@GenMatch_final.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Load packages and data\n",
    "#library(dplyr)\n",
    "library(tableone)\n",
    "library(MatchIt)\n",
    "library(Matching)\n",
    "library(pROC)\n",
    "library(ROCR)\n",
    "library(ggplot2)\n",
    "library(epitools)\n",
    "library(plyr)\n",
    "library(ipw)\n",
    "library(sandwich)\n",
    "library(survey)\n",
    "library(reshape2)\n",
    "library(ggplot2)\n",
    "library(gridExtra)\n",
    "library(xlsx)\n",
    "library(optmatch)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>ID</th><th scope=col>hospital_number</th><th scope=col>female</th><th scope=col>age</th><th scope=col>pim2</th><th scope=col>pelod</th><th scope=col>p_bloodculture</th><th scope=col>comorbidity</th><th scope=col>rf_pneumonia</th><th scope=col>rf_sepsis</th><th scope=col>⋯</th><th scope=col>mortality_30day</th><th scope=col>mortality_28day</th><th scope=col>mortality_100day</th><th scope=col>ventilation_free_day</th><th scope=col>picu_free_days</th><th scope=col>fio2</th><th scope=col>map</th><th scope=col>amplitude</th><th scope=col>OI_cat</th><th scope=col>ICUduration</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>1001    </td><td>1       </td><td>1       </td><td>0.02    </td><td>87.3    </td><td>42      </td><td>1       </td><td>0       </td><td>0       </td><td>1       </td><td>⋯       </td><td>1       </td><td>1       </td><td>1       </td><td> 0      </td><td> 0      </td><td> NA     </td><td>NA      </td><td>      NA</td><td>1       </td><td> 3      </td></tr>\n",
       "\t<tr><td>1003    </td><td>1       </td><td>0       </td><td>0.47    </td><td> 2.8    </td><td>11      </td><td>0       </td><td>0       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td>0       </td><td>21      </td><td>26      </td><td> NA     </td><td>NA      </td><td>      NA</td><td>1       </td><td> 3      </td></tr>\n",
       "\t<tr><td>1005    </td><td>1       </td><td>1       </td><td>0.10    </td><td> 0.6    </td><td> 1      </td><td>0       </td><td>0       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td>0       </td><td>21      </td><td>20      </td><td> NA     </td><td>NA      </td><td>      NA</td><td>2       </td><td> 9      </td></tr>\n",
       "\t<tr><td>1006    </td><td>1       </td><td>1       </td><td>1.72    </td><td> 8.2    </td><td>12      </td><td>0       </td><td>0       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td>0       </td><td> 7      </td><td> 0      </td><td>100     </td><td>23      </td><td>70.00000</td><td>4       </td><td>59      </td></tr>\n",
       "\t<tr><td>1008    </td><td>1       </td><td>1       </td><td>2.92    </td><td> 6.8    </td><td> 2      </td><td>0       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td>0       </td><td>27      </td><td>24      </td><td> NA     </td><td>NA      </td><td>      NA</td><td>1       </td><td> 5      </td></tr>\n",
       "\t<tr><td>1009    </td><td>1       </td><td>0       </td><td>0.70    </td><td> 2.8    </td><td>10      </td><td>0       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>1       </td><td>1       </td><td>1       </td><td> 0      </td><td> 0      </td><td>100     </td><td>32      </td><td>43.66667</td><td>1       </td><td> 9      </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllllllllllllllllllllllllllllllllllll}\n",
       " ID & hospital\\_number & female & age & pim2 & pelod & p\\_bloodculture & comorbidity & rf\\_pneumonia & rf\\_sepsis & ⋯ & mortality\\_30day & mortality\\_28day & mortality\\_100day & ventilation\\_free\\_day & picu\\_free\\_days & fio2 & map & amplitude & OI\\_cat & ICUduration\\\\\n",
       "\\hline\n",
       "\t 1001     & 1        & 1        & 0.02     & 87.3     & 42       & 1        & 0        & 0        & 1        & ⋯        & 1        & 1        & 1        &  0       &  0       &  NA      & NA       &       NA & 1        &  3      \\\\\n",
       "\t 1003     & 1        & 0        & 0.47     &  2.8     & 11       & 0        & 0        & 1        & 0        & ⋯        & 0        & 0        & 0        & 21       & 26       &  NA      & NA       &       NA & 1        &  3      \\\\\n",
       "\t 1005     & 1        & 1        & 0.10     &  0.6     &  1       & 0        & 0        & 1        & 0        & ⋯        & 0        & 0        & 0        & 21       & 20       &  NA      & NA       &       NA & 2        &  9      \\\\\n",
       "\t 1006     & 1        & 1        & 1.72     &  8.2     & 12       & 0        & 0        & 1        & 0        & ⋯        & 0        & 0        & 0        &  7       &  0       & 100      & 23       & 70.00000 & 4        & 59      \\\\\n",
       "\t 1008     & 1        & 1        & 2.92     &  6.8     &  2       & 0        & 1        & 1        & 0        & ⋯        & 0        & 0        & 0        & 27       & 24       &  NA      & NA       &       NA & 1        &  5      \\\\\n",
