R/nsqip_clean_data.R
In UKBiostatCIRCL/UK-Biostat-CIRCL: UKBiostatCIRCL: A package for the common functions of the Biostat CIRCL at the University of Kentucky
Defines functions
nsqip_clean_data
Documented in
nsqip_clean_data
#### NSQIP Raw Data Cleaning ####
#' Cleaning the raw National Surgical Quality Improvement Program (NSQIP) Dataset
#'
#' This function is designed to perform routine data cleaning routines when using the raw NSQIP data.
#' Cleaning functions include missing data recoding, variable categorization/dichotomization, converting string
#' entries to usable formats, and changing variable classes. This function outputs a cleaned dataframe and variable summary.
#'
#' NOTE: This function contains the sub-function update_label() to change variable labels.
#'
#' @param dataset the filtered and merged NSQIP data frame outputted by the "nsqip_subset_pufs" function
#'
#' @export
## The function
nsqip_clean_data <- function(dataset = dat){
###########################
## Handle Missing Values ##
###########################
## Replacing Null's with NA's
dataset[ dataset == "NULL" ] <- NA
## Replacing -99's with NA's
dataset[ dataset == -99 ] <- NA
## Replacing "Unknown" with NA
dataset[ dataset == "Unknown" ] <- NA
## Replacing "Unknown" with NA
dataset[ dataset == "None assigned" ] <- NA
## Remove unused/erroneous variables
dataset$admsyr <- NULL
#############################
## Create Factor Variables ##
#############################
## Sex
dataset$sex <- factor(x = dataset$sex,
levels = c("male","female"),
labels = c("Male","Female"))
## Race
dataset$racenew <- factor(x = dataset$racenew,
levels = c("White","Black or African American","Asian","American Indian or Alaska Native",
"Native Hawaiian or Pacific Islander","Unknown/Not Reported"))
## Deep incisional SSI
dataset$wndinfd <- factor(x = dataset$wndinfd,
levels = c("No Complication","Deep Incisional SSI"))
## Renal failure
dataset$oprenafl <- factor(x = dataset$oprenafl,
levels = c("No Complication","Acute Renal Failure"))
## DVT
dataset$othdvt <- factor(x = dataset$othdvt,
levels = c("No Complication","DVT Requiring Therap","DVT Requiring Therapy"),
labels = c("No Complication","DVT Requiring Therapy","DVT Requiring Therapy"))
## Cardiac arrest
dataset$cdarrest <- factor(x = dataset$cdarrest,
levels = c("No Complication","Cardiac Arrest Requiring CPR"))
## Myocardial infarction
dataset$cdmi <- factor(x = dataset$cdmi,
levels = c("No Complication","Myocardial Infarction"))
## Wound closure
dataset$woundclosure <- factor(x = dataset$woundclosure,
levels = c("All layers of incision (deep and superficial) fully closed",
"Only deep layers closed; superficial left open",
"No layers of incision are surgically closed"))
## Factor variables that don't require re-ordering of levels
vars <- c("ventilat","hxcopd","ascites","hxchf","hypermed","renafail","dialysis","discancr","wndinf","steroid","wtloss","bleeddis","transfus","prsepis",
"smoke","caseid","inout","emergncy","fnstatus2","wndclas","orgspcssi","supinfec","dehis","oupneumo","othbleed","urninfec","reintub","failwean",
"othsysep","othseshock","renainsf","pulembol","cnscva")
for(i in vars){
dataset[, i] <- factor(x = dataset[, i])
}
rm(i, vars)
############################
## Dichotomize Lab Values ##
############################
## Pre-Operative Albumin < 3
dataset$pralbum_lt3 <- factor(x = as.numeric(dataset$pralbum < 3),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Alkaline Phosphatase > 125
dataset$pralkph_gt125 <- factor(x = as.numeric(dataset$pralkph > 125),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Bilirubin > 1
dataset$prbili_gt1 <- factor(x = as.numeric(dataset$prbili > 1),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative BUN > 40
dataset$prbun_gt40 <- factor(x = as.numeric(dataset$prbun > 40),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Creatinine > 1.2
dataset$prcreat_gt1p2 <- factor(x = as.numeric(dataset$prcreat > 1.2),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Hematocrit < 30%
dataset$prhct_lt30 <- factor(x = as.numeric(dataset$prhct < 30),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Hematocrit > 45%
dataset$prhct_gt45 <- factor(x = as.numeric(dataset$prhct > 45),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Platelets < 150,000
dataset$prplate_lt150 <- factor(x = as.numeric(dataset$prplate < 150),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Platelets > 400,000
dataset$prplate_gt400 <- factor(x = as.numeric(dataset$prplate > 400),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative SGOT > 40
dataset$prsgot_gt40 <- factor(x = as.numeric(dataset$prsgot > 40),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Sodium < 135
dataset$prsodm_lt135 <- factor(x = as.numeric(dataset$prsodm < 135),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Sodium > 145
dataset$prsodm_gt145 <- factor(x = as.numeric(dataset$prsodm > 145),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative White Blood Count <= 4,500
dataset$prwbc_lte4p5 <- factor(x = as.numeric(dataset$prwbc <= 4.5),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative White Blood Count > 11,000
dataset$prwbc_gt11 <- factor(x = as.numeric(dataset$prwbc > 11),
levels = c(0, 1),
labels = c("No","Yes"))
##########################
## Create New Variables ##
##########################
## Impute the age values of "90+" to "90"
dataset$age_90imp <- as.character(dataset$age)
dataset$age_90imp[ dataset$age_90imp == "90+" ] <- "90"
dataset$age_90imp <- as.numeric(dataset$age_90imp)
## Preop sepsis
dataset$prsepis_any <- factor(x = as.numeric(dataset$prsepis %in% c("Sepsis","SIRS")),
levels = c(0, 1),
labels = c("No","Yes"))
## BMI
if( is.factor(dataset$weight) ){
dataset$weight <- as.numeric( as.character(dataset$weight) )
}
if( is.factor(dataset$height) ){
dataset$height <- as.numeric( as.character(dataset$height) )
}
dataset$weight[ dataset$weight == -99 ] <- NA
dataset$height[ dataset$height == -99 ] <- NA
dataset$bmi <- ((dataset$weight / dataset$height / dataset$height) * 703)
## BMI > 30
dataset$bmi_gte30 <- factor(x = as.numeric(dataset$bmi >= 30),
levels = c(0, 1),
labels = c("BMI < 30","BMI 30+"))
## BMI Group
dataset$bmi_group <- cut(dataset$bmi,
c(0, 18.5, 25, 30, 35, 40, 9999),
include.lowest = TRUE,
right = FALSE)
dataset$bmi_group <- factor(x = dataset$bmi_group,
levels = c("[0,18.5)","[18.5,25)","[25,30)","[30,35)","[35,40)","[40,1e+04]"),
labels = c("<18.5","18.5-24.9","25.0-29.9","30.0-34.9","35.0-39.9","40.0+"))
## ASA Class Group
dataset$asaclas[ dataset$asaclas == "None assigned" ] <- NA
dataset$asaclas_group[ dataset$asaclas == "1-No Disturb" | dataset$asaclas == "2-Mild Disturb" ] <- "I-II"
dataset$asaclas_group[ dataset$asaclas == "3-Severe Disturb" ] <- "III"
dataset$asaclas_group[ dataset$asaclas == "4-Life Threat" | dataset$asaclas == "5-Moribund" ] <- "IV-V"
dataset$asaclas_group[ is.na(dataset$asaclas) ] <- NA
dataset$asaclas_group <- factor(x = dataset$asaclas_group,
levels = c("I-II","III","IV-V"))
## ASA Class > III
dataset$asaclas_gtIII <- factor(x = as.numeric(dataset$asaclas_group == "IV-V"),
levels = c(0, 1),
