R/remove_non_complete.R

In UMCSTaR/medicareAnalytics: Analytic Tools for Medicare Data Research Projects

#' Remove non-complete cases
#' 
#' @details included variables are the commoly used varaibles in the models. Variables included:
#'    "flg_cmp_po_severe", "flg_male", "age_at_admit", "e_race_wbho", "e_admit_type",
#'    "AHRQ_score", "e_ses_5grp","facility_clm_yr", "had_assist_surg",
#'    "e_hosp_beds_4grp","flg_hosp_ICU_hosp", "val_hosp_mcday2inptday_ratio",
#'    "val_hosp_rn2bed_ratio", "id_physician_npi","facility_prvnumgrp"
#'
#' @param data medicare data
#' @param variables variables that need to be used to exclude missing values, for example
#'    years of experience. If need to add multiple variables, use syntax c("var1", "var2")
#'
#' @return
#' @export
#'
#' @examples
remove_non_complete <- function(data,
                                variables = c(
                                  "flg_cmp_po_severe", 
                                  "flg_male", 
                                  "age_at_admit", 
                                  "e_race_wbho",
                                  "e_admit_type",
                                  "AHRQ_score", 
                                  "e_ses_5grp",
                                  "facility_clm_yr", 
                                  "had_assist_surg",
                                  "e_hosp_beds_4grp",
                                  "flg_hosp_ICU_hosp", 
                                  "val_hosp_mcday2inptday_ratio", 
                                  "val_hosp_rn2bed_ratio", 
                                  "id_physician_npi", 
                                  "facility_prvnumgrp"
                                )) {
  # model variables 
  model_vars = unique(variables)
  
  data %>% 
    drop_na(!!model_vars)
  
}