       "\t 1009     & 1        & 0        & 0.70     &  2.8     & 10       & 0        & 1        & 1        & 0        & ⋯        & 1        & 1        & 1        &  0       &  0       & 100      & 32       & 43.66667 & 1        &  9      \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "ID | hospital_number | female | age | pim2 | pelod | p_bloodculture | comorbidity | rf_pneumonia | rf_sepsis | ⋯ | mortality_30day | mortality_28day | mortality_100day | ventilation_free_day | picu_free_days | fio2 | map | amplitude | OI_cat | ICUduration | \n",
       "|---|---|---|---|---|---|\n",
       "| 1001     | 1        | 1        | 0.02     | 87.3     | 42       | 1        | 0        | 0        | 1        | ⋯        | 1        | 1        | 1        |  0       |  0       |  NA      | NA       |       NA | 1        |  3       | \n",
       "| 1003     | 1        | 0        | 0.47     |  2.8     | 11       | 0        | 0        | 1        | 0        | ⋯        | 0        | 0        | 0        | 21       | 26       |  NA      | NA       |       NA | 1        |  3       | \n",
       "| 1005     | 1        | 1        | 0.10     |  0.6     |  1       | 0        | 0        | 1        | 0        | ⋯        | 0        | 0        | 0        | 21       | 20       |  NA      | NA       |       NA | 2        |  9       | \n",
       "| 1006     | 1        | 1        | 1.72     |  8.2     | 12       | 0        | 0        | 1        | 0        | ⋯        | 0        | 0        | 0        |  7       |  0       | 100      | 23       | 70.00000 | 4        | 59       | \n",
       "| 1008     | 1        | 1        | 2.92     |  6.8     |  2       | 0        | 1        | 1        | 0        | ⋯        | 0        | 0        | 0        | 27       | 24       |  NA      | NA       |       NA | 1        |  5       | \n",
       "| 1009     | 1        | 0        | 0.70     |  2.8     | 10       | 0        | 1        | 1        | 0        | ⋯        | 1        | 1        | 1        |  0       |  0       | 100      | 32       | 43.66667 | 1        |  9       | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  ID   hospital_number female age  pim2 pelod p_bloodculture comorbidity\n",
       "1 1001 1               1      0.02 87.3 42    1              0          \n",
       "2 1003 1               0      0.47  2.8 11    0              0          \n",
       "3 1005 1               1      0.10  0.6  1    0              0          \n",
       "4 1006 1               1      1.72  8.2 12    0              0          \n",
       "5 1008 1               1      2.92  6.8  2    0              1          \n",
       "6 1009 1               0      0.70  2.8 10    0              1          \n",
       "  rf_pneumonia rf_sepsis ⋯ mortality_30day mortality_28day mortality_100day\n",
       "1 0            1         ⋯ 1               1               1               \n",
       "2 1            0         ⋯ 0               0               0               \n",
       "3 1            0         ⋯ 0               0               0               \n",
       "4 1            0         ⋯ 0               0               0               \n",
       "5 1            0         ⋯ 0               0               0               \n",
       "6 1            0         ⋯ 1               1               1               \n",
       "  ventilation_free_day picu_free_days fio2 map amplitude OI_cat ICUduration\n",
       "1  0                    0              NA  NA        NA  1       3         \n",
       "2 21                   26              NA  NA        NA  1       3         \n",
       "3 21                   20              NA  NA        NA  2       9         \n",
       "4  7                    0             100  23  70.00000  4      59         \n",
       "5 27                   24              NA  NA        NA  1       5         \n",
       "6  0                    0             100  32  43.66667  1       9         "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#load data\n",
    "kkh_data <- read.csv(\"pards_data_1107.csv\")\n",
    "#kkh_data <- read.csv(\"pards_data_0816.csv\")\n",
    "df<-data.frame(kkh_data)\n",
    "kkh<-df[!is.na(df$d2_oi),]\n",
    "kkh <- kkh[which(!((kkh$hfv==1)&(kkh$HFV_verify==0))), ]\n",
    "#kkh <- kkh[which((kkh$Non_CPAP==1)), ]\n",
    "####check column names\n",
    "var<-colnames(kkh, do.NULL = FALSE)[1:9]\n",
    "#head(kkh)\n",
    "\n",
    "####select features\n",
    "ID<-kkh$StudyID\n",
    "hospital_number<-kkh$Hospital_number\n",
    "female<-as.numeric(kkh$Gender=='Female')\n",
    "age<-kkh$Age\n",
    "pim2<-kkh$pim2\n",
    "pelod<-kkh$pelod\n",
    "p_bloodculture<-as.numeric(kkh$Positive_blood_culture=='Yes')\n",
    "early_HFV<-kkh$early_HFV\n",
    "D1_HFV<-kkh$D1_HFV\n",
    "D2_HFV<-kkh$D2_HFV\n",
    "D3_HFV<-kkh$D3_HFV\n",
    "D4_HFV<-kkh$D4_HFV\n",
    "D5_HFV<-kkh$D5_HFV\n",
