---
title: "EPOC Paper One Supplementary analysis"
author: "Helen Badge and Ian Harris"
date: '`r paste(format(Sys.Date(), "%d %b %Y"))`'
output:
#  pdf_document:
#    toc: yes
  html_document:
    toc: yes
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
knitr::opts_chunk$set(tidy.opts=list(width.cutoff=80),tidy=TRUE)
```

  
#### Load packages
```{r load packages}
library(ggplot2)
library(dplyr)
library(tidyverse)
library(caTools)
library(StepReg)
library(finalfit)
library(tableone)
library(caret)
library(MASS)
library(DataExplorer)
library(dplyr)
library(statsr)
library(tidyr)
library(skimr)
library(caTools)
library(gmodels)
library(ROCR)
library(Amelia)
library(chron)
```
#### Load data
```{r load data}
data <- read.csv("epoc_with_fu_data_R_LATEST.csv", na.strings = c("", "999", "NA", "N/A", "NR"))
mydata_supp <- as.data.frame(data)

# Recode id_number= 886 due to error in data 
  # TG_Ab_1_compliant_missing_data: recode to FALSE
  # TG_Ab_1_compliant: recode to TRUE because  ab_ceph = 1 AND ab_vanc = 1 AND others not missing 
mydata_supp <- mydata_supp %>%
  mutate(TG_Ab_1_compliant_missing_data = ifelse(id_number == 886, FALSE, TG_Ab_1_compliant_missing_data),
         TG_Ab_1_compliant = ifelse(id_number == 886, TRUE, TG_Ab_1_compliant))

# remove ineligible and those with no follow up
mydata_supp <- mydata_supp %>% 
  filter((optout_fu > 0) & (eligibility == 1))

```
## Describe additional demographic, acute care and prophylaxis variables 
```{r Paper 1: Supplementary descriptive code}
## Supplemetary results 
## Calculate median and IQR for continuous variables (Manuscript Table One)
mydata_supp %>%
summarise(mean_age = mean(mydata_supp$age), median_age = median(mydata_supp$age),iqr_age = IQR(mydata_supp$age), sd_age = sd(mydata_supp$age), count = n())

mydata_supp %>%
summarise(mean_bmi = mean(mydata_supp$bmi), median_bmi = median(mydata_supp$bmi),iqr_bmi = IQR(mydata_supp$bmi), sd_bmi = sd(mydata_supp$bmi), count = n())

mydata_supp %>%
summarise(mean_los = mean(mydata_supp$los), median_los = median(mydata_supp$los),iqr_los = IQR(mydata_supp$los), sd_los = sd(mydata_supp$los), count = n())

mydata_supp$surg_duration <- as.numeric(mydata_supp$surg_duration)

mydata_supp %>%
summarise(mean_surg_duration = mean(mydata_supp$surg_duration), median_surg_duration = median(mydata_supp$surg_duration),iqr_surg_duration = IQR(mydata_supp$surg_duration), sd_surg_duration = sd(mydata_supp$surg_duration), count = n())

#Calculate descriptives for other variables re indications for vancomycin (Manuscript Table One & line 269)
table(mydata_supp$ab_contra, dnn = "Contraindication for an Antibiotic", exclude = NULL)
table(mydata_supp$ab_allergy_type, dnn = "Type of antibiotic allergy", exclude = NULL)
table(mydata_supp$ab_resist,  dnn = "History Antibiotic Resistant infections", exclude = NULL)
table(mydata_supp$ab_resist_spec,  dnn = "Type Antibiotic Resistant infections", exclude = NULL)
table(mydata_supp$ab_hosp_adm,  dnn = "Previous hospital admission", exclude = NULL)

## Supplementary data, Surgeon and discharge destination  
table(mydata_supp$dc_dest,  dnn = "Discharge destination", exclude = NULL)
table(mydata_supp$surgeon, dnn = "Surgeon", exclude = NULL)

