rm(list=ls())
#install.packages("survival",destdir="D:/R_total/Rbag")
options("scipen"=100,"digits"=4) # Disable scientific notation
library(haven)
data <- read_sav("E:/data/work/hospital.sav")


attach(data)
names(data)

# Part A: Initial Data Cleaning
# 1. Remove outliers
table(data$cancer_diagnosis)
data$cancer_diagnosis <- gsub('[\n]','',data$cancer_diagnosis)
# Remove leading and trailing spaces
data$cancer_diagnosis <- trimws(data$cancer_diagnosis, which = c("both","left","right"), whitespace = "[ \t\r\n]")
table(data$cancer_diagnosis)
newdata <- data.frame(table(data$cancer_diagnosis))

# 2. Check missing value ratio
pMiss <- function(x){sum(is.na(x))/length(x)*100}
apply(data,2,pMiss)

# Part B: Statistical Analysis
# B-1. Select lung cancer population
# Lung cancer variable
data$lung_a <- ifelse(
  grepl("lung",data$cancer_diagnosis)==TRUE & grepl("cancer",data$cancer_diagnosis)==TRUE,"100", "0"
)
data$lung_a
table(data$lung_a)

data1 <- data[data$lung_a=="100",]
table(data1$cancer_diagnosis)

data1$lung_b <- ifelse(
  data1$cancer_diagnosis=="lung mass, breast cancer" | data1$cancer_diagnosis=="colorectal cancer lung metastasis","non", "1"
)
data1$lung_b
table(data1$lung_b)

# Export all lung cancer data
data_lung <- data1[data1$lung_b=="1",]
table(data_lung$cancer_diagnosis)

data_lung <- data_lung[!duplicated(data_lung$case_number),]

############################## Table 1
# A. Table 1: General Description
# 1.
shapiro.test(data_lung$age)
shapiro.test(data_lung$age[data_lung$planned_admission==0])
shapiro.test(data_lung$age[data_lung$planned_admission==1])

wilcox.test(data_lung$age ~ data_lung$planned_admission)

# 2.
shapiro.test(data_lung$BMI)
shapiro.test(data_lung$BMI[data_lung$planned_admission==0])
shapiro.test(data_lung$BMI[data_lung$planned_admission==1])

wilcox.test(data_lung$BMI ~ data_lung$planned_admission)

# 3.
shapiro.test(data_lung$SOFA_score)
shapiro.test(data_lung$SOFA_score[data_lung$planned_admission==0])
shapiro.test(data_lung$SOFA_score[data_lung$planned_admission==1])

wilcox.test(data_lung$SOFA_score ~ data_lung$planned_admission)

#
shapiro.test(data_lung$APACHE_score)
shapiro.test(data_lung$APACHE_score[data_lung$planned_admission==0])
shapiro.test(data_lung$APACHE_score[data_lung$planned_admission==1])

wilcox.test(data_lung$APACHE_score ~ data_lung$planned_admission)

library(tableone)
myVars <- c("age","gender","BMI",
            
            "SOFA_score","APACHE_score",
            "survival_90d","ICU_death","hospital_death",
            "mechanical_ventilation","sedation_treatment","conventional_oxygen",
            "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis")
catVars <- c("gender","mechanical_ventilation","sedation_treatment","conventional_oxygen",
             "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis",
             "survival_90d","ICU_death","hospital_death") # Define categorical variables
nonvar <- c("age","BMI","SOFA_score","APACHE_score")  # Specify which variables are non-normally distributed

table <- CreateTableOne(vars = myVars, ## Variables to display
                        factorVars = catVars,   # Categorical variables
                        strata = "planned_admission", # Stratification
                        data = data_lung,
                        addOverall = TRUE) # Add overall column

table1 <- print(table, # Create table function (including conditions 1.2)
                showAllLevels=TRUE, # Show all variables
                catDigits = 1,contDigits = 1,pDigits = 3, # Number of decimal places
                nonnormal = nonvar,
                quote = FALSE, # Do not show quotes
                noSpaces = TRUE,        # Delete spaces for alignment
                printToggle = TRUE) # Display output results

#write.csv(table1, "E:/outcome/6.csv",fileEncoding="GBK")

names(data_lung)

# A
Table <- xtabs(~data_lung$planned_admission + data_lung$ICU_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$planned_admission + data_lung$hospital_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$planned_admission + data_lung$survival_90d, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

# B
Table <- xtabs(~data_lung$planned_admission + data_lung$gender, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

# D

Table <- xtabs(~data_lung$planned_admission + data_lung$mechanical_ventilation, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$planned_admission + data_lung$sedation_treatment, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$planned_admission + data_lung$conventional_oxygen, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