    "D6_HFV<-kkh$D6_HFV\n",
    "D7_HFV<-kkh$D7_HFV\n",
    "ICUadmissiondate <- as.Date(kkh$ICUadmissiondate,format=\"%d/%m/%Y\")\n",
    "ICUdischargedate <- as.Date(kkh$ICUdischargedate,format=\"%d/%m/%Y\")\n",
    "\n",
    "\n",
    "#comorbidity\n",
    "comorbidity<-kkh$comorbidity\n",
    "#risk factor of ARDS\n",
    "rf_pneumonia<-as.numeric(kkh$PNEUMONIA=='Checked')\n",
    "rf_sepsis<-as.numeric(kkh$SEPSIS=='Checked')\n",
    "rf_aspiration<-as.numeric(kkh$ASPIRATION=='Checked')\n",
    "rf_transfusion<-as.numeric(kkh$TRANSFUSION=='Checked')\n",
    "rf_trauma<-as.numeric(kkh$TRAUMA=='Checked')\n",
    "rf_drowning<-as.numeric(kkh$DROWNING=='Checked')\n",
    "\n",
    "#other useful features\n",
    "MODS<-as.numeric(kkh$MODS=='Yes')\n",
    "OI<-kkh$d2_oi\n",
    "OI_cat<-kkh$OI_category\n",
    "treatment<-as.numeric(kkh$Received_HFV=='Yes')\n",
    "died_in_ICU<-kkh$died_in_ICU\n",
    "mortality_30day<-kkh$mort_30\n",
    "mortality_28day<-kkh$mort_28\n",
    "mortality_100day<-ifelse(kkh$Died.within.100.days==\"Yes\",1,0)\n",
    "ventilation_free_day<-kkh$ventilation_free_day\n",
    "picu_free_days<-kkh$picu_free_days\n",
    "fio2<-kkh$avg_fio2max\n",
    "map<-kkh$avg_mapmax\n",
    "amplitude<-kkh$avg_amplitudemax\n",
    "\n",
    "\n",
    "####prepare for the data\n",
    "mydata<-cbind(ID,hospital_number,female,age,pim2,pelod,p_bloodculture,comorbidity,rf_pneumonia,rf_sepsis,rf_aspiration,rf_transfusion\n",
    "              ,rf_trauma\n",
    "              ,rf_drowning,MODS,OI\n",
    "              ,treatment,died_in_ICU,D1_HFV,D2_HFV,D3_HFV,D4_HFV,D5_HFV,D6_HFV,D7_HFV,ICUadmissiondate,ICUdischargedate\n",
    "              ,mortality_30day,mortality_28day,mortality_100day,ventilation_free_day,picu_free_days, fio2, map, amplitude,OI_cat)\n",
    "\n",
    "\n",
    "mydata<-data.frame(mydata)\n",
    "mydata$ICUduration<-mydata$ICUdischargedate-mydata$ICUadmissiondate+1\n",
    "#mydata$OI_cat<-as.factor(mydata$OI_category)\n",
    "#mydata$OI_cat<-C(mydata$OI_cat, contr.treatment, base=3)\n",
    "\n",
    "head(mydata)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "2162"
      ],
      "text/latex": [
       "2162"
      ],
      "text/markdown": [
       "2162"
      ],
      "text/plain": [
       "[1] 2162"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>treatment</th><th scope=col>ID</th><th scope=col>hospital_number</th><th scope=col>start_day</th><th scope=col>end_day</th><th scope=col>HFV_indicator</th><th scope=col>event</th><th scope=col>died_in_ICU</th><th scope=col>mortality_28day</th><th scope=col>ICUduration</th><th scope=col>⋯</th><th scope=col>comorbidity</th><th scope=col>rf_pneumonia</th><th scope=col>rf_sepsis</th><th scope=col>rf_aspiration</th><th scope=col>rf_transfusion</th><th scope=col>rf_trauma</th><th scope=col>rf_drowning</th><th scope=col>MODS</th><th scope=col>OI</th><th scope=col>OI_cat</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>1</th><td>0    </td><td>1001 </td><td>1    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>1    </td><td>1    </td><td> 3   </td><td>⋯    </td><td>0    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td> 5.06</td><td>1    </td></tr>\n",
       "\t<tr><th scope=row>329</th><td>0    </td><td>1001 </td><td>1    </td><td>1    </td><td>2    </td><td>0    </td><td>0    </td><td>1    </td><td>1    </td><td> 3   </td><td>⋯    </td><td>0    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td> 5.06</td><td>1    </td></tr>\n",
       "\t<tr><th scope=row>657</th><td>0    </td><td>1001 </td><td>1    </td><td>2    </td><td>3    </td><td>0    </td><td>1    </td><td>1    </td><td>1    </td><td> 3   </td><td>⋯    </td><td>0    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td> 5.06</td><td>1    </td></tr>\n",
       "\t<tr><th scope=row>2</th><td>0    </td><td>1003 </td><td>1    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 3   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 5.11</td><td>1    </td></tr>\n",
       "\t<tr><th scope=row>330</th><td>0    </td><td>1003 </td><td>1    </td><td>1    </td><td>2    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 3   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 5.11</td><td>1    </td></tr>\n",
       "\t<tr><th scope=row>658</th><td>0    </td><td>1003 </td><td>1    </td><td>2    </td><td>3    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 3   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 5.11</td><td>1    </td></tr>\n",
       "\t<tr><th scope=row>3</th><td>0    </td><td>1005 </td><td>1    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 9   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 8.01</td><td>2    </td></tr>\n",
       "\t<tr><th scope=row>331</th><td>0    </td><td>1005 </td><td>1    </td><td>1    </td><td>2    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 9   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 8.01</td><td>2    </td></tr>\n",