# Supplementary data - Describe Antibiotic prophylaxis
table(mydata_supp$ab_ceph,  dnn = "Cephazolin used", exclude = NULL)
table(mydata_supp$ab_fluclox,  dnn = "Flucloxacillin", exclude = NULL)
table(mydata_supp$ab_vanc,  dnn = "vancomycin", exclude = NULL)
table(mydata_supp$ab_genta,  dnn = "gentamicin", exclude = NULL)
table(mydata_supp$ab_clinda,  dnn = "Clindamycin", exclude = NULL)
table(mydata_supp$ab_astre,  dnn = "Astronam", exclude = NULL)
table(mydata_supp$ab_fluro,  dnn = "Fluoroquinolone", exclude = NULL)
table(mydata_supp$ab_alt1,  dnn = "Alternate antibiotic - One", exclude = NULL)
table(mydata_supp$ab_alt2,  dnn = "Alternate antibiotic - Two", exclude = NULL)
table(mydata_supp$ab_alt1_name,  dnn = "Names alternate antibiotic - One", exclude = NULL)
table(mydata_supp$ab_alt2_name,  dnn = "Names alternate antibiotic - Two", exclude = NULL)

# Number prophylactic medications taken by each person
#VTE - transform yesno variables to numeric 
mydata_supp <- mydata_supp %>%
mutate(vte_clex_count = vte_clex)
mydata_supp$vte_clex_count <- as.numeric(mydata_supp$vte_clex_count)

mydata_supp <- mydata_supp %>%
mutate(vte_frag_count = vte_frag)
mydata_supp$vte_frag_count <- as.numeric(mydata_supp$vte_frag_count)

mydata_supp <- mydata_supp %>%
mutate(vte_fondap_count = vte_fondap)
mydata_supp$vte_fondap_count <- as.numeric(mydata_supp$vte_fondap_count)

mydata_supp <- mydata_supp %>%
mutate(vte_riva_count = vte_riva)
mydata_supp$vte_riva_count <- as.numeric(mydata_supp$vte_riva_count)

mydata_supp <- mydata_supp %>%
mutate(vte_dabig_count = vte_dabig)
mydata_supp$vte_dabig_count <- as.numeric(mydata_supp$vte_dabig_count)

mydata_supp <- mydata_supp %>%
mutate(vte_hep_count = vte_hep)
mydata_supp$vte_hep_count <- as.numeric(mydata_supp$vte_hep_count)

mydata_supp <- mydata_supp %>%
mutate(vte_unfhep_count = vte_unfhep)
mydata_supp$vte_unfhep_count <- as.numeric(mydata_supp$vte_unfhep_count)

mydata_supp <- mydata_supp %>%
mutate(vte_aspir_count = vte_aspir)
mydata_supp$vte_aspir_count <- as.numeric(mydata_supp$vte_aspir_count)
table(mydata_supp$vte_aspir_count)

mydata_supp <- mydata_supp %>%
mutate(vte_warf_count = vte_warf)
mydata_supp$vte_warf_count <- as.numeric(mydata_supp$vte_warf_count)

mydata_supp <- mydata_supp %>%
mutate(vte_alt1_count = vte_alt1)
mydata_supp$vte_alt1_count <- as.numeric(mydata_supp$vte_alt1_count)

#Count number VTE prophylaxis medications taken 
mydata_supp <-mydata_supp %>%
mutate(vte_prophylmeds_count = vte_clex_count + vte_frag_count + vte_fondap_count + vte_riva_count + vte_dabig_count + vte_hep_count + vte_unfhep_count + vte_aspir_count + vte_warf_count + vte_alt1_count)
table(mydata_supp$vte_prophylmeds_count, dnn = "Number prophylactic VTE medications taken by each person")

#AB - transform yesno variables to numeric 
mydata_supp <- mydata_supp %>%
mutate(ab_ceph_count = ab_ceph)
mydata_supp$ab_ceph_count <- as.numeric(mydata_supp$ab_ceph_count)

mydata_supp <- mydata_supp %>%
mutate(ab_fluclox_count = ab_fluclox)
mydata_supp$ab_fluclox_count <- as.numeric(mydata_supp$ab_fluclox_count)

mydata_supp <- mydata_supp %>%
mutate(ab_vanc_count = ab_vanc)
mydata_supp$ab_vanc_count <- as.numeric(mydata_supp$ab_vanc_count)

mydata_supp <- mydata_supp %>%
mutate(ab_genta_count = ab_genta)
mydata_supp$ab_genta_count <- as.numeric(mydata_supp$ab_genta_count)

mydata_supp <- mydata_supp %>%
mutate(ab_clinda_count = ab_clinda)
mydata_supp$ab_clinda_count <- as.numeric(mydata_supp$ab_clinda_count)

mydata_supp <- mydata_supp %>%
mutate(ab_astre_count = ab_astre)
mydata_supp$ab_astre_count <- as.numeric(mydata_supp$ab_astre_count)

mydata_supp <- mydata_supp %>%
mutate(ab_fluro_count = ab_fluro)
mydata_supp$ab_fluro_count <- as.numeric(mydata_supp$ab_fluro_count)

mydata_supp <- mydata_supp %>%
mutate(ab_alt1_count = ab_alt1)
mydata_supp$ab_alt1_count <- as.numeric(mydata_supp$ab_alt1_count)

mydata_supp <- mydata_supp %>%
mutate(ab_alt2_count = ab_alt2)
mydata_supp$ab_alt2_count <- as.numeric(mydata_supp$ab_alt2_count)