labels = c("ASA Class I, II, or III","ASA Class IV or V"))
## Non-white race
dataset$racenew_nonwhite <- factor(x = dataset$racenew,
levels = c("White","Black or African American","Asian","American Indian or Alaska Native",
"Native Hawaiian or Pacific Islander","Unknown/Not Reported"),
labels = c("White","Non-White","Non-White","Non-White",
"Non-White","Unknown/Not Reported"))
## Total hospital LOS
dataset$hospital_los_total <- dataset$tothlos
## Pre-operative hospital LOS
dataset$hospital_los_preop <- dataset$htooday
## Post-operative hospital LOS
dataset$hospital_los_postop <- (dataset$hospital_los_total - dataset$hospital_los_preop)
## Any diabetes
dataset$diabetes_present <- factor(x = as.numeric(dataset$diabetes %in% c("INSULIN","NON-INSULIN","ORAL")),
levels = c(0, 1),
labels = c("No Diabetes","Diabetes"))
## Any dyspnea
dataset$dyspnea_present <- factor(x = as.numeric(dataset$dyspnea %in% c("MODERATE EXERTION","AT REST")),
levels = c(0, 1),
labels = c("No Dyspnea","Dyspnea"))
##################################################
## Create Post-Operative Complication Variables ##
##################################################
## Standalone complications
dataset$comp_wound_disruption <- factor(x = dataset$dehis,
levels = c("No Complication","Wound Disruption"),
labels = c("No","Yes"))
dataset$comp_pneumonia <- factor(x = dataset$oupneumo,
levels = c("No Complication","Pneumonia"),
labels = c("No","Yes"))
dataset$comp_transfusion <- factor(x = dataset$othbleed,
levels = c("No Complication","Transfusions/Intraop/Postop"),
labels = c("No","Yes"))
dataset$comp_uti <- factor(x = dataset$urninfec,
levels = c("No Complication","Urinary Tract Infection"),
labels = c("No","Yes"))
## Individual complication variables
dataset$comp_ssi_super <- factor(x = dataset$supinfec,
levels = c("No Complication","Superficial Incisional SSI"),
labels = c("No","Yes"))
dataset$comp_ssi_deep_incis <- factor(x = dataset$wndinfd,
levels = c("No Complication","Deep Incisional SSI"),
labels = c("No","Yes"))
dataset$comp_ssi_organ_space <- factor(x = dataset$orgspcssi,
levels = c("No Complication","Organ/Space SSI"),
labels = c("No","Yes"))
## Any SSI
dataset$comp_ssi_any <- apply(X = dataset[, c("comp_ssi_super","comp_ssi_deep_incis","comp_ssi_organ_space") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") |
(!is.na(x[3]) & x[3] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) & is.na(x[3]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_ssi_any <- factor(x = dataset$comp_ssi_any)
## Individual complication variables
dataset$comp_unpl_intub <- factor(x = dataset$reintub,
levels = c("No Complication","Unplanned Intubation"),
labels = c("No","Yes"))
dataset$comp_vent48 <- factor(x = dataset$failwean,
levels = c("No Complication","On Ventilator greater than 48 Hours"),
labels = c("No","Yes"))
## Unplanned intubation or ventilator > 48 hours
dataset$comp_unpl_intub_or_vent48 <- apply(X = dataset[, c("comp_unpl_intub","comp_vent48") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_unpl_intub_or_vent48 <- factor(x = dataset$comp_unpl_intub_or_vent48)
## Individual complication variables
dataset$comp_sepsis <- factor(x = dataset$othsysep,
levels = c("No Complication","Sepsis"),
labels = c("No","Yes"))
dataset$comp_septic_shock <- factor(x = dataset$othseshock,
levels = c("No Complication","Septic Shock"),
labels = c("No","Yes"))
## Sepsis or septic shock
dataset$comp_sepsis_or_septic_shock <- apply(X = dataset[, c("comp_sepsis","comp_septic_shock") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_sepsis_or_septic_shock <- factor(x = dataset$comp_sepsis_or_septic_shock)
## Individual complication variables
dataset$comp_renal_insuff <- factor(x = dataset$renainsf,
levels = c("No Complication","Progressive Renal Insufficiency"),
labels = c("No","Yes"))
dataset$comp_renal_failure <- factor(x = dataset$oprenafl,
levels = c("No Complication","Acute Renal Failure"),
labels = c("No","Yes"))
## Renal insufficiency or failure
dataset$comp_renal_insuff_or_failure <- apply(X = dataset[, c("comp_renal_insuff","comp_renal_failure") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_renal_insuff_or_failure <- factor(x = dataset$comp_renal_insuff_or_failure)
## Individual complication variables
dataset$comp_pulm_embolism <- factor(x = dataset$pulembol,
levels = c("No Complication","Pulmonary Embolism"),
labels = c("No","Yes"))
dataset$comp_dvt <- factor(x = dataset$othdvt,
levels = c("No Complication","DVT Requiring Therapy"),
labels = c("No","Yes"))
## Pulmonary embolism or DVT
dataset$comp_pulm_embolism_or_dvt <- apply(X = dataset[, c("comp_pulm_embolism","comp_dvt") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_pulm_embolism_or_dvt <- factor(x = dataset$comp_pulm_embolism_or_dvt)
## Individual complication variables
dataset$comp_cardiac_arrest <- factor(x = dataset$cdarrest,
levels = c("No Complication","Cardiac Arrest Requiring CPR"),
labels = c("No","Yes"))
dataset$comp_mi <- factor(x = dataset$cdmi,
levels = c("No Complication","Myocardial Infarction"),
labels = c("No","Yes"))
dataset$comp_stroke <- factor(x = dataset$cnscva,
levels = c("No Complication","Stroke/CVA"),
labels = c("No","Yes"))
## Cardiac arrest, MI, or stroke
dataset$comp_cardiac_arrest_mi_stroke <- apply(X = dataset[, c("comp_cardiac_arrest","comp_mi","comp_stroke") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") |
(!is.na(x[3]) & x[3] == "Yes")){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) & is.na(x[3]) ){
return(NA)
} else{ return("No") }
} )
## Mortality
dataset$comp_mortality <- as.numeric( !is.na(dataset$dopertod) & dataset$dopertod >= 0 & dataset$dopertod <= 30)
dataset$comp_mortality[ is.na(dataset$comp_mortality) ] <- 0
dataset$comp_mortality <- factor(x = dataset$comp_mortality,
levels = c(0, 1),
labels = c("No","Yes"))
## Any complication, including mortality
dataset$comp_any_including_mortality <- apply(X = dataset[, c("comp_wound_disruption","comp_pneumonia","comp_transfusion",
"comp_uti","comp_ssi_super","comp_ssi_deep_incis",
"comp_ssi_organ_space","comp_unpl_intub","comp_vent48",
"comp_sepsis","comp_septic_shock","comp_renal_insuff",
"comp_renal_failure","comp_pulm_embolism","comp_dvt",
"comp_cardiac_arrest","comp_mi","comp_stroke",
"comp_mortality") ],
MARGIN = 1,
FUN = function(x){
if( sum(is.na(x)) == length(x) ){
return(NA)
} else if( sum(x == "Yes", na.rm = TRUE) > 0 ){
return("Yes")
} else{ return("No") }
} )
## Any complication, excluding mortality
dataset$comp_any_excluding_mortality <- apply(X = dataset[, c("comp_wound_disruption","comp_pneumonia","comp_transfusion",
"comp_uti","comp_ssi_super","comp_ssi_deep_incis",
"comp_ssi_organ_space","comp_unpl_intub","comp_vent48",
"comp_sepsis","comp_septic_shock","comp_renal_insuff",
"comp_renal_failure","comp_pulm_embolism","comp_dvt",
"comp_cardiac_arrest","comp_mi","comp_stroke") ],
MARGIN = 1,
FUN = function(x){
if( sum(is.na(x)) == length(x) ){
return(NA)
} else if( sum(x == "Yes", na.rm = TRUE) > 0 ){
return("Yes")
} else{ return("No") }
} )
##########################################
## Create a Variable Summary Data Frame ##
##########################################