       "\t<tr><th scope=row>659</th><td>0    </td><td>1005 </td><td>1    </td><td>2    </td><td>3    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 9   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 8.01</td><td>2    </td></tr>\n",
       "\t<tr><th scope=row>987</th><td>0    </td><td>1005 </td><td>1    </td><td>3    </td><td>4    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 9   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 8.01</td><td>2    </td></tr>\n",
       "\t<tr><th scope=row>1315</th><td>0    </td><td>1005 </td><td>1    </td><td>4    </td><td>5    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 9   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 8.01</td><td>2    </td></tr>\n",
       "\t<tr><th scope=row>1643</th><td>0    </td><td>1005 </td><td>1    </td><td>5    </td><td>6    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 9   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 8.01</td><td>2    </td></tr>\n",
       "\t<tr><th scope=row>1971</th><td>0    </td><td>1005 </td><td>1    </td><td>6    </td><td>7    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 9   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td> 8.01</td><td>2    </td></tr>\n",
       "\t<tr><th scope=row>4</th><td>1    </td><td>1006 </td><td>1    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>59   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td>20.63</td><td>4    </td></tr>\n",
       "\t<tr><th scope=row>332</th><td>1    </td><td>1006 </td><td>1    </td><td>1    </td><td>2    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>59   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td>20.63</td><td>4    </td></tr>\n",
       "\t<tr><th scope=row>660</th><td>1    </td><td>1006 </td><td>1    </td><td>2    </td><td>3    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>59   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td>20.63</td><td>4    </td></tr>\n",
       "\t<tr><th scope=row>988</th><td>1    </td><td>1006 </td><td>1    </td><td>3    </td><td>4    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>59   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td>20.63</td><td>4    </td></tr>\n",
       "\t<tr><th scope=row>1316</th><td>1    </td><td>1006 </td><td>1    </td><td>4    </td><td>5    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>59   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td>20.63</td><td>4    </td></tr>\n",
       "\t<tr><th scope=row>1644</th><td>1    </td><td>1006 </td><td>1    </td><td>5    </td><td>6    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>59   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td>20.63</td><td>4    </td></tr>\n",
       "\t<tr><th scope=row>1972</th><td>1    </td><td>1006 </td><td>1    </td><td>6    </td><td>7    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>59   </td><td>⋯    </td><td>0    </td><td>1    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>0    </td><td>1    </td><td>20.63</td><td>4    </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllllllllllllllllllllllll}\n",
       "  & treatment & ID & hospital\\_number & start\\_day & end\\_day & HFV\\_indicator & event & died\\_in\\_ICU & mortality\\_28day & ICUduration & ⋯ & comorbidity & rf\\_pneumonia & rf\\_sepsis & rf\\_aspiration & rf\\_transfusion & rf\\_trauma & rf\\_drowning & MODS & OI & OI\\_cat\\\\\n",
       "\\hline\n",
       "\t1 & 0     & 1001  & 1     & 0     & 1     & 0     & 0     & 1     & 1     &  3    & ⋯     & 0     & 0     & 1     & 0     & 0     & 0     & 0     & 1     &  5.06 & 1    \\\\\n",
       "\t329 & 0     & 1001  & 1     & 1     & 2     & 0     & 0     & 1     & 1     &  3    & ⋯     & 0     & 0     & 1     & 0     & 0     & 0     & 0     & 1     &  5.06 & 1    \\\\\n",
       "\t657 & 0     & 1001  & 1     & 2     & 3     & 0     & 1     & 1     & 1     &  3    & ⋯     & 0     & 0     & 1     & 0     & 0     & 0     & 0     & 1     &  5.06 & 1    \\\\\n",
       "\t2 & 0     & 1003  & 1     & 0     & 1     & 0     & 0     & 0     & 0     &  3    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  5.11 & 1    \\\\\n",
       "\t330 & 0     & 1003  & 1     & 1     & 2     & 0     & 0     & 0     & 0     &  3    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  5.11 & 1    \\\\\n",
       "\t658 & 0     & 1003  & 1     & 2     & 3     & 0     & 0     & 0     & 0     &  3    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  5.11 & 1    \\\\\n",
       "\t3 & 0     & 1005  & 1     & 0     & 1     & 0     & 0     & 0     & 0     &  9    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  8.01 & 2    \\\\\n",
       "\t331 & 0     & 1005  & 1     & 1     & 2     & 0     & 0     & 0     & 0     &  9    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  8.01 & 2    \\\\\n",
