,
  meds_data,
  variables,
  by = "clinicid",
  meds_database_name = NULL,
  variable_details = chmsflow::variable_details
)

Arguments

data

data.frame Main cycle data to merge medication variables into.

meds_data

data.frame Wide-format medication data (cycles 1-2). Must contain clinicid, ATC code columns atc_101a-atc_115a, atc_131a-atc_135a, atc_201a-atc_215a, atc_231a-atc_235a, and matching time-last-taken columns mhr_101b-mhr_115b, mhr_131b-mhr_135b, mhr_201b-mhr_215b, mhr_231b-mhr_235b. Column names are normalized to lowercase before recoding, so uppercase variants (e.g., CLINICID, ATC_101A) are accepted.

variables

character Medication variable names to derive (e.g., "any_htn_med").

by

character Respondent identifier column. Default is "clinicid".

meds_database_name

character Name of the meds database in variable_details. Defaults to the name of the meds_data argument. Override when passing a transformed object (e.g., head()).

variable_details

data.frame Variable details table. Defaults to chmsflow::variable_details.

Value

data.frame data with derived medication variables merged in as numeric columns.

See also

recode_meds_cycles3to6(), aggregate_meds_by_person()

Examples

result <- recode_meds_cycles1to2(
  cycle1,
  cycle1_meds,
  c("any_htn_med", "diab_med")
)
#> Warning:  is missing from variable details therefore cannot be recoded
head(result[, c("clinicid", "any_htn_med", "diab_med")])
#>   clinicid any_htn_med diab_med
#> 1        1           1        0
#> 2        2           1        0
#> 3        3           1        0
#> 4        4           1        1
#> 5        5           1        0
#> 6        6           1        0