## Create a variable summary data frame
variable_summary <- data.frame(Variable = names(dataset),
Label = names(dataset),
Variable_Type = NA,
Variable_Category = NA,
stringsAsFactors = FALSE)
## Identify continuous variables
vars_cont <- c("pufyear","workrvu","age","age_90imp","admyr","operyr","height","weight","bmi",
"dprna","dprbun","dprcreat","dpralbum","dprbili","dprsgot","dpralkph","dprwbc","dprhct","dprplate","dprptt",
"dprpt","dprinr",
"prsodm","prbun","prcreat","pralbum","prbili","prsgot","pralkph","prwbc","prhct","prplate","prptt","prinr",
"prpt",
"otherwrvu1","otherwrvu2","otherwrvu3","otherwrvu4","otherwrvu5","otherwrvu6","otherwrvu7","otherwrvu8",
"otherwrvu9","otherwrvu10",
"conwrvu1","conwrvu2","conwrvu3","conwrvu4","conwrvu5","conwrvu6","conwrvu7","conwrvu8","conwrvu9","conwrvu10",
"optime","hdisdt","yrdeath","tothlos","htooday","hospital_los_total","hospital_los_preop","hospital_los_postop",
"dsupinfec","dwndinfd","dorgspcssi","ddehis","doupneumo","dreintub","dpulembol","dfailwean","drenainsf",
"doprenafl","durninfec","dcnscva","dcdarrest","dcdmi","dothbleed","dothdvt","dothsysep","dothseshock",
"retorpodays","retor2podays","readmpodays1","readmpodays2","readmpodays3","readmpodays4","readmpodays5",
"dothcdiff")
## Identify ordinal variables
vars_ord <- c("wndclas","asaclas_group","bmi_group","admqtr",
"nsupinfec","nwndinfd","norgspcssi","ndehis","noupneumo","nreintub","npulembol","nfailwean","nrenainsf",
"noprenafl","nurninfec","ncnscva","ncdarrest","ncdmi","nothbleed","nothdvt","nothsysep","nothseshock",
"nothcdiff","fnstatus2","woundclosure")
## Define variable types
variable_summary$Variable_Type <- "Categorical"
variable_summary$Variable_Type[ variable_summary$Variable %in% vars_cont ] <- "Continuous"
variable_summary$Variable_Type[ variable_summary$Variable %in% vars_ord ] <- "Ordinal"
## Identify common patient characteristic variables
vars_pat <- c("age_90imp","sex","racenew","ethnicityhispanic","bmi","bmi_gte30",
"smoke","diabetes_present","dyspnea_present","fnstatus2",
"ventilat","hxcopd","ascites","hxchf","hypermed","renafail",
"dialysis","discancr","wndinf","steroid","wtloss","bleeddis",
"transfus","prsepis_any")
## Identify common surgical charactistic variables
vars_surg <- c("optime","cpt","asaclas_group","wndclas","inout","electsurg","emergncy","woundclosure")
## Identify lab value variables, continuous
vars_labs_cont <- c("prsodm","prbun","prcreat","pralbum","prbili","prsgot","pralkph","prwbc","prhct","prplate","prptt","prinr","prpt")
## Identify lab value variables, dichotomous
vars_labs_dich <- c("pralbum_lt3","pralkph_gt125","prbili_gt1","prbun_gt40","prcreat_gt1p2","prhct_lt30","prhct_gt45","prplate_lt150","prplate_gt400",
"prsgot_gt40","prsodm_lt135","prsodm_gt145","prwbc_lte4p5","prwbc_gt11")
## Identify outcome variables
vars_outs <- c("hospital_los_total","comp_mortality","comp_any_including_mortality","comp_any_excluding_mortality",
"comp_ssi_any","comp_wound_disruption","comp_pneumonia","comp_transfusion","comp_uti","comp_unpl_intub_or_vent48",
"comp_sepsis_or_septic_shock","comp_renal_insuff_or_failure","comp_pulm_embolism_or_dvt","comp_cardiac_arrest_mi_stroke")
## Define the variable categories
variable_summary$Variable_Category <- "Other"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_pat ] <- "Patient Characteristics"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_surg ] <- "Surgical Characteristics"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_labs_cont ] <- "Continous Lab Values"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_labs_dich ] <- "Dichotomous Lab Values"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_outs ] <- "Outcomes"
## A function to update variable labels
update_label <- function(df = variable_summary, variable, label){
variable_summary$Label[ variable_summary$Variable == variable ] <- label
return(variable_summary)
}
## Manually update variable labels
variable_summary <- update_label(variable = "age_90imp", label = "Age (years)")
variable_summary <- update_label(variable = "sex", label = "Female")
variable_summary <- update_label(variable = "racenew", label = "Race")
variable_summary <- update_label(variable = "ethnicityhispanic", label = "Hispanic ethnicity")
variable_summary <- update_label(variable = "bmi", label = "Body mass index (kg/m2)")
variable_summary <- update_label(variable = "bmi_gte30", label = "Body mass index >= 30 kg/m2")
variable_summary <- update_label(variable = "bmi_group", label = "Body mass index (kg/m2)")
variable_summary <- update_label(variable = "cpt", label = "Primary CPT code")
variable_summary <- update_label(variable = "inout", label = "Outpatient")
variable_summary <- update_label(variable = "electsurg", label = "Elective surgery, patient coming from home")
variable_summary <- update_label(variable = "smoke", label = "Current smoker within one year")
variable_summary <- update_label(variable = "fnstatus2", label = "Functional status prior to surgery")
variable_summary <- update_label(variable = "ventilat", label = "Ventilator dependent")
variable_summary <- update_label(variable = "hxcopd", label = "History of severe COPD")
variable_summary <- update_label(variable = "ascites", label = "Ascites with 30 days prior to surgery")
variable_summary <- update_label(variable = "hxchf", label = "Congestive heart failure within 30 days prior to surgery")
variable_summary <- update_label(variable = "hypermed", label = "Hypertension requiring medication")
variable_summary <- update_label(variable = "renafail", label = "Acute renal failure (pre-op)")
variable_summary <- update_label(variable = "dialysis", label = "Currently on dialysis (pre-op)")
variable_summary <- update_label(variable = "discancr", label = "Disseminated cancer")
variable_summary <- update_label(variable = "wndinf", label = "Open wound (with or without infection)")
variable_summary <- update_label(variable = "steroid", label = "Steroid/Immunosuppressant usre for a chronic condition")
variable_summary <- update_label(variable = "wtloss", label = "Greater than 10% loss of body weight in the 6 months prior to surgery")
variable_summary <- update_label(variable = "bleeddis", label = "Bleeding disorder")
variable_summary <- update_label(variable = "transfus", label = "Pre-operative transfusion (RBC wihtin 72 hours prior to surgery start time)")
variable_summary <- update_label(variable = "diabetes_present", label = "Diabetes mellitus requiring therapy")
variable_summary <- update_label(variable = "dyspnea_present", label = "Dyspnea at rest or with moderate exertion")
variable_summary <- update_label(variable = "prsepis_any", label = "Sepsis within 48 hours prior to surgery")
variable_summary <- update_label(variable = "emergncy", label = "Emergency case")
variable_summary <- update_label(variable = "wndclas", label = "Wound classification")
variable_summary <- update_label(variable = "optime", label = "Total operation time (minutes)")
variable_summary <- update_label(variable = "woundclosure", label = "Surgical wound closure")
