R/examples_invalids.R

In markhwhiteii/biosensequality: A Data Quality Report Generator for NSSP BioSense Platform

#' Generate Examples of Invalid Fields
#'
#' @description
#' NOTE: THIS FUNCTION IS NOT MEANT TO BE USED ON ITS OWN. IT IS USED INTERNALLY BY `write_reports`.
#' This function will return the C_BioSense_ID and field name for patient-visits where the entry was invalid.
#'
#' @param i A facility ID, as a string.
#' @param invalid_examples A list of data frames that are the examples output from the `_invalid` functions.
#' @return A data frame containing the patient-visit IDs and fields for this visit that were invalid.
#' @import dplyr
#' @export
examples_invalids <- function(i, invalid_examples) {

  # too many little different things for every df to do it as a
  # for loop or lapply/sapply; someone cleverer could figure it out sometime?

  ## admit source
  output <- invalid_examples[[1]] %>% # get examples
    filter(C_Biosense_Facility_ID==i & Invalid_Admit_Source==TRUE) %>% # for invalids in this facility
    select(C_BioSense_ID, Admit_Source) %>% # select vars we want
    distinct() # get rid of dupes

  ## age
  output <- invalid_examples[[2]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Age==TRUE) %>%
    select(C_BioSense_ID, Age_Reported) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## age units
  output <- invalid_examples[[2]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Age_Units==TRUE) %>%
    select(C_BioSense_ID, Age_Units_Reported) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## any e in visit
  output <- invalid_examples[[3]] %>%
    filter(C_Biosense_Facility_ID==i & No_E_In_Visit==TRUE) %>%
    select(C_BioSense_ID, No_E_In_Visit) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## blood pressure units
  output <- invalid_examples[[4]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Blood_Pressure_Units==TRUE) %>%
    select(C_BioSense_ID, Systolic_Diastolic_Blood_Pressure_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing bp units
  output <- invalid_examples[[4]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_BP_Units_Given_BP==TRUE) %>%
    select(C_BioSense_ID, Missing_BP_Units_Given_BP) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing bp
  output <- invalid_examples[[4]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_BP_Given_BP_Units==TRUE) %>%
    select(C_BioSense_ID, Missing_BP_Given_BP_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## admit reason description
  output <- invalid_examples[[5]] %>%
    filter(C_Biosense_Facility_ID==i & Admit_Reason_Description==TRUE) %>%
    select(C_BioSense_ID, Admit_Reason_Description) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## chief complaint text
  output <- invalid_examples[[5]] %>%
    filter(C_Biosense_Facility_ID==i & Chief_Complaint_Text==TRUE) %>%
    select(C_BioSense_ID, Chief_Complaint_Text) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## patient country
  output <- invalid_examples[[6]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Patient_Country==TRUE) %>%
    select(C_BioSense_ID, Patient_Country) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing death given dd
  output <- invalid_examples[[7]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Death_Given_Discharge_Disposition==TRUE) %>%
    select(C_BioSense_ID, Missing_Death_Given_Discharge_Disposition) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing death date time given death
  output <- invalid_examples[[7]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Death_Date_Time_Given_Indicator==TRUE) %>%
    select(C_BioSense_ID, Missing_Death_Date_Time_Given_Indicator) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing death given ddt
  output <- invalid_examples[[7]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Death_Indicator_Given_Date_Time==TRUE) %>%
    select(C_BioSense_ID, Missing_Death_Indicator_Given_Date_Time) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## diagnosis type
  output <- invalid_examples[[8]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Diagnosis_Type==TRUE) %>%
    select(C_BioSense_ID, Diagnosis_Type) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## discharge disposition
  output <- invalid_examples[[9]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Discharge_Disposition==TRUE) %>%
    select(C_BioSense_ID, Discharge_Disposition) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## discharge disposition
  output <- invalid_examples[[10]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Ethnicity_Code==TRUE) %>%
    select(C_BioSense_ID, Ethnicity_Code) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## facility type
  output <- invalid_examples[[11]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Facility_Type_Code==TRUE) %>%
    select(C_BioSense_ID, Facility_Type_Code) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## mrn, fpid mismatch
  output <- invalid_examples[[12]] %>%
    filter(C_Biosense_Facility_ID==i & FPID_MRN_Mismatch==TRUE) %>%