       "\t659 & 0     & 1005  & 1     & 2     & 3     & 0     & 0     & 0     & 0     &  9    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  8.01 & 2    \\\\\n",
       "\t987 & 0     & 1005  & 1     & 3     & 4     & 0     & 0     & 0     & 0     &  9    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  8.01 & 2    \\\\\n",
       "\t1315 & 0     & 1005  & 1     & 4     & 5     & 0     & 0     & 0     & 0     &  9    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  8.01 & 2    \\\\\n",
       "\t1643 & 0     & 1005  & 1     & 5     & 6     & 0     & 0     & 0     & 0     &  9    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  8.01 & 2    \\\\\n",
       "\t1971 & 0     & 1005  & 1     & 6     & 7     & 0     & 0     & 0     & 0     &  9    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 0     &  8.01 & 2    \\\\\n",
       "\t4 & 1     & 1006  & 1     & 0     & 1     & 0     & 0     & 0     & 0     & 59    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 1     & 20.63 & 4    \\\\\n",
       "\t332 & 1     & 1006  & 1     & 1     & 2     & 0     & 0     & 0     & 0     & 59    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 1     & 20.63 & 4    \\\\\n",
       "\t660 & 1     & 1006  & 1     & 2     & 3     & 0     & 0     & 0     & 0     & 59    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 1     & 20.63 & 4    \\\\\n",
       "\t988 & 1     & 1006  & 1     & 3     & 4     & 0     & 0     & 0     & 0     & 59    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 1     & 20.63 & 4    \\\\\n",
       "\t1316 & 1     & 1006  & 1     & 4     & 5     & 0     & 0     & 0     & 0     & 59    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 1     & 20.63 & 4    \\\\\n",
       "\t1644 & 1     & 1006  & 1     & 5     & 6     & 0     & 0     & 0     & 0     & 59    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 1     & 20.63 & 4    \\\\\n",
       "\t1972 & 1     & 1006  & 1     & 6     & 7     & 1     & 0     & 0     & 0     & 59    & ⋯     & 0     & 1     & 0     & 0     & 0     & 0     & 0     & 1     & 20.63 & 4    \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | treatment | ID | hospital_number | start_day | end_day | HFV_indicator | event | died_in_ICU | mortality_28day | ICUduration | ⋯ | comorbidity | rf_pneumonia | rf_sepsis | rf_aspiration | rf_transfusion | rf_trauma | rf_drowning | MODS | OI | OI_cat | \n",
       "|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n",
       "| 1 | 0     | 1001  | 1     | 0     | 1     | 0     | 0     | 1     | 1     |  3    | ⋯     | 0     | 0     | 1     | 0     | 0     | 0     | 0     | 1     |  5.06 | 1     | \n",
       "| 329 | 0     | 1001  | 1     | 1     | 2     | 0     | 0     | 1     | 1     |  3    | ⋯     | 0     | 0     | 1     | 0     | 0     | 0     | 0     | 1     |  5.06 | 1     | \n",
       "| 657 | 0     | 1001  | 1     | 2     | 3     | 0     | 1     | 1     | 1     |  3    | ⋯     | 0     | 0     | 1     | 0     | 0     | 0     | 0     | 1     |  5.06 | 1     | \n",
       "| 2 | 0     | 1003  | 1     | 0     | 1     | 0     | 0     | 0     | 0     |  3    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  5.11 | 1     | \n",
       "| 330 | 0     | 1003  | 1     | 1     | 2     | 0     | 0     | 0     | 0     |  3    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  5.11 | 1     | \n",
       "| 658 | 0     | 1003  | 1     | 2     | 3     | 0     | 0     | 0     | 0     |  3    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  5.11 | 1     | \n",
       "| 3 | 0     | 1005  | 1     | 0     | 1     | 0     | 0     | 0     | 0     |  9    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  8.01 | 2     | \n",
       "| 331 | 0     | 1005  | 1     | 1     | 2     | 0     | 0     | 0     | 0     |  9    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  8.01 | 2     | \n",
       "| 659 | 0     | 1005  | 1     | 2     | 3     | 0     | 0     | 0     | 0     |  9    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  8.01 | 2     | \n",
       "| 987 | 0     | 1005  | 1     | 3     | 4     | 0     | 0     | 0     | 0     |  9    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  8.01 | 2     | \n",
       "| 1315 | 0     | 1005  | 1     | 4     | 5     | 0     | 0     | 0     | 0     |  9    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  8.01 | 2     | \n",
       "| 1643 | 0     | 1005  | 1     | 5     | 6     | 0     | 0     | 0     | 0     |  9    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  8.01 | 2     | \n",
       "| 1971 | 0     | 1005  | 1     | 6     | 7     | 0     | 0     | 0     | 0     |  9    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 0     |  8.01 | 2     | \n",
       "| 4 | 1     | 1006  | 1     | 0     | 1     | 0     | 0     | 0     | 0     | 59    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 1     | 20.63 | 4     | \n",