variable_summary <- update_label(variable = "asaclas_group", label = "ASA classification")
variable_summary <- update_label(variable = "hospital_los_total", label = "Lenght of total hospital stay (days)")
variable_summary <- update_label(variable = "prsodm", label = "Pre-operative serum sodium")
variable_summary <- update_label(variable = "prsodm_lt135", label = "Pre-operative serum sodium < 135")
variable_summary <- update_label(variable = "prsodm_gt145", label = "Pre-operative serum sodium > 145")
variable_summary <- update_label(variable = "prbun", label = "Pre-operative blood urea nitrogen (BUN)")
variable_summary <- update_label(variable = "prbun_gt40", label = "Pre-operative blood urea nitrogen (BUN) > 40")
variable_summary <- update_label(variable = "prcreat", label = "Pre-operative serum creatinine")
variable_summary <- update_label(variable = "prcreat_gt1p2", label = "Pre-operative serum creatinine > 1.2")
variable_summary <- update_label(variable = "pralbum", label = "Pre-operative serum albumin")
variable_summary <- update_label(variable = "pralbum_lt3", label = "Pre-operative serum albumin < 3")
variable_summary <- update_label(variable = "prbili", label = "Pre-operative total bilirubin")
variable_summary <- update_label(variable = "prbili_gt1", label = "Pre-operative total bilirubin > 1")
variable_summary <- update_label(variable = "prsgot", label = "Pre-operative serum glutamic-oxaloacetic transaminase (SGOT)")
variable_summary <- update_label(variable = "prsgot_gt40", label = "Pre-operative serum glutamic-oxaloacetic transaminase (SGOT) > 40")
variable_summary <- update_label(variable = "pralkph", label = "Pre-operative alkaline phosphatase")
variable_summary <- update_label(variable = "pralkph_gt125", label = "Pre-operative alkaline phosphatase > 125")
variable_summary <- update_label(variable = "prwbc", label = "Pre-operative white blood cell (WBC) count")
variable_summary <- update_label(variable = "prwbc_lte4p5", label = "Pre-operative white blood cell (WBC) count <= 4.5")
variable_summary <- update_label(variable = "prwbc_gt11", label = "Pre-operative white blood cell (WBC) count > 11")
variable_summary <- update_label(variable = "prhct", label = "Pre-operative hematocrit")
variable_summary <- update_label(variable = "prhct_lt30", label = "Pre-operative hematocrit < 30")
variable_summary <- update_label(variable = "prhct_gt45", label = "Pre-operative hematocrit > 45")
variable_summary <- update_label(variable = "prplate", label = "Pre-operative platelet count")
variable_summary <- update_label(variable = "prplate_lt150", label = "Pre-operative platelet count < 150")
variable_summary <- update_label(variable = "prplate_gt400", label = "Pre-operative platelet count > 400")
variable_summary <- update_label(variable = "prptt", label = "Pre-operative partial thromboplastin time (PTT)")
variable_summary <- update_label(variable = "prinr", label = "Pre-operative international normalized ratio (INR) of prothrombin time (PT) values")
variable_summary <- update_label(variable = "prpt", label = "Pre-operative prothrombin time (PT)")
variable_summary <- update_label(variable = "comp_mortality", label = "Death within 30 days or surgery")
variable_summary <- update_label(variable = "comp_wound_disruption", label = "Post-operative wound disruption")
variable_summary <- update_label(variable = "comp_pneumonia", label = "Post-operative pneumonia")
variable_summary <- update_label(variable = "comp_transfusion", label = "Intra-/post-operative transfusion (RBC within the first 72 hours of surgery start time)")
variable_summary <- update_label(variable = "comp_uti", label = "Post-operative UTI")
variable_summary <- update_label(variable = "comp_ssi_super", label = "Post-operative superficial incisional SSI")
variable_summary <- update_label(variable = "comp_ssi_deep_incis", label = "Post-operative deep incisional SSI")
variable_summary <- update_label(variable = "comp_ssi_organ_space", label = "Post-operative organ/space SSI")
variable_summary <- update_label(variable = "comp_ssi_any", label = "Post-operative SSI (superficial incisional, deep incisional, or organ/space)")
variable_summary <- update_label(variable = "comp_unpl_intub", label = "Post-operative unplanned intubation")
variable_summary <- update_label(variable = "comp_vent48", label = "Post-operative venilator use > 48 hours")
variable_summary <- update_label(variable = "comp_unpl_intub_or_vent48", label = "Post-operative unplanned intubation or ventilator use > 48 hours")
variable_summary <- update_label(variable = "comp_sepsis", label = "Post-operative sepsis")
variable_summary <- update_label(variable = "comp_septic_shock", label = "Post-operative septic shock")
variable_summary <- update_label(variable = "comp_sepsis_or_septic_shock", label = "Post-operative sepsis or septic shock")
variable_summary <- update_label(variable = "comp_renal_insuff", label = "Post-operative progressive renal insufficiency")
variable_summary <- update_label(variable = "comp_renal_failure", label = "Post-operative acute renal failure")
variable_summary <- update_label(variable = "comp_renal_insuff_or_failure", label = "Post-operative progressive renal insufficiency or acute renal failure")
variable_summary <- update_label(variable = "comp_pulm_embolism", label = "Post-operative pulmonary embolism")
variable_summary <- update_label(variable = "comp_dvt", label = "Post-operative DVT requiring therapy")
variable_summary <- update_label(variable = "comp_pulm_embolism_or_dvt", label = "Post-operative pulmonary embolism or DVT requiring therapy")
variable_summary <- update_label(variable = "comp_cardiac_arrest", label = "Intra-/post-operative cardiac arrest requiring CPR")
variable_summary <- update_label(variable = "comp_mi", label = "Intra-/post-operative myocardial infarction")
variable_summary <- update_label(variable = "comp_stroke", label = "Post-operative stroke/CVA")
variable_summary <- update_label(variable = "comp_cardiac_arrest_mi_stroke", label = "Post-operative cardiac arrest requiring CPR, myocardial infarction, or stroke/CVA")
variable_summary <- update_label(variable = "comp_any_including_mortality", label = "Post-operative complication")
variable_summary <- update_label(variable = "comp_any_excluding_mortality", label = "Post-operative complication")
#variable_summary <- update_label(variable = "", label = "")
######################################
## Return the Cleaned Up Data Files ##
######################################
## Remove empty factor levels
for(i in 1:ncol(dataset)){
if( is.factor(dataset[,i]) ){
dataset[,i] <- droplevels(dataset[,i])
}
}
rm(i)
## Return the cleaned up dataset and variable summary
return(
list(
Data = dataset,
VariableSummary = variable_summary)
)
}
UKBiostatCIRCL/UK-Biostat-CIRCL documentation built on June 30, 2022, 4:42 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions nsqip_clean_data
Documented in nsqip_clean_data
#### NSQIP Raw Data Cleaning ####
#' Cleaning the raw National Surgical Quality Improvement Program (NSQIP) Dataset
#'
#' This function is designed to perform routine data cleaning routines when using the raw NSQIP data.