    select(C_BioSense_ID, FPID_MRN_Mismatch) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## sex
  output <- invalid_examples[[13]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Administrative_Sex==TRUE) %>%
    select(C_BioSense_ID, Administrative_Sex) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## height units
  output <- invalid_examples[[14]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Height_Units==TRUE) %>%
    select(C_BioSense_ID, Height_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing height units
  output <- invalid_examples[[14]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Height_Units_Given_Height==TRUE) %>%
    select(C_BioSense_ID, Missing_Height_Units_Given_Height) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing height
  output <- invalid_examples[[14]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Height_Given_Height_Units==TRUE) %>%
    select(C_BioSense_ID, Missing_Height_Given_Height_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## patient class
  output <- invalid_examples[[15]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Patient_Class_Code==TRUE) %>%
    select(C_BioSense_ID, Patient_Class_Code) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## invalid pulse ox units
  output <- invalid_examples[[16]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Pulse_Oximetry_Units==TRUE) %>%
    select(C_BioSense_ID, Initial_Pulse_Oximetry_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## invalid pulse ox
  output <- invalid_examples[[16]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Pulse_Oximetry==TRUE) %>%
    select(C_BioSense_ID, Initial_Pulse_Oximetry) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing pulse ox units
  output <- invalid_examples[[16]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Oximetry_Units_Given_Pulse_Ox==TRUE) %>%
    select(C_BioSense_ID, Missing_Oximetry_Units_Given_Pulse_Ox) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing pulse ox
  output <- invalid_examples[[16]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Pulse_Ox_Given_Oximetry_Units==TRUE) %>%
    select(C_BioSense_ID, Missing_Pulse_Ox_Given_Oximetry_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## race
  output <- invalid_examples[[17]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Race_Code==TRUE) %>%
    select(C_BioSense_ID, Race_Code) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## smoking status
  output <- invalid_examples[[18]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Smoking_Status_Code==TRUE) %>%
    select(C_BioSense_ID, Smoking_Status_Code) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## patient state
  output <- invalid_examples[[19]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Patient_State==TRUE) %>%
    select(C_BioSense_ID, Patient_State) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## invalid temp units
  output <- invalid_examples[[20]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Temp_Units==TRUE) %>%
    select(C_BioSense_ID, Initial_Temp_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## temperature out of range
  output <- invalid_examples[[20]] %>%
    filter(C_Biosense_Facility_ID==i & Temp_OOR==TRUE) %>%
    select(C_BioSense_ID, Initial_Temp) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing temp
  output <- invalid_examples[[20]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Temp_Given_Units==TRUE) %>%
    select(C_BioSense_ID, Missing_Temp_Given_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing temp units
  output <- invalid_examples[[20]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Units_Given_Temp==TRUE) %>%
    select(C_BioSense_ID, Missing_Units_Given_Temp) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## invalid weight units
  output <- invalid_examples[[21]] %>%
    filter(C_Biosense_Facility_ID==i & Invalid_Weight_Units==TRUE) %>%
    select(C_BioSense_ID, Weight_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing weight units
  output <- invalid_examples[[21]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Weight_Units_Given_Weight==TRUE) %>%
    select(C_BioSense_ID, Missing_Weight_Units_Given_Weight) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  ## missing weight
  output <- invalid_examples[[21]] %>%
    filter(C_Biosense_Facility_ID==i & Missing_Weight_Given_Weight_Units==TRUE) %>%
    select(C_BioSense_ID, Missing_Weight_Given_Weight_Units) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  # patient zip
  output <- invalid_examples[[22]] %>%
    filter(C_Biosense_Facility_ID==i & Zip_Invalid==TRUE) %>%
    select(C_BioSense_ID, Patient_Zip) %>%
    distinct() %>%
    full_join(output, by="C_BioSense_ID")

  return(
    suppressWarnings( # it'll throw that it is converting it all to character; we do not care
      output %>% # take output
        gather(Field, Invalid_Entry, 2:ncol(.), na.rm=TRUE) # gather it all, get rid of NAs
    )
  )
}