#Count number prophylaxis antibiotics taken 
mydata_supp <-mydata_supp %>%
mutate(ab_prophylmeds_count = ab_ceph_count + ab_fluclox_count + ab_vanc_count + ab_genta_count + ab_clinda_count + ab_astre_count + ab_fluro_count + ab_alt1_count + ab_alt2_count)
table(mydata_supp$ab_prophylmeds_count, dnn = "Number prophylactic antibiotics taken by each person")

# Supplementary data - Describe preop cardiac meds
table(mydata_supp$cc_hd_meds,  dnn = "Medication for heart disease", exclude = NULL)

table(mydata_supp$cardiac_aspirin, dnn = "Preoperative medication commonly also used for VTE prophylaxis- Aspirin", exclude = NULL)
table(mydata_supp$cardiac_warfarin , dnn = "Preoperative medication commonly also used for VTE prophylaxis- Warfarin", exclude = NULL)
table(mydata_supp$cardiac_heparin, dnn = "Preoperative medication commonly also used for VTE prophylaxis- Heparin", exclude = NULL)
table(mydata_supp$cardiac_drug_other, dnn = "Preoperative medication commonly also used for VTE prophylaxis- Other type", exclude = NULL)
table(mydata_supp$cardiac_drug_other_type, dnn = "Name of other preoperative medication commonly also used for VTE prophylaxis", exclude = NULL)

## Supplementary data - Describe VTE prophylaxis 
table(mydata_supp$vte_clex,  dnn = "Clexane used", exclude = NULL)
table(mydata_supp$vte_frag,  dnn = "Fragmin used", exclude = NULL)
table(mydata_supp$vte_fondap,  dnn = "Fondaparinux used", exclude = NULL)
table(mydata_supp$vte_riva,  dnn = "Rivaroxaban used", exclude = NULL)
table(mydata_supp$vte_dabig,  dnn = "Dabigatran etexilate used", exclude = NULL)
table(mydata_supp$vte_hep,  dnn = "Heparin used", exclude = NULL)
table(mydata_supp$vte_unfhep,  dnn = "Unfractionated heparin used", exclude = NULL)
table(mydata_supp$vte_aspir,  dnn = "Apirin used", exclude = NULL)
table(mydata_supp$vte_warf,  dnn = "Warfarin used", exclude = NULL)
table(mydata_supp$vte_alt1,  dnn = "Other chemoprophyalxis used - one", exclude = NULL)
table(mydata_supp$vte_alt1_name,  dnn = "Name 0ther chemoprophyalxis used - one", exclude = NULL)

#Supplementary data: Mechanical VTE prophylaxis
table(mydata_supp$vte_icp_foot,  dnn = "Mechnical VTE prophylaxis: Foot pumps used", exclude = NULL)
table(mydata_supp$vte_icp_calf,  dnn = "Mechnical VTE prophylaxis: Intermittent calf compressors(ICP) used", exclude = NULL)
table(mydata_supp$vte_gcs,  dnn = "Mechnical VTE prophylaxis: Graduated compression stockins (GCS) used", exclude = NULL)
table(mydata_supp$vte_mech_yesno,  dnn = "HOw many people used any mechnical VTE prophylaxis", exclude = NULL)

#Day first mobilised
table(mydata_supp$vte_1stmobil_day0,  dnn = "First mobilised day 0", exclude = NULL)
```
# Supplementary Data - Describe details of complication outcomes

```{r Paper 1: Supplementary complications}
##Supplementary Data - Describe details of complication outcomes
#Cause of death 
#need to add cc_death_yesno = 1 AND surgery related - 10 people coded with a cause of death not related to surgery who didn't die  - added 1014 cause of death = surgery related (PE) into REDcap
table(mydata_supp$cc_death_surgeryrelated,  dnn = "Cause of death related to surgery", exclude = NULL)