#' Cleaning functions include missing data recoding, variable categorization/dichotomization, converting string
#' entries to usable formats, and changing variable classes. This function outputs a cleaned dataframe and variable summary.
#'
#' NOTE: This function contains the sub-function update_label() to change variable labels.
#'
#' @param dataset the filtered and merged NSQIP data frame outputted by the "nsqip_subset_pufs" function
#'
#' @export
## The function
nsqip_clean_data <- function(dataset = dat){
###########################
## Handle Missing Values ##
###########################
## Replacing Null's with NA's
dataset[ dataset == "NULL" ] <- NA
## Replacing -99's with NA's
dataset[ dataset == -99 ] <- NA
## Replacing "Unknown" with NA
dataset[ dataset == "Unknown" ] <- NA
## Replacing "Unknown" with NA
dataset[ dataset == "None assigned" ] <- NA
## Remove unused/erroneous variables
dataset$admsyr <- NULL
#############################
## Create Factor Variables ##
#############################
## Sex
dataset$sex <- factor(x = dataset$sex,
levels = c("male","female"),
labels = c("Male","Female"))
## Race
dataset$racenew <- factor(x = dataset$racenew,
levels = c("White","Black or African American","Asian","American Indian or Alaska Native",
"Native Hawaiian or Pacific Islander","Unknown/Not Reported"))
## Deep incisional SSI
dataset$wndinfd <- factor(x = dataset$wndinfd,
levels = c("No Complication","Deep Incisional SSI"))
## Renal failure
dataset$oprenafl <- factor(x = dataset$oprenafl,
levels = c("No Complication","Acute Renal Failure"))
## DVT
dataset$othdvt <- factor(x = dataset$othdvt,
levels = c("No Complication","DVT Requiring Therap","DVT Requiring Therapy"),
labels = c("No Complication","DVT Requiring Therapy","DVT Requiring Therapy"))
## Cardiac arrest
dataset$cdarrest <- factor(x = dataset$cdarrest,
levels = c("No Complication","Cardiac Arrest Requiring CPR"))
## Myocardial infarction
dataset$cdmi <- factor(x = dataset$cdmi,
levels = c("No Complication","Myocardial Infarction"))
## Wound closure
dataset$woundclosure <- factor(x = dataset$woundclosure,
levels = c("All layers of incision (deep and superficial) fully closed",
"Only deep layers closed; superficial left open",
"No layers of incision are surgically closed"))
## Factor variables that don't require re-ordering of levels
vars <- c("ventilat","hxcopd","ascites","hxchf","hypermed","renafail","dialysis","discancr","wndinf","steroid","wtloss","bleeddis","transfus","prsepis",
"smoke","caseid","inout","emergncy","fnstatus2","wndclas","orgspcssi","supinfec","dehis","oupneumo","othbleed","urninfec","reintub","failwean",
"othsysep","othseshock","renainsf","pulembol","cnscva")
for(i in vars){
dataset[, i] <- factor(x = dataset[, i])
}
rm(i, vars)
############################
## Dichotomize Lab Values ##
############################
## Pre-Operative Albumin < 3
dataset$pralbum_lt3 <- factor(x = as.numeric(dataset$pralbum < 3),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Alkaline Phosphatase > 125
dataset$pralkph_gt125 <- factor(x = as.numeric(dataset$pralkph > 125),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Bilirubin > 1
dataset$prbili_gt1 <- factor(x = as.numeric(dataset$prbili > 1),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative BUN > 40
dataset$prbun_gt40 <- factor(x = as.numeric(dataset$prbun > 40),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Creatinine > 1.2
dataset$prcreat_gt1p2 <- factor(x = as.numeric(dataset$prcreat > 1.2),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Hematocrit < 30%
dataset$prhct_lt30 <- factor(x = as.numeric(dataset$prhct < 30),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Hematocrit > 45%
dataset$prhct_gt45 <- factor(x = as.numeric(dataset$prhct > 45),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Platelets < 150,000
dataset$prplate_lt150 <- factor(x = as.numeric(dataset$prplate < 150),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Platelets > 400,000
dataset$prplate_gt400 <- factor(x = as.numeric(dataset$prplate > 400),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative SGOT > 40
dataset$prsgot_gt40 <- factor(x = as.numeric(dataset$prsgot > 40),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Sodium < 135
dataset$prsodm_lt135 <- factor(x = as.numeric(dataset$prsodm < 135),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative Sodium > 145
dataset$prsodm_gt145 <- factor(x = as.numeric(dataset$prsodm > 145),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative White Blood Count <= 4,500
dataset$prwbc_lte4p5 <- factor(x = as.numeric(dataset$prwbc <= 4.5),
levels = c(0, 1),
labels = c("No","Yes"))
## Pre-Operative White Blood Count > 11,000
dataset$prwbc_gt11 <- factor(x = as.numeric(dataset$prwbc > 11),
levels = c(0, 1),
labels = c("No","Yes"))
##########################
## Create New Variables ##
##########################
## Impute the age values of "90+" to "90"
dataset$age_90imp <- as.character(dataset$age)
dataset$age_90imp[ dataset$age_90imp == "90+" ] <- "90"
dataset$age_90imp <- as.numeric(dataset$age_90imp)
## Preop sepsis
dataset$prsepis_any <- factor(x = as.numeric(dataset$prsepis %in% c("Sepsis","SIRS")),
levels = c(0, 1),
labels = c("No","Yes"))
## BMI
if( is.factor(dataset$weight) ){
dataset$weight <- as.numeric( as.character(dataset$weight) )
}
if( is.factor(dataset$height) ){
dataset$height <- as.numeric( as.character(dataset$height) )
}
dataset$weight[ dataset$weight == -99 ] <- NA
dataset$height[ dataset$height == -99 ] <- NA
dataset$bmi <- ((dataset$weight / dataset$height / dataset$height) * 703)
## BMI > 30
dataset$bmi_gte30 <- factor(x = as.numeric(dataset$bmi >= 30),
levels = c(0, 1),
labels = c("BMI < 30","BMI 30+"))
## BMI Group
dataset$bmi_group <- cut(dataset$bmi,
c(0, 18.5, 25, 30, 35, 40, 9999),
include.lowest = TRUE,
right = FALSE)
dataset$bmi_group <- factor(x = dataset$bmi_group,
levels = c("[0,18.5)","[18.5,25)","[25,30)","[30,35)","[35,40)","[40,1e+04]"),
labels = c("<18.5","18.5-24.9","25.0-29.9","30.0-34.9","35.0-39.9","40.0+"))
## ASA Class Group
dataset$asaclas[ dataset$asaclas == "None assigned" ] <- NA
dataset$asaclas_group[ dataset$asaclas == "1-No Disturb" | dataset$asaclas == "2-Mild Disturb" ] <- "I-II"
dataset$asaclas_group[ dataset$asaclas == "3-Severe Disturb" ] <- "III"
dataset$asaclas_group[ dataset$asaclas == "4-Life Threat" | dataset$asaclas == "5-Moribund" ] <- "IV-V"
dataset$asaclas_group[ is.na(dataset$asaclas) ] <- NA
dataset$asaclas_group <- factor(x = dataset$asaclas_group,
levels = c("I-II","III","IV-V"))
## ASA Class > III
dataset$asaclas_gtIII <- factor(x = as.numeric(dataset$asaclas_group == "IV-V"),
levels = c(0, 1),
labels = c("ASA Class I, II, or III","ASA Class IV or V"))
## Non-white race
dataset$racenew_nonwhite <- factor(x = dataset$racenew,
levels = c("White","Black or African American","Asian","American Indian or Alaska Native",
"Native Hawaiian or Pacific Islander","Unknown/Not Reported"),