# F
Table <- xtabs(~data_lung$planned_admission + data_lung$shock_diagnosis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$planned_admission + data_lung$sepsis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$planned_admission + data_lung$respiratory_failure, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$planned_admission + data_lung$ARDS, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$planned_admission + data_lung$acute_kidney_injury, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

############################## Table 2
names(data_lung)
table(data_lung$transfer_source)

data_lung$transfer_source_new <- ifelse(
  data_lung$transfer_source == 4,3,data_lung$transfer_source
)
data_lung$transfer_source_new
table(data_lung$transfer_source_new)

# A. Table 1: General Description
# 1.

shapiro.test(data_lung$age)
shapiro.test(data_lung$age[data_lung$transfer_source_new==1])
shapiro.test(data_lung$age[data_lung$transfer_source_new==2])
shapiro.test(data_lung$age[data_lung$transfer_source_new==3])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(age ~ transfer_source_new, data=data_lung)

# 2.
shapiro.test(data_lung$BMI)
shapiro.test(data_lung$BMI[data_lung$transfer_source_new==1])
shapiro.test(data_lung$BMI[data_lung$transfer_source_new==2])
shapiro.test(data_lung$BMI[data_lung$transfer_source_new==3])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(BMI ~ transfer_source_new, data=data_lung)

# 3.
shapiro.test(data_lung$SOFA_score)
shapiro.test(data_lung$SOFA_score[data_lung$transfer_source_new==1])
shapiro.test(data_lung$SOFA_score[data_lung$transfer_source_new==2])
shapiro.test(data_lung$SOFA_score[data_lung$transfer_source_new==3])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(SOFA_score ~ transfer_source_new, data=data_lung)

#
shapiro.test(data_lung$APACHE_score)
shapiro.test(data_lung$APACHE_score[data_lung$transfer_source_new==1])
shapiro.test(data_lung$APACHE_score[data_lung$transfer_source_new==2])
shapiro.test(data_lung$APACHE_score[data_lung$transfer_source_new==3])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(APACHE_score ~ transfer_source_new, data=data_lung)

library(tableone)
myVars <- c("age","gender","BMI",
            
            "SOFA_score","APACHE_score",
            "survival_90d","ICU_death","hospital_death",
            "mechanical_ventilation","sedation_treatment","conventional_oxygen",
            "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis")
catVars <- c("gender","mechanical_ventilation","sedation_treatment","conventional_oxygen",
             "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis",
             "survival_90d","ICU_death","hospital_death") # Define categorical variables
nonvar <- c("age","BMI","SOFA_score","APACHE_score")  # Specify which variables are non-normally distributed

table <- CreateTableOne(vars = myVars, ## Variables to display
                        factorVars = catVars,   # Categorical variables
                        strata = "transfer_source_new", # Stratification
                        data = data_lung,
                        addOverall = TRUE) # Add overall column

table1 <- print(table, # Create table function (including conditions 1.2)
                showAllLevels=TRUE, # Show all variables
                catDigits = 1,contDigits = 1,pDigits = 3, # Number of decimal places
                nonnormal = nonvar,
                quote = FALSE, # Do not show quotes
                noSpaces = TRUE,        # Delete spaces for alignment
                printToggle = TRUE) # Display output results

#write.csv(table1, "E:/outcome/9.csv",fileEncoding="GBK")

names(data_lung)

# A
Table <- xtabs(~data_lung$transfer_source_new + data_lung$ICU_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

Table <- xtabs(~data_lung$transfer_source_new + data_lung$hospital_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

Table <- xtabs(~data_lung$transfer_source_new + data_lung$survival_90d, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

# B
Table <- xtabs(~data_lung$transfer_source_new + data_lung$gender, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

# C
Table <- xtabs(~data_lung$transfer_source_new + data_lung$mechanical_ventilation, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$transfer_source_new + data_lung$sedation_treatment, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$transfer_source_new + data_lung$conventional_oxygen, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

# D
Table <- xtabs(~data_lung$transfer_source_new + data_lung$shock_diagnosis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

Table <- xtabs(~data_lung$transfer_source_new + data_lung$sepsis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

Table <- xtabs(~data_lung$transfer_source_new + data_lung$respiratory_failure, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$transfer_source_new + data_lung$ARDS, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

Table <- xtabs(~data_lung$transfer_source_new + data_lung$acute_kidney_injury, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
fisher.test(Table)

############################## Table 3
names(data_lung)
table(data_lung$transfer_source)

