Categorizing logic in patient type with dplyr

tidyverse
1

Hi Team
I hope you will help me in this to classify emergency patient types by looking at the col 1 and 3.
The expected result would be the col 4 ed_ip_split_2 .
ed_only (NA in col 3 if_admitted_id and EDxxxx prefix in col 1 ed_ip_id),
ed_admit (ed_ip_id with prefix ED and IPxxxxx populated in col 3 if_admitted_id), and
ip_only (both col 1 and 3 populated with IPxxxxx prefix).
Repex is here.
Thank you very much.

# ed ip split

ed_ip_data = tibble::tribble(
                  ~ed_ip_id, ~ed_ip_split, ~if_admitted_id, ~ed_ip_split_2,
                "ED1700147",         "ED",              NA,      "ed_only",
                "ED1700147",         "ED",              NA,      "ed_only",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
               "IP04229022",         "IP",    "IP04229022",      "ip_only",
               "IP04229022",         "IP",    "IP04229022",      "ip_only",
               "IP04229022",         "IP",    "IP04229022",      "ip_only",
               "IP04229022",         "IP",    "IP04229022",      "ip_only"
               )

ed_ip_data
#> # A tibble: 10 x 4
#>    ed_ip_id   ed_ip_split if_admitted_id ed_ip_split_2
#>    <chr>      <chr>       <chr>          <chr>        
#>  1 ED1700147  ED          NA             ed_only      
#>  2 ED1700147  ED          NA             ed_only      
#>  3 ED1698078  ED          IP04218624     ed_admit     
#>  4 ED1698078  ED          IP04218624     ed_admit     
#>  5 ED1698078  ED          IP04218624     ed_admit     
#>  6 ED1698078  ED          IP04218624     ed_admit     
#>  7 IP04229022 IP          IP04229022     ip_only      
#>  8 IP04229022 IP          IP04229022     ip_only      
#>  9 IP04229022 IP          IP04229022     ip_only      
#> 10 IP04229022 IP          IP04229022     ip_only
2

Will this work for you?

library(tidyverse)
ed_ip_data = tibble::tribble(
                  ~ed_ip_id, ~ed_ip_split, ~if_admitted_id, ~ed_ip_split_2,
                "ED1700147",         "ED",              NA,      "ed_only",
                "ED1700147",         "ED",              NA,      "ed_only",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
                "ED1698078",         "ED",    "IP04218624",     "ed_admit",
               "IP04229022",         "IP",    "IP04229022",      "ip_only",
               "IP04229022",         "IP",    "IP04229022",      "ip_only",
               "IP04229022",         "IP",    "IP04229022",      "ip_only",
               "IP04229022",         "IP",    "IP04229022",      "ip_only"
               )
ed_ip_data |> mutate(New_ed_ip_split = case_when(
  is.na(if_admitted_id) ~ "ed_only",
  str_detect(ed_ip_split, "ED") & str_detect(if_admitted_id, "^IP") ~ "ed_admit",
  str_detect(ed_ip_split, "IP") & str_detect(if_admitted_id, "^IP") ~ "ip_only",
  TRUE ~ "Error"
))
#> # A tibble: 10 × 5
#>    ed_ip_id   ed_ip_split if_admitted_id ed_ip_split_2 New_ed_ip_split
#>    <chr>      <chr>       <chr>          <chr>         <chr>          
#>  1 ED1700147  ED          <NA>           ed_only       ed_only        
#>  2 ED1700147  ED          <NA>           ed_only       ed_only        
#>  3 ED1698078  ED          IP04218624     ed_admit      ed_admit       
#>  4 ED1698078  ED          IP04218624     ed_admit      ed_admit       
#>  5 ED1698078  ED          IP04218624     ed_admit      ed_admit       
#>  6 ED1698078  ED          IP04218624     ed_admit      ed_admit       
#>  7 IP04229022 IP          IP04229022     ip_only       ip_only        
#>  8 IP04229022 IP          IP04229022     ip_only       ip_only        
#>  9 IP04229022 IP          IP04229022     ip_only       ip_only        
#> 10 IP04229022 IP          IP04229022     ip_only       ip_only

Created on 2025-03-21 with reprex v2.1.1

1 Like
3

@FJCC Thank you very much for your prompt response. G