labels = c("White","Non-White","Non-White","Non-White",
"Non-White","Unknown/Not Reported"))
## Total hospital LOS
dataset$hospital_los_total <- dataset$tothlos
## Pre-operative hospital LOS
dataset$hospital_los_preop <- dataset$htooday
## Post-operative hospital LOS
dataset$hospital_los_postop <- (dataset$hospital_los_total - dataset$hospital_los_preop)
## Any diabetes
dataset$diabetes_present <- factor(x = as.numeric(dataset$diabetes %in% c("INSULIN","NON-INSULIN","ORAL")),
levels = c(0, 1),
labels = c("No Diabetes","Diabetes"))
## Any dyspnea
dataset$dyspnea_present <- factor(x = as.numeric(dataset$dyspnea %in% c("MODERATE EXERTION","AT REST")),
levels = c(0, 1),
labels = c("No Dyspnea","Dyspnea"))
##################################################
## Create Post-Operative Complication Variables ##
##################################################
## Standalone complications
dataset$comp_wound_disruption <- factor(x = dataset$dehis,
levels = c("No Complication","Wound Disruption"),
labels = c("No","Yes"))
dataset$comp_pneumonia <- factor(x = dataset$oupneumo,
levels = c("No Complication","Pneumonia"),
labels = c("No","Yes"))
dataset$comp_transfusion <- factor(x = dataset$othbleed,
levels = c("No Complication","Transfusions/Intraop/Postop"),
labels = c("No","Yes"))
dataset$comp_uti <- factor(x = dataset$urninfec,
levels = c("No Complication","Urinary Tract Infection"),
labels = c("No","Yes"))
## Individual complication variables
dataset$comp_ssi_super <- factor(x = dataset$supinfec,
levels = c("No Complication","Superficial Incisional SSI"),
labels = c("No","Yes"))
dataset$comp_ssi_deep_incis <- factor(x = dataset$wndinfd,
levels = c("No Complication","Deep Incisional SSI"),
labels = c("No","Yes"))
dataset$comp_ssi_organ_space <- factor(x = dataset$orgspcssi,
levels = c("No Complication","Organ/Space SSI"),
labels = c("No","Yes"))
## Any SSI
dataset$comp_ssi_any <- apply(X = dataset[, c("comp_ssi_super","comp_ssi_deep_incis","comp_ssi_organ_space") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") |
(!is.na(x[3]) & x[3] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) & is.na(x[3]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_ssi_any <- factor(x = dataset$comp_ssi_any)
## Individual complication variables
dataset$comp_unpl_intub <- factor(x = dataset$reintub,
levels = c("No Complication","Unplanned Intubation"),
labels = c("No","Yes"))
dataset$comp_vent48 <- factor(x = dataset$failwean,
levels = c("No Complication","On Ventilator greater than 48 Hours"),
labels = c("No","Yes"))
## Unplanned intubation or ventilator > 48 hours
dataset$comp_unpl_intub_or_vent48 <- apply(X = dataset[, c("comp_unpl_intub","comp_vent48") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_unpl_intub_or_vent48 <- factor(x = dataset$comp_unpl_intub_or_vent48)
## Individual complication variables
dataset$comp_sepsis <- factor(x = dataset$othsysep,
levels = c("No Complication","Sepsis"),
labels = c("No","Yes"))
dataset$comp_septic_shock <- factor(x = dataset$othseshock,
levels = c("No Complication","Septic Shock"),
labels = c("No","Yes"))
## Sepsis or septic shock
dataset$comp_sepsis_or_septic_shock <- apply(X = dataset[, c("comp_sepsis","comp_septic_shock") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_sepsis_or_septic_shock <- factor(x = dataset$comp_sepsis_or_septic_shock)
## Individual complication variables
dataset$comp_renal_insuff <- factor(x = dataset$renainsf,
levels = c("No Complication","Progressive Renal Insufficiency"),
labels = c("No","Yes"))
dataset$comp_renal_failure <- factor(x = dataset$oprenafl,
levels = c("No Complication","Acute Renal Failure"),
labels = c("No","Yes"))
## Renal insufficiency or failure
dataset$comp_renal_insuff_or_failure <- apply(X = dataset[, c("comp_renal_insuff","comp_renal_failure") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_renal_insuff_or_failure <- factor(x = dataset$comp_renal_insuff_or_failure)
## Individual complication variables
dataset$comp_pulm_embolism <- factor(x = dataset$pulembol,
levels = c("No Complication","Pulmonary Embolism"),
labels = c("No","Yes"))
dataset$comp_dvt <- factor(x = dataset$othdvt,
levels = c("No Complication","DVT Requiring Therapy"),
labels = c("No","Yes"))
## Pulmonary embolism or DVT
dataset$comp_pulm_embolism_or_dvt <- apply(X = dataset[, c("comp_pulm_embolism","comp_dvt") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") ){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) ){
return(NA)
} else{ return("No") }
} )
dataset$comp_pulm_embolism_or_dvt <- factor(x = dataset$comp_pulm_embolism_or_dvt)
## Individual complication variables
dataset$comp_cardiac_arrest <- factor(x = dataset$cdarrest,
levels = c("No Complication","Cardiac Arrest Requiring CPR"),
labels = c("No","Yes"))
dataset$comp_mi <- factor(x = dataset$cdmi,
levels = c("No Complication","Myocardial Infarction"),
labels = c("No","Yes"))
dataset$comp_stroke <- factor(x = dataset$cnscva,
levels = c("No Complication","Stroke/CVA"),
labels = c("No","Yes"))
## Cardiac arrest, MI, or stroke
dataset$comp_cardiac_arrest_mi_stroke <- apply(X = dataset[, c("comp_cardiac_arrest","comp_mi","comp_stroke") ],
MARGIN = 1,
FUN = function(x){
if( (!is.na(x[1]) & x[1] == "Yes") |
(!is.na(x[2]) & x[2] == "Yes") |
(!is.na(x[3]) & x[3] == "Yes")){
return("Yes")
} else if( is.na(x[1]) & is.na(x[2]) & is.na(x[3]) ){
return(NA)
} else{ return("No") }
} )
## Mortality
dataset$comp_mortality <- as.numeric( !is.na(dataset$dopertod) & dataset$dopertod >= 0 & dataset$dopertod <= 30)
dataset$comp_mortality[ is.na(dataset$comp_mortality) ] <- 0
dataset$comp_mortality <- factor(x = dataset$comp_mortality,
levels = c(0, 1),
labels = c("No","Yes"))
## Any complication, including mortality
dataset$comp_any_including_mortality <- apply(X = dataset[, c("comp_wound_disruption","comp_pneumonia","comp_transfusion",
"comp_uti","comp_ssi_super","comp_ssi_deep_incis",
"comp_ssi_organ_space","comp_unpl_intub","comp_vent48",
"comp_sepsis","comp_septic_shock","comp_renal_insuff",
"comp_renal_failure","comp_pulm_embolism","comp_dvt",
"comp_cardiac_arrest","comp_mi","comp_stroke",
"comp_mortality") ],
MARGIN = 1,
FUN = function(x){
if( sum(is.na(x)) == length(x) ){
return(NA)
} else if( sum(x == "Yes", na.rm = TRUE) > 0 ){
return("Yes")
} else{ return("No") }
} )
## Any complication, excluding mortality
dataset$comp_any_excluding_mortality <- apply(X = dataset[, c("comp_wound_disruption","comp_pneumonia","comp_transfusion",
"comp_uti","comp_ssi_super","comp_ssi_deep_incis",
"comp_ssi_organ_space","comp_unpl_intub","comp_vent48",
"comp_sepsis","comp_septic_shock","comp_renal_insuff",
"comp_renal_failure","comp_pulm_embolism","comp_dvt",
"comp_cardiac_arrest","comp_mi","comp_stroke") ],
MARGIN = 1,
FUN = function(x){
if( sum(is.na(x)) == length(x) ){
return(NA)
} else if( sum(x == "Yes", na.rm = TRUE) > 0 ){
return("Yes")
} else{ return("No") }
} )
##########################################