# Grouping
data_lung$treatment_history_group <- ifelse(
  data_lung$radiation_history == 1 | data_lung$chemotherapy_history == 1,0,1
)
data_lung$treatment_history_group
table(data_lung$treatment_history_group)

# A. Table 1: General Description
# 1.
shapiro.test(data_lung$age)
shapiro.test(data_lung$age[data_lung$treatment_history_group==0])
shapiro.test(data_lung$age[data_lung$treatment_history_group==1])

wilcox.test(data_lung$age ~ data_lung$treatment_history_group)

# 2.
shapiro.test(data_lung$BMI)
shapiro.test(data_lung$BMI[data_lung$treatment_history_group==0])
shapiro.test(data_lung$BMI[data_lung$treatment_history_group==1])

wilcox.test(data_lung$BMI ~ data_lung$treatment_history_group)

# 3.
shapiro.test(data_lung$SOFA_score)
shapiro.test(data_lung$SOFA_score[data_lung$treatment_history_group==0])
shapiro.test(data_lung$SOFA_score[data_lung$treatment_history_group==1])

wilcox.test(data_lung$SOFA_score ~ data_lung$treatment_history_group)

#
shapiro.test(data_lung$APACHE_score)
shapiro.test(data_lung$APACHE_score[data_lung$treatment_history_group==0])
shapiro.test(data_lung$APACHE_score[data_lung$treatment_history_group==1])

wilcox.test(data_lung$APACHE_score ~ data_lung$treatment_history_group)

library(tableone)
myVars <- c("age","gender","BMI",
            
            "SOFA_score","APACHE_score",
            "survival_90d","ICU_death","hospital_death",
            "mechanical_ventilation","sedation_treatment","conventional_oxygen",
            "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis")
catVars <- c("gender","mechanical_ventilation","sedation_treatment","conventional_oxygen",
             "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis",
             "survival_90d","ICU_death","hospital_death") # Define categorical variables
nonvar <- c("age","BMI","SOFA_score","APACHE_score")  # Specify which variables are non-normally distributed

table <- CreateTableOne(vars = myVars, ## Variables to display
                        factorVars = catVars,   # Categorical variables
                        strata = "treatment_history_group", # Stratification
                        data = data_lung,
                        addOverall = TRUE) # Add overall column

table1 <- print(table, # Create table function (including conditions 1.2)
                showAllLevels=TRUE, # Show all variables
                catDigits = 1,contDigits = 1,pDigits = 3, # Number of decimal places
                nonnormal = nonvar,
                quote = FALSE, # Do not show quotes
                noSpaces = TRUE,        # Delete spaces for alignment
                printToggle = TRUE) # Display output results

#write.csv(table1, "E:/outcome/51.csv",fileEncoding="GBK")

names(data_lung)

# A
Table <- xtabs(~data_lung$treatment_history_group + data_lung$ICU_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$treatment_history_group + data_lung$hospital_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$treatment_history_group + data_lung$survival_90d, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

# B
Table <- xtabs(~data_lung$treatment_history_group + data_lung$gender, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

# D

Table <- xtabs(~data_lung$treatment_history_group + data_lung$mechanical_ventilation, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$treatment_history_group + data_lung$sedation_treatment, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$treatment_history_group + data_lung$conventional_oxygen, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

# F
Table <- xtabs(~data_lung$treatment_history_group + data_lung$shock_diagnosis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$treatment_history_group + data_lung$sepsis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$treatment_history_group + data_lung$respiratory_failure, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$treatment_history_group + data_lung$ARDS, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$treatment_history_group + data_lung$acute_kidney_injury, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

############################## Table 4
names(data_lung)
table(data_lung$other_infection_medications)
# Grouping
data_lung$fungal_gp <- ifelse(
  data_lung$echinocandins == 1 | data_lung$triazoles == 1,1,0
)
data_lung$fungal_gp
table(data_lung$fungal_gp)

data_lung$bacterial_gp <- ifelse(
  data_lung$carbapenems == 1 | data_lung$beta_lactam == 1 | data_lung$glycopeptides == 1 | data_lung$tigecycline == 1,1,0
)
data_lung$bacterial_gp
table(data_lung$bacterial_gp)

data_lung$anti_infection_gp <- ifelse(
  data_lung$fungal_gp == 0 & data_lung$bacterial_gp == 0,0,
  ifelse(
    data_lung$fungal_gp == 1 & data_lung$bacterial_gp == 1,1,2
  )
)
data_lung$anti_infection_gp
table(data_lung$anti_infection_gp)

h <- data_lung[data_lung$anti_infection_gp == 0,]

table(h$fungal_gp)
table(h$bacterial_gp)
table(h$other_infection_medications)

data_lung$anti_infection_group <- ifelse(
  data_lung$anti_infection_gp == 0 & data_lung$other_infection_medications == 0,0,
  ifelse(
    data_lung$anti_infection_gp == 1,1,2
  )
)
data_lung$anti_infection_group
table(data_lung$anti_infection_group)