       "| 332 | 1     | 1006  | 1     | 1     | 2     | 0     | 0     | 0     | 0     | 59    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 1     | 20.63 | 4     | \n",
       "| 660 | 1     | 1006  | 1     | 2     | 3     | 0     | 0     | 0     | 0     | 59    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 1     | 20.63 | 4     | \n",
       "| 988 | 1     | 1006  | 1     | 3     | 4     | 0     | 0     | 0     | 0     | 59    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 1     | 20.63 | 4     | \n",
       "| 1316 | 1     | 1006  | 1     | 4     | 5     | 0     | 0     | 0     | 0     | 59    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 1     | 20.63 | 4     | \n",
       "| 1644 | 1     | 1006  | 1     | 5     | 6     | 0     | 0     | 0     | 0     | 59    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 1     | 20.63 | 4     | \n",
       "| 1972 | 1     | 1006  | 1     | 6     | 7     | 1     | 0     | 0     | 0     | 59    | ⋯     | 0     | 1     | 0     | 0     | 0     | 0     | 0     | 1     | 20.63 | 4     | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "     treatment ID   hospital_number start_day end_day HFV_indicator event\n",
       "1    0         1001 1               0         1       0             0    \n",
       "329  0         1001 1               1         2       0             0    \n",
       "657  0         1001 1               2         3       0             1    \n",
       "2    0         1003 1               0         1       0             0    \n",
       "330  0         1003 1               1         2       0             0    \n",
       "658  0         1003 1               2         3       0             0    \n",
       "3    0         1005 1               0         1       0             0    \n",
       "331  0         1005 1               1         2       0             0    \n",
       "659  0         1005 1               2         3       0             0    \n",
       "987  0         1005 1               3         4       0             0    \n",
       "1315 0         1005 1               4         5       0             0    \n",
       "1643 0         1005 1               5         6       0             0    \n",
       "1971 0         1005 1               6         7       0             0    \n",
       "4    1         1006 1               0         1       0             0    \n",
       "332  1         1006 1               1         2       0             0    \n",
       "660  1         1006 1               2         3       0             0    \n",
       "988  1         1006 1               3         4       0             0    \n",
       "1316 1         1006 1               4         5       0             0    \n",
       "1644 1         1006 1               5         6       0             0    \n",
       "1972 1         1006 1               6         7       1             0    \n",
       "     died_in_ICU mortality_28day ICUduration ⋯ comorbidity rf_pneumonia\n",
       "1    1           1                3          ⋯ 0           0           \n",
       "329  1           1                3          ⋯ 0           0           \n",
       "657  1           1                3          ⋯ 0           0           \n",
       "2    0           0                3          ⋯ 0           1           \n",
       "330  0           0                3          ⋯ 0           1           \n",
       "658  0           0                3          ⋯ 0           1           \n",
       "3    0           0                9          ⋯ 0           1           \n",
       "331  0           0                9          ⋯ 0           1           \n",
       "659  0           0                9          ⋯ 0           1           \n",
       "987  0           0                9          ⋯ 0           1           \n",
       "1315 0           0                9          ⋯ 0           1           \n",
       "1643 0           0                9          ⋯ 0           1           \n",
       "1971 0           0                9          ⋯ 0           1           \n",
       "4    0           0               59          ⋯ 0           1           \n",
       "332  0           0               59          ⋯ 0           1           \n",
       "660  0           0               59          ⋯ 0           1           \n",
       "988  0           0               59          ⋯ 0           1           \n",
       "1316 0           0               59          ⋯ 0           1           \n",
       "1644 0           0               59          ⋯ 0           1           \n",
       "1972 0           0               59          ⋯ 0           1           \n",
       "     rf_sepsis rf_aspiration rf_transfusion rf_trauma rf_drowning MODS OI   \n",
       "1    1         0             0              0         0           1     5.06\n",
