Generate a data table with basic inpatient stay information. At the minimum, it will include an encounter and hospital ID, along with other information if cohort is included in the input.

This function creates a data table of simulated encounter IDs and hospital IDs. The creation is either based on user's desired number of encounters and unique hospitals, or based on a given set of encounter IDs. It can be used to create long format data tables where users have control over average repeat frequency.

Usage

generate_id_hospital(
  nid = 1000,
  n_hospitals = 10,
  avg_repeats = 1.5,
  include_prop = 1,
  cohort = NULL,
  by_los = FALSE,
  seed = NULL
)

Arguments

nid

(integer)
Optional, number of unique encounter IDs to simulate

n_hospitals

(integer)
Optional, number of hospitals to simulate and assign to encounter IDs

avg_repeats

(numeric)
The average number of repeats per row in the final data table

include_prop

(numeric)
A number between 0 and 1, for the proportion of unique rows in cohort to include in the final data table

cohort

(data.table)
Optional, resembling the GEMINI "admdad" table to build the returned data table from

by_los

(logical)
Optional, whether to assign more repeats to longer hospital stays or not. Default to FALSE. When TRUE, two additional columns are required in the input cohort dataset - admission_date_time and discharge_date_time for calculating length of stay.

seed

(integer)
Optional, a number for setting the seed for reproducible results

Value

(data.table)
A data.table object with the same columns as cohort, but with some rows excluded and/or repeated based on user specifications. If cohort is not included, then it will have the following fields:

• genc_id (integer): Mock encounter number, may be repeated in multiple rows based on avg_repeats

• hospital_num (integer): Mock hospital ID number

Examples

sample_cohort <- data.table::data.table(genc_id = 1:100, hospital_num = rep(1:5, each = 20))
generate_id_hospital(cohort = sample_cohort, include_prop = 0.8, avg_repeats = 1.5, by_los = TRUE, seed = 1)
#>      genc_id hospital_num admission_date_time discharge_date_time   los
#>        <int>        <int>              <POSc>              <POSc> <num>
#>   1:      68            4                <NA>                <NA>    NA
#>   2:      39            2                <NA>                <NA>    NA
#>   3:       1            1                <NA>                <NA>    NA
#>   4:      34            2                <NA>                <NA>    NA
#>   5:      87            5                <NA>                <NA>    NA
#>  ---                                                                   
#> 129:      19            1                <NA>                <NA>    NA
#> 130:      19            1                <NA>                <NA>    NA
#> 131:      19            1                <NA>                <NA>    NA
#> 132:      19            1                <NA>                <NA>    NA
#> 133:      19            1                <NA>                <NA>    NA
generate_id_hospital(nid = 1000, n_hospitals = 10, avg_repeats = 1)
#>       genc_id hospital_num
#>         <int>        <int>
#>    1:       1            4
#>    2:       2            2
#>    3:       3            1
#>    4:       4            7
#>    5:       5            9
#>   ---                     
#>  996:     996            9
#>  997:     997            6
#>  998:     998            6
#>  999:     999            6
#> 1000:    1000            1