# A. Table 1: General Description
# 1.
shapiro.test(data_lung$age)
shapiro.test(data_lung$age[data_lung$anti_infection_group == 0])
shapiro.test(data_lung$age[data_lung$anti_infection_group == 1])
shapiro.test(data_lung$age[data_lung$anti_infection_group == 2])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(age ~ anti_infection_group, data=data_lung)

# 2.
shapiro.test(data_lung$BMI)
shapiro.test(data_lung$BMI[data_lung$anti_infection_group == 0])
shapiro.test(data_lung$BMI[data_lung$anti_infection_group == 1])
shapiro.test(data_lung$BMI[data_lung$anti_infection_group == 2])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(BMI ~ anti_infection_group, data=data_lung)

# 3.
shapiro.test(data_lung$SOFA_score)
shapiro.test(data_lung$SOFA_score[data_lung$anti_infection_group == 0])
shapiro.test(data_lung$SOFA_score[data_lung$anti_infection_group == 1])
shapiro.test(data_lung$SOFA_score[data_lung$anti_infection_group == 2])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(SOFA_score ~ anti_infection_group, data=data_lung)

#
shapiro.test(data_lung$APACHE_score)
shapiro.test(data_lung$APACHE_score[data_lung$anti_infection_group == 0])
shapiro.test(data_lung$APACHE_score[data_lung$anti_infection_group == 1])
shapiro.test(data_lung$APACHE_score[data_lung$anti_infection_group == 2])

# Kruskal-Wallis H test - non-parametric test (3 groups) - differences between groups
kruskal.test(APACHE_score ~ anti_infection_group, data=data_lung)

library(tableone)
myVars <- c("age","gender","BMI",
            
            "SOFA_score","APACHE_score",
            "survival_90d","ICU_death","hospital_death",
            "mechanical_ventilation","sedation_treatment","conventional_oxygen",
            "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis")
catVars <- c("gender","mechanical_ventilation","sedation_treatment","conventional_oxygen",
             "sepsis","respiratory_failure","ARDS","acute_kidney_injury","shock_diagnosis",
             "survival_90d","ICU_death","hospital_death") # Define categorical variables
nonvar <- c("age","BMI","SOFA_score","APACHE_score")  # Specify which variables are non-normally distributed

table <- CreateTableOne(vars = myVars, ## Variables to display
                        factorVars = catVars,   # Categorical variables
                        strata = "anti_infection_group", # Stratification
                        data = data_lung,
                        addOverall = TRUE) # Add overall column

table1 <- print(table, # Create table function (including conditions 1.2)
                showAllLevels=TRUE, # Show all variables
                catDigits = 1,contDigits = 1,pDigits = 3, # Number of decimal places
                nonnormal = nonvar,
                quote = FALSE, # Do not show quotes
                noSpaces = TRUE,        # Delete spaces for alignment
                printToggle = TRUE) # Display output results

#write.csv(table1, "E:/outcome/123.csv",fileEncoding="GBK")

names(data_lung)

# A
Table <- xtabs(~data_lung$anti_infection_group + data_lung$ICU_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$anti_infection_group + data_lung$hospital_death, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$anti_infection_group + data_lung$survival_90d, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

# B
Table <- xtabs(~data_lung$anti_infection_group + data_lung$gender, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

# D

Table <- xtabs(~data_lung$anti_infection_group + data_lung$mechanical_ventilation, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

Table <- xtabs(~data_lung$anti_infection_group + data_lung$sedation_treatment, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$anti_infection_group + data_lung$conventional_oxygen, data=data_lung)
Table
prop.table(Table,1) # Calculate row frequency ratio
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

# F
Table <- xtabs(~data_lung$anti_infection_group + data_lung$shock_diagnosis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$anti_infection_group + data_lung$sepsis, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$anti_infection_group + data_lung$respiratory_failure, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$anti_infection_group + data_lung$ARDS, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction

ab <- chisq.test(Table,correct = F) # Chi-square test correct = F No correction
round(ab$p.value,3)    # Extract p-value

Table <- xtabs(~data_lung$anti_infection_group + data_lung$acute_kidney_injury, data=data_lung)
Table
chisq.test(Table)$expected   # View expected frequencies
chisq.test(Table,correct = F) # Chi-square test correct = F No correction