       "329  1         0             0              0         0           1     5.06\n",
       "657  1         0             0              0         0           1     5.06\n",
       "2    0         0             0              0         0           0     5.11\n",
       "330  0         0             0              0         0           0     5.11\n",
       "658  0         0             0              0         0           0     5.11\n",
       "3    0         0             0              0         0           0     8.01\n",
       "331  0         0             0              0         0           0     8.01\n",
       "659  0         0             0              0         0           0     8.01\n",
       "987  0         0             0              0         0           0     8.01\n",
       "1315 0         0             0              0         0           0     8.01\n",
       "1643 0         0             0              0         0           0     8.01\n",
       "1971 0         0             0              0         0           0     8.01\n",
       "4    0         0             0              0         0           1    20.63\n",
       "332  0         0             0              0         0           1    20.63\n",
       "660  0         0             0              0         0           1    20.63\n",
       "988  0         0             0              0         0           1    20.63\n",
       "1316 0         0             0              0         0           1    20.63\n",
       "1644 0         0             0              0         0           1    20.63\n",
       "1972 0         0             0              0         0           1    20.63\n",
       "     OI_cat\n",
       "1    1     \n",
       "329  1     \n",
       "657  1     \n",
       "2    1     \n",
       "330  1     \n",
       "658  1     \n",
       "3    2     \n",
       "331  2     \n",
       "659  2     \n",
       "987  2     \n",
       "1315 2     \n",
       "1643 2     \n",
       "1971 2     \n",
       "4    4     \n",
       "332  4     \n",
       "660  4     \n",
       "988  4     \n",
       "1316 4     \n",
       "1644 4     \n",
       "1972 4     "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#create new dataset for cox ph model with time-dependent adjustment for treatment and covariates\n",
    "#df2 is the final dataset\n",
    "df<-melt(mydata,id.vars=c(\"ID\",\"hospital_number\",\"ICUduration\",\"died_in_ICU\",\"mortality_30day\",\"mortality_28day\",\"mortality_100day\",\"treatment\",\n",
    "                          \"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\n",
    "                          \"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\",\"rf_transfusion\", \n",
    "                          \"rf_trauma\", \"rf_drowning\",\"MODS\",\"OI\",\"OI_cat\"), \n",
    "         measure.vars=c(\"D1_HFV\",\"D2_HFV\",\"D3_HFV\",\"D4_HFV\",\"D5_HFV\",\"D6_HFV\",\"D7_HFV\"),\n",
    "         variable.name = \"end_day\",value.name = \"HFV_indicator\",factorsAsStrings = FALSE)\n",
    "\n",
    "df<-df[order(df$ID), , drop = FALSE]\n",
    "df$end_day<-revalue(df$end_day, c(\"D1_HFV\"=1,\"D2_HFV\"=2, \"D3_HFV\"=3,\"D4_HFV\"=4,\"D5_HFV\"=5, \"D6_HFV\"=6,\"D7_HFV\"=7))\n",
    "df$end_day<-as.numeric(df$end_day)\n",
    "df$start_day<-df$end_day-1\n",
    "\n",
    "df1<-df[df$start_day<df$ICUduration,]\n",
    "df1$event<-ifelse(df1$end_day<df1$ICUduration, 0,df1$died_in_ICU)\n",
    "\n",
    "df1 <- df1[c(\"treatment\",\"ID\",\"hospital_number\",\"start_day\",\"end_day\",\"HFV_indicator\",\"event\",\"died_in_ICU\",\"mortality_28day\",\"ICUduration\",\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\"\n",
    "         ,\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\", \"rf_drowning\",\"MODS\",\"OI\",\"OI_cat\")]\n",
    "\n",
    "\n",
    "length(df1$ID)\n",
    "df2 <- do.call(\"rbind\", lapply(unique(df1$ID), function(id) {\n",
    "    df1_id <- df1[df1$ID == id, ]\n",
    "    n_id <- nrow(df1_id)\n",
    "    n_all <- df1_id$ICUduration[1]\n",
    "    if(n_all>7){\n",
    "        n_all<-7\n",
    "    }\n",
    "    if (n_id == n_all) {\n",
    "        df1_id\n",
    "    } else {\n",
    "        last_row <- df1_id[n_id, ]\n",
    "        df1_add <- do.call(\"rbind\", lapply((n_id + 1):n_all, function(r) {\n",
    "            new_row <- last_row\n",
    "            new_row$start_day <- r - 1\n",
    "            new_row$end_day <- r\n",
    "            new_row$HFV_indicator<-0\n",
    "            if(r==n_all){\n",
    "                new_row$event<- new_row$mortality_28day\n",
    "            } else{\n",
    "                new_row$event<- 0\n",
    "            }\n",
    "            new_row\n",
    "        }))\n",
    "        rbind(df1_id, df1_add)\n",
    "    }\n",
    "}))\n",
    "#head(df2)\n",
