fall.Rmd

---
title: "fall"
output:
  word_document: default
  html_notebook: default
  pdf_document: default
  
---
a.table3
table2



```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
library(dplyr)
library(tidyverse)
library(rio)
library(tidyr)
library(dplyr)
library(stringr)
library(tableone)
## Loading the csv file
url <- "https://raw.githubusercontent.com/titco/titco-I/master/titco-I-full-dataset-v1.csv"
titco_fall <- import(url) %>% as_tibble()

## Filter all fall patients

fall <- filter(titco_fall, moi == "Fall" )


## to find the sample size
nrow(fall)

## Find no and % of males and females
m <- fall$sex == "Male"
n.m <- sum(m)
p.m <- round(mean(m) * 100)
f <- fall$sex == "Female"
n.f <- sum(f)
p.f <- round(mean(f)*100)
## Summarize age
ages  <- fall$age
ages <- as.numeric(ages)

summary(ages)
## Break age into groups
fall["agegroup"] <- agegroup <- cut(ages,
                                     breaks = c(0, 17, 25, 45, 65, 85),
                                     labels =  c("less than 18","18-24", "25-44", "45-64", "65+"))
table1 <- table(agegroup)
## Find percentage transferred
t <- fall$tran == "Yes"
n.t <- sum(t,na.rm=T)
p.t <- round(mean(t,na.rm=T)*100)
## Find mode of transport 
am <- fall$mot == "Ambulance"
n.am <- sum(am,na.rm=T)
p.am <- round(mean(am,na.rm=T)*100)
## Find percentage of Type of injury
ti <- fall$ti == "Blunt"
n.ti <- sum(ti,na.rm=T)
p.ti <- round(mean(ti,na.rm=T)*100)
## Create GCS group
a.gcs <- fall$gcs_t_1
## Break GCS into groups
fall["gcsgroup"] <- gcsgroup <- cut(a.gcs,
                                     breaks = c(2, 8, 13, 15),
                                     labels = c( "3-8", "9-13", "14-15"))
gcs.table <- table(gcsgroup)
## Find number of surgery done
tos <- fall$tos 
surg1<- ifelse(tos == 0, 0,
              ifelse(tos == 999, 0, 1))
surgery1 <- fall["Surgery"] <- surg1 == 1
n.s <- sum(surgery1, na.rm = TRUE)
p.s <- round(mean(surgery1, na.rm = TRUE)*100)

## Find mortality
died <- fall$died == "Yes"
n.died <- sum(died, na.rm = TRUE)
p.died <- round(mean(died, na.rm = TRUE) * 100)
## Find discharged against medical advice
dama <- fall$dama == "Yes"
n.dama <- sum(dama,na.rm = T)
p.dama <- round(mean(dama,na.rm = T)*100)
## Creating Table for TBI patients demography using table one
table1.data <- with(fall,
                    data.frame("Age" = age,
                               "Sex" = sex,
                               "Age group" = agegroup,
                               
                               "Type of injury" = ti,
                               "Mode of transport" = mot,
                               "Transferred" = ifelse(tran == "Yes", "Transferred", "Direct"),
                               "Systolic blood pressure" = sbp_1,
                               "Heart rate" = hr_1,
                               "Saturation" = spo2_1,
                               "Respiratory rate" = rr_1,
                               "Glasgow coma scale" = gcsgroup,
                               "Surgery" = ifelse(tos == "0", "Conservative", "Operative"),
                               "Mortality" = ifelse(died == "Yes", "Died", "Survived"),
                          
                               check.names = FALSE
                               ))
a.table3 <- CreateTableOne(data=table1.data)
a.table3 <- knitr::kable(print(a.table3, caption = "Table 1. Characteristics of Fall Patients", showAllLevels = TRUE, printToggle = FALSE))



summary(fall$niss)
issbreaks <- c(1,9,16,25,109)
isslabels <- c("Mild","Moderate","Severe","Profound")
fall$isspv <-cut(fall$iss,
                breaks = issbreaks,
                right = FALSE,
                labels = isslabels)
table2 <-  table(fall$isspv)



## delay

fall_d <- select(fall,doi, toi, doar,toar,dodd,todd,died,age,sex,tran)
## removing all NA's
fall_d <- fall_d[complete.cases(fall_d),]
nrow(fall_d)

col2 <- c("doi","toi")
y <- fall_d$date_time_injury <- apply (fall_d[,col2],1, paste,collapse ="")

date.time.injury <- as.POSIXct(y)


col3 <- c("doar","toar")
x <- fall_d$date_time_arrival <- apply (fall_d[,col3],1,paste,collapse ="")
date.time.arrival <- as.POSIXct(x)

delay <-difftime(date.time.arrival, date.time.injury)

fall_d["delay"] <- delay_hours <- as.numeric(delay/60)

summary(delay_hours)
m.delay <- median(delay_hours)
q1.delay <- quantile(delay_hours,0.25)
q3.delay <- quantile(delay_hours,0.75)

range.delay <- range(delay_hours)


## length of hospital stay

col3 <- c("doar","toar")
x <- fall_d$date_time_arrival <- apply (fall_d[,col3],1,paste,collapse ="")
date.time.arrival <- as.POSIXct(x)

col4 <- c("dodd","todd")
z <- fall_d$date_time_discharge <- apply(fall_d[,col4],1,paste,collapse = "")
data.time.discharge <- as.POSIXct(z)

los <-difftime(data.time.discharge, date.time.arrival)

fall_d["LOS"] <- los_hrs <- as.numeric(delay/60)
summary(los_hrs)





```