#Supplementary data re VTE complications
#0, PE and DVT | 1, PE only | 2, Symptomatic DVT | 3, Assymptomatic DVT | 4, DVT, unknown if symptomatic | 5, NONE, No VTE |  999, Not stated/Unknown
table(mydata_supp$cc_vte_first_type,  dnn = "Type 1st VTE", exclude = NULL)
table(mydata_supp$cc_vte_second_type,  dnn = "Type 2nd VTE", exclude = NULL)
table(mydata_supp$cc_vte_third_type,  dnn = "Type 3rd VTE", exclude = NULL)
table(mydata_supp$cc_vte_fourth_type,  dnn = "Type 4th VTE", exclude = NULL)

table(mydata_supp$cc_vte_followup,  dnn = "Time period for First VTE", exclude = NULL)
table(mydata_supp$cc_vte_second_followup,  dnn = "Time period for Second VTE", exclude = NULL)
table(mydata_supp$cc_vte_third_followup,  dnn = "Time period for Third VTE", exclude = NULL)
table(mydata_supp$cc_vte_fourth_followup, dnn = "Time period for Fourth VTE", exclude = NULL)
# Scores: 0, During Acute admission | 1, Between acute discharge and 35 days | 2, Between 36-90 days | 3, Between 91-365 days | 999, Not stated/Unknown

#Bleeding outcomes - Joint and non-joint
table(mydata_supp$cc_type_jtbleed,  dnn = "Joint related bleeding complications", exclude = NULL)
table(mydata_supp$cc_type_jtbleed_type,  dnn = "Severity of joint related bleeding complications", exclude = NULL)
table(mydata_supp$cc_type_otherbleed,  dnn = "Non-joint related bleeding complications", exclude = NULL)
table(mydata_supp$cc_type_otherbleed_type,  dnn = "Severity of non-joint related bleeding complications", exclude = NULL)

#Timeframe for bleeding is a checkbox variable - creates separate variable for each possible code: 0, During Acute admission | 1, Between acute discharge and 35 days | 2, Between 36-90 days | 3, Between 91-365 days | 999, Not stated/Unknown
table(mydata_supp$cc_type_jtbleed_time___0,  dnn = " Experienced joint related bleeding complications during acute admission", exclude = NULL)
table(mydata_supp$cc_type_jtbleed_time___1,  dnn = "Experienced joint related bleeding complications betweeen acute discharge and 35 days", exclude = NULL)
table(mydata_supp$cc_type_jtbleed_time___2,  dnn = "Experienced joint related bleeding complications betweeen 36 - 90 days", exclude = NULL)
table(mydata_supp$cc_type_jtbleed_time___3,  dnn = "Experienced joint related bleeding complications betweeen 91 - 365 days", exclude = NULL)

#Timeframes non-joint bleeding
table(mydata_supp$cc_type_otherbleed_time___0,  dnn = "Non-joint related bleeding complications duing acute admission", exclude = NULL)
table(mydata_supp$cc_type_otherbleed_time___1,  dnn = "Non-joint related bleeding complications betwen acute discharge and 35 days", exclude = NULL)
table(mydata_supp$cc_type_otherbleed_time___2,  dnn = "Non-joint related bleeding complications betwen 36-90 days", exclude = NULL)
table(mydata_supp$cc_type_otherbleed_time___3,  dnn = "Non-joint related bleeding complications betwen 91-365 days", exclude = NULL)
# Severity: 0, Type 1 (no healthcare action)| 1, Type 2 (overt bleed, healthcare provided) | 2, Type 3 (major intervention, readmission/reoperation) | 3, Type 5 (Fatal bleeding) | 999, Not stated/Unknown

# Describe breakdown of joint infection 
table(mydata_supp$cc_type_jtinfect_type, dnn = "Type Joint infection",  exclude = NULL)

# Number of readmissions 
table(mydata_supp$cc_readmiss_yesno,  dnn = "Had 1st Hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_2nd_yesno,  dnn = "Had 2nd hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_3rd_yesno,  dnn = "Had 3rd hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_4th_yesno,  dnn = "Had 4th hospital readmission", exclude = NULL)

#Describe reasons for readmission.
## Note scores: 0,  Readmitted, Joint specific reason (including VTE) | 1, Readmitted re surgery related but NOT joint specific | 999, Not stated/Unknown. Reasons for non-joint readmission and reoperation summarised from raw data, multiple fields provide information but mainly free text, 24 readmissions, 16 reoperations
table(mydata_supp$cc_readmiss_first_reas,  dnn = "Reason for 1st hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_2nd_reas,  dnn = "Reason for 2nd hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_3rd_reas,  dnn = "Reason for 3rd hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_4th_reas, dnn = "Reason for 4th hospital readmission", exclude = NULL)