## Create a Variable Summary Data Frame ##
##########################################
## Create a variable summary data frame
variable_summary <- data.frame(Variable = names(dataset),
Label = names(dataset),
Variable_Type = NA,
Variable_Category = NA,
stringsAsFactors = FALSE)
## Identify continuous variables
vars_cont <- c("pufyear","workrvu","age","age_90imp","admyr","operyr","height","weight","bmi",
"dprna","dprbun","dprcreat","dpralbum","dprbili","dprsgot","dpralkph","dprwbc","dprhct","dprplate","dprptt",
"dprpt","dprinr",
"prsodm","prbun","prcreat","pralbum","prbili","prsgot","pralkph","prwbc","prhct","prplate","prptt","prinr",
"prpt",
"otherwrvu1","otherwrvu2","otherwrvu3","otherwrvu4","otherwrvu5","otherwrvu6","otherwrvu7","otherwrvu8",
"otherwrvu9","otherwrvu10",
"conwrvu1","conwrvu2","conwrvu3","conwrvu4","conwrvu5","conwrvu6","conwrvu7","conwrvu8","conwrvu9","conwrvu10",
"optime","hdisdt","yrdeath","tothlos","htooday","hospital_los_total","hospital_los_preop","hospital_los_postop",
"dsupinfec","dwndinfd","dorgspcssi","ddehis","doupneumo","dreintub","dpulembol","dfailwean","drenainsf",
"doprenafl","durninfec","dcnscva","dcdarrest","dcdmi","dothbleed","dothdvt","dothsysep","dothseshock",
"retorpodays","retor2podays","readmpodays1","readmpodays2","readmpodays3","readmpodays4","readmpodays5",
"dothcdiff")
## Identify ordinal variables
vars_ord <- c("wndclas","asaclas_group","bmi_group","admqtr",
"nsupinfec","nwndinfd","norgspcssi","ndehis","noupneumo","nreintub","npulembol","nfailwean","nrenainsf",
"noprenafl","nurninfec","ncnscva","ncdarrest","ncdmi","nothbleed","nothdvt","nothsysep","nothseshock",
"nothcdiff","fnstatus2","woundclosure")
## Define variable types
variable_summary$Variable_Type <- "Categorical"
variable_summary$Variable_Type[ variable_summary$Variable %in% vars_cont ] <- "Continuous"
variable_summary$Variable_Type[ variable_summary$Variable %in% vars_ord ] <- "Ordinal"
## Identify common patient characteristic variables
vars_pat <- c("age_90imp","sex","racenew","ethnicityhispanic","bmi","bmi_gte30",
"smoke","diabetes_present","dyspnea_present","fnstatus2",
"ventilat","hxcopd","ascites","hxchf","hypermed","renafail",
"dialysis","discancr","wndinf","steroid","wtloss","bleeddis",
"transfus","prsepis_any")
## Identify common surgical charactistic variables
vars_surg <- c("optime","cpt","asaclas_group","wndclas","inout","electsurg","emergncy","woundclosure")
## Identify lab value variables, continuous
vars_labs_cont <- c("prsodm","prbun","prcreat","pralbum","prbili","prsgot","pralkph","prwbc","prhct","prplate","prptt","prinr","prpt")
## Identify lab value variables, dichotomous
vars_labs_dich <- c("pralbum_lt3","pralkph_gt125","prbili_gt1","prbun_gt40","prcreat_gt1p2","prhct_lt30","prhct_gt45","prplate_lt150","prplate_gt400",
"prsgot_gt40","prsodm_lt135","prsodm_gt145","prwbc_lte4p5","prwbc_gt11")
## Identify outcome variables
vars_outs <- c("hospital_los_total","comp_mortality","comp_any_including_mortality","comp_any_excluding_mortality",
"comp_ssi_any","comp_wound_disruption","comp_pneumonia","comp_transfusion","comp_uti","comp_unpl_intub_or_vent48",
"comp_sepsis_or_septic_shock","comp_renal_insuff_or_failure","comp_pulm_embolism_or_dvt","comp_cardiac_arrest_mi_stroke")
## Define the variable categories
variable_summary$Variable_Category <- "Other"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_pat ] <- "Patient Characteristics"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_surg ] <- "Surgical Characteristics"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_labs_cont ] <- "Continous Lab Values"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_labs_dich ] <- "Dichotomous Lab Values"
variable_summary$Variable_Category[ variable_summary$Variable %in% vars_outs ] <- "Outcomes"
## A function to update variable labels
update_label <- function(df = variable_summary, variable, label){
variable_summary$Label[ variable_summary$Variable == variable ] <- label
return(variable_summary)
}
## Manually update variable labels
variable_summary <- update_label(variable = "age_90imp", label = "Age (years)")
variable_summary <- update_label(variable = "sex", label = "Female")
variable_summary <- update_label(variable = "racenew", label = "Race")
variable_summary <- update_label(variable = "ethnicityhispanic", label = "Hispanic ethnicity")
variable_summary <- update_label(variable = "bmi", label = "Body mass index (kg/m2)")
variable_summary <- update_label(variable = "bmi_gte30", label = "Body mass index >= 30 kg/m2")
variable_summary <- update_label(variable = "bmi_group", label = "Body mass index (kg/m2)")
variable_summary <- update_label(variable = "cpt", label = "Primary CPT code")
variable_summary <- update_label(variable = "inout", label = "Outpatient")
variable_summary <- update_label(variable = "electsurg", label = "Elective surgery, patient coming from home")
variable_summary <- update_label(variable = "smoke", label = "Current smoker within one year")
variable_summary <- update_label(variable = "fnstatus2", label = "Functional status prior to surgery")
variable_summary <- update_label(variable = "ventilat", label = "Ventilator dependent")
variable_summary <- update_label(variable = "hxcopd", label = "History of severe COPD")
variable_summary <- update_label(variable = "ascites", label = "Ascites with 30 days prior to surgery")
variable_summary <- update_label(variable = "hxchf", label = "Congestive heart failure within 30 days prior to surgery")
variable_summary <- update_label(variable = "hypermed", label = "Hypertension requiring medication")
variable_summary <- update_label(variable = "renafail", label = "Acute renal failure (pre-op)")
variable_summary <- update_label(variable = "dialysis", label = "Currently on dialysis (pre-op)")
variable_summary <- update_label(variable = "discancr", label = "Disseminated cancer")
variable_summary <- update_label(variable = "wndinf", label = "Open wound (with or without infection)")
variable_summary <- update_label(variable = "steroid", label = "Steroid/Immunosuppressant usre for a chronic condition")
variable_summary <- update_label(variable = "wtloss", label = "Greater than 10% loss of body weight in the 6 months prior to surgery")
variable_summary <- update_label(variable = "bleeddis", label = "Bleeding disorder")
variable_summary <- update_label(variable = "transfus", label = "Pre-operative transfusion (RBC wihtin 72 hours prior to surgery start time)")
variable_summary <- update_label(variable = "diabetes_present", label = "Diabetes mellitus requiring therapy")
variable_summary <- update_label(variable = "dyspnea_present", label = "Dyspnea at rest or with moderate exertion")
variable_summary <- update_label(variable = "prsepis_any", label = "Sepsis within 48 hours prior to surgery")
variable_summary <- update_label(variable = "emergncy", label = "Emergency case")
variable_summary <- update_label(variable = "wndclas", label = "Wound classification")