#Reasons for joint related readmission at each followup
#"cc_readmiss_first_reas_jt"" breaks joint specific reasons into: 1, PE | 2, DVT | 3, Deep joint infection (reoperation) | 4, Joint/wound Infection (readmitted and IV antibiotics) | 5, Wound/operated leg Bleeding | 6, Joint/wound bleeding | 7,  Joint Dislocation | 8, Wound dehiscence | 9, severe joint stiffness |10, Other joint related reason | 999, Not stated/ Unknown
table(mydata_supp$cc_readmiss_first_reas_jt,  dnn = "Joint related reasons for first hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmis_second_reas_jt, dnn = "Joint related reasons for second hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_third_reas_jt,  dnn = "Joint related reasons for third hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_fourth_reas_jt,  dnn = "Joint related reasons for fourth hospital readmission", exclude = NULL)

#Time period for readmissions - would need to do crosstabs to look at by joint/nonjoint
table(mydata_supp$cc_readmiss_1st_followup,  dnn = "Time period when experienced 1st hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_2nd_followup,  dnn = "Time period when experienced 2nd hospital readmission", exclude = NULL)
table(mydata_supp$cc_readmiss_3rd_followup,  dnn = "Time period when experienced 3rd hospital readmission", exclude = NULL)

# Number of reoperations 
table(mydata_supp$cc_reop_yesno,  dnn = "Had 1st Hospital reoperation", exclude = NULL)
table(mydata_supp$cc_reop_2nd_yesno,  dnn = "Had 2nd hospital reoperation", exclude = NULL)
table(mydata_supp$cc_reop_3rd_yesno,  dnn = "Had 3rd hospital reoperation", exclude = NULL)
table(mydata_supp$cc_reop_4th_yesno,  dnn = "Had 4th hospital reoperation", exclude = NULL)

# Describe reasons for reoperation.
#Note scores = 0,  Reoperation, Joint specific reason (including VTE) | 1, Operation related to surgery related but NOT joint specific | 2, Reason for operation unclear | 999, Not stated/Unknown
table(mydata_supp$cc_reop_yesno,  dnn = "Reoperation", exclude = NULL)
table(mydata_supp$cc_reop_first_reas,  dnn = "Reason for 1st reoperation", exclude = NULL)
table(mydata_supp$cc_reop_2nd_reas,  dnn = "Reason for 2nd reoperation", exclude = NULL)
table(mydata_supp$cc_reop_3rd_reas,  dnn = "Reason for 3rd reoperation", exclude = NULL)
table(mydata_supp$cc_reop_4th_reas, dnn = "Reason for 4th reoperation",  exclude = NULL)

#Reasons for joint related reoperation at each followup
# "cc_reop_first_reas_jt" can break down into joint specific reasons: 1, Joint revision | 2, Joint Wash out | 3, Joint MUA | 4, Open reduction of dislocated joint | 5, Closed reduction of dislocated joint | 6, Joint, repair wound dehiscence | 7, Joint, exploratory , e.g. severe pain  | 8, Joint, manage bleeding  | 9, Other joint related reoperation | 10, Joint specific soft tissue repair | 11, Surgery re vascualr injury | 12, Surgery re compartment syndrome | 13, Amputation | 14, Internal fixation of fracture | 15, CLosed reduction of dislocated joint | 999, Not specified / Unknown
table(mydata_supp$cc_reop_first_reas_jt, dnn = "Joint related reasons for first reoperation", exclude = NULL)
table(mydata_supp$cc_reop_2nd_reas_jt, dnn = "Joint related reasons for second reoperation", exclude = NULL)
table(mydata_supp$cc_reop_3rd_reas_jt, dnn = "Joint related reasons for third reoperation", exclude = NULL)
table(mydata_supp$cc_reop_4th_reas_jt, dnn = "Joint related reasons for fourth reoperation", exclude = NULL)

#Time period for reoperations
table(mydata_supp$cc_reop_yesno,  dnn = "Reopyesno", exclude = NULL)
table(mydata_supp$cc_reop_1st_followup,  dnn = "Time period when experienced 1st reoperation", exclude = NULL)
table(mydata_supp$cc_reop_2nd_followup,  dnn = "Time period when experienced 2nd reoperation", exclude = NULL)
table(mydata_supp$cc_reop_3rd_followup ,  dnn = "Time period when experienced 3rd reoperation", exclude = NULL)
```