variable_summary <- update_label(variable = "optime", label = "Total operation time (minutes)")
variable_summary <- update_label(variable = "woundclosure", label = "Surgical wound closure")
variable_summary <- update_label(variable = "asaclas_group", label = "ASA classification")
variable_summary <- update_label(variable = "hospital_los_total", label = "Lenght of total hospital stay (days)")
variable_summary <- update_label(variable = "prsodm", label = "Pre-operative serum sodium")
variable_summary <- update_label(variable = "prsodm_lt135", label = "Pre-operative serum sodium < 135")
variable_summary <- update_label(variable = "prsodm_gt145", label = "Pre-operative serum sodium > 145")
variable_summary <- update_label(variable = "prbun", label = "Pre-operative blood urea nitrogen (BUN)")
variable_summary <- update_label(variable = "prbun_gt40", label = "Pre-operative blood urea nitrogen (BUN) > 40")
variable_summary <- update_label(variable = "prcreat", label = "Pre-operative serum creatinine")
variable_summary <- update_label(variable = "prcreat_gt1p2", label = "Pre-operative serum creatinine > 1.2")
variable_summary <- update_label(variable = "pralbum", label = "Pre-operative serum albumin")
variable_summary <- update_label(variable = "pralbum_lt3", label = "Pre-operative serum albumin < 3")
variable_summary <- update_label(variable = "prbili", label = "Pre-operative total bilirubin")
variable_summary <- update_label(variable = "prbili_gt1", label = "Pre-operative total bilirubin > 1")
variable_summary <- update_label(variable = "prsgot", label = "Pre-operative serum glutamic-oxaloacetic transaminase (SGOT)")
variable_summary <- update_label(variable = "prsgot_gt40", label = "Pre-operative serum glutamic-oxaloacetic transaminase (SGOT) > 40")
variable_summary <- update_label(variable = "pralkph", label = "Pre-operative alkaline phosphatase")
variable_summary <- update_label(variable = "pralkph_gt125", label = "Pre-operative alkaline phosphatase > 125")
variable_summary <- update_label(variable = "prwbc", label = "Pre-operative white blood cell (WBC) count")
variable_summary <- update_label(variable = "prwbc_lte4p5", label = "Pre-operative white blood cell (WBC) count <= 4.5")
variable_summary <- update_label(variable = "prwbc_gt11", label = "Pre-operative white blood cell (WBC) count > 11")
variable_summary <- update_label(variable = "prhct", label = "Pre-operative hematocrit")
variable_summary <- update_label(variable = "prhct_lt30", label = "Pre-operative hematocrit < 30")
variable_summary <- update_label(variable = "prhct_gt45", label = "Pre-operative hematocrit > 45")
variable_summary <- update_label(variable = "prplate", label = "Pre-operative platelet count")
variable_summary <- update_label(variable = "prplate_lt150", label = "Pre-operative platelet count < 150")
variable_summary <- update_label(variable = "prplate_gt400", label = "Pre-operative platelet count > 400")
variable_summary <- update_label(variable = "prptt", label = "Pre-operative partial thromboplastin time (PTT)")
variable_summary <- update_label(variable = "prinr", label = "Pre-operative international normalized ratio (INR) of prothrombin time (PT) values")
variable_summary <- update_label(variable = "prpt", label = "Pre-operative prothrombin time (PT)")
variable_summary <- update_label(variable = "comp_mortality", label = "Death within 30 days or surgery")
variable_summary <- update_label(variable = "comp_wound_disruption", label = "Post-operative wound disruption")
variable_summary <- update_label(variable = "comp_pneumonia", label = "Post-operative pneumonia")
variable_summary <- update_label(variable = "comp_transfusion", label = "Intra-/post-operative transfusion (RBC within the first 72 hours of surgery start time)")
variable_summary <- update_label(variable = "comp_uti", label = "Post-operative UTI")
variable_summary <- update_label(variable = "comp_ssi_super", label = "Post-operative superficial incisional SSI")
variable_summary <- update_label(variable = "comp_ssi_deep_incis", label = "Post-operative deep incisional SSI")
variable_summary <- update_label(variable = "comp_ssi_organ_space", label = "Post-operative organ/space SSI")
variable_summary <- update_label(variable = "comp_ssi_any", label = "Post-operative SSI (superficial incisional, deep incisional, or organ/space)")
variable_summary <- update_label(variable = "comp_unpl_intub", label = "Post-operative unplanned intubation")
variable_summary <- update_label(variable = "comp_vent48", label = "Post-operative venilator use > 48 hours")
variable_summary <- update_label(variable = "comp_unpl_intub_or_vent48", label = "Post-operative unplanned intubation or ventilator use > 48 hours")
variable_summary <- update_label(variable = "comp_sepsis", label = "Post-operative sepsis")
variable_summary <- update_label(variable = "comp_septic_shock", label = "Post-operative septic shock")
variable_summary <- update_label(variable = "comp_sepsis_or_septic_shock", label = "Post-operative sepsis or septic shock")
variable_summary <- update_label(variable = "comp_renal_insuff", label = "Post-operative progressive renal insufficiency")
variable_summary <- update_label(variable = "comp_renal_failure", label = "Post-operative acute renal failure")
variable_summary <- update_label(variable = "comp_renal_insuff_or_failure", label = "Post-operative progressive renal insufficiency or acute renal failure")
variable_summary <- update_label(variable = "comp_pulm_embolism", label = "Post-operative pulmonary embolism")
variable_summary <- update_label(variable = "comp_dvt", label = "Post-operative DVT requiring therapy")
variable_summary <- update_label(variable = "comp_pulm_embolism_or_dvt", label = "Post-operative pulmonary embolism or DVT requiring therapy")
variable_summary <- update_label(variable = "comp_cardiac_arrest", label = "Intra-/post-operative cardiac arrest requiring CPR")
variable_summary <- update_label(variable = "comp_mi", label = "Intra-/post-operative myocardial infarction")
variable_summary <- update_label(variable = "comp_stroke", label = "Post-operative stroke/CVA")
variable_summary <- update_label(variable = "comp_cardiac_arrest_mi_stroke", label = "Post-operative cardiac arrest requiring CPR, myocardial infarction, or stroke/CVA")
variable_summary <- update_label(variable = "comp_any_including_mortality", label = "Post-operative complication")
variable_summary <- update_label(variable = "comp_any_excluding_mortality", label = "Post-operative complication")
#variable_summary <- update_label(variable = "", label = "")
######################################
## Return the Cleaned Up Data Files ##
######################################
## Remove empty factor levels
for(i in 1:ncol(dataset)){
if( is.factor(dataset[,i]) ){
dataset[,i] <- droplevels(dataset[,i])
}
}
rm(i)
## Return the cleaned up dataset and variable summary
return(
list(
Data = dataset,
VariableSummary = variable_summary)
)
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.