    "df2[1:20,]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# GenMatch with Matching package"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Loading required namespace: rgenoud\n"
     ]
    }
   ],
   "source": [
    "\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "var_list2<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI_cat\")\n",
    "treatment<-data.frame(mydata[,\"treatment\"])\n",
    "#mydata_wth_pscore<-data.frame(cbind(mydata[, var_list,pscore]))\n",
    "gen_data<-data.frame(mydata[, var_list])\n",
    "#mydata_wth_pscore$female<-as.factor(mydata_wth_pscore$female)\n",
    "\n",
    "# treatment~age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(rf_pneumonia)+factor(rf_sepsis)\n",
    "#                  +factor(rf_aspiration)+factor(rf_transfusion)+factor(rf_trauma)+factor(rf_drowning)+factor(MODS)+factor(OI_cat)\n",
    "gen.out<-GenMatch(mydata$treatment,mydata[,var_list], pop.size=328,replace=FALSE,exact=force_exact,distance.tolerance=4.0,print.level=0)\n",
    "gen_treat_data<-gen_data\n",
    "gen_treat_data$treatment<-mydata[,\"treatment\"]\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Estimate...  0 \n",
      "SE.........  0 \n",
      "T-stat.....  NaN \n",
      "p.val......  NA \n",
      "\n",
      "Original number of observations..............  328 \n",
      "Original number of treated obs...............  122 \n",
      "Matched number of observations...............  118 \n",
      "Matched number of observations  (unweighted).  118 \n",
      "\n",
      "Number of obs dropped by 'exact' or 'caliper'  4 \n",
      "\n"
     ]
    }
   ],
   "source": [
    "m.out<-Match(Tr=gen_treat_data$treatment, X=gen_treat_data[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=3.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "\tOne Sample t-test\n",
       "\n",
       "data:  diffy\n",
       "t = -1.239, df = 117, p-value = 0.2178\n",
       "alternative hypothesis: true mean is not equal to 0\n",
       "95 percent confidence interval:\n",
       " -3.5673768  0.8216141\n",
       "sample estimates:\n",
       "mean of x \n",
       "-1.372881 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tOne Sample t-test\n",
       "\n",
       "data:  diffy1\n",
       "t = -2.1615, df = 117, p-value = 0.03269\n",
       "alternative hypothesis: true mean is not equal to 0\n",
       "95 percent confidence interval:\n",
       " -4.141030 -0.181004\n",
       "sample estimates:\n",
       "mean of x \n",
       "-2.161017 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tMcNemar's Chi-squared test with continuity correction\n",
       "\n",
       "data:  hfv$mort_100day and non_hfv$mort_100day\n",
       "McNemar's chi-squared = 0.078431, df = 1, p-value = 0.7794\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"Non-HFOV group 100day mortality n:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "51"
      ],
      "text/latex": [
       "51"
      ],
      "text/markdown": [
       "51"
      ],
      "text/plain": [
       "[1] 51"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"%:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.432203389830508"
      ],
      "text/latex": [
       "0.432203389830508"
      ],
      "text/markdown": [
       "0.432203389830508"
      ],
      "text/plain": [
       "[1] 0.4322034"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"HFOV group 100day mortality n:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "54"
      ],
      "text/latex": [
       "54"
      ],
      "text/markdown": [
       "54"
      ],
      "text/plain": [
       "[1] 54"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"%:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.457627118644068"
      ],
      "text/latex": [
       "0.457627118644068"
      ],
      "text/markdown": [
       "0.457627118644068"
      ],
      "text/plain": [
       "[1] 0.4576271"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"Non-HFOV group 28day mortality n:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "21"
      ],
      "text/latex": [
       "21"
      ],
      "text/markdown": [
       "21"
      ],
      "text/plain": [
       "[1] 21"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"%:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.177966101694915"
      ],
      "text/latex": [
       "0.177966101694915"
      ],
      "text/markdown": [
       "0.177966101694915"
      ],
      "text/plain": [
       "[1] 0.1779661"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"HFOV group 28day mortality n:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "38"
      ],
      "text/latex": [
       "38"
      ],
      "text/markdown": [
       "38"
      ],
      "text/plain": [
       "[1] 38"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"%:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.322033898305085"
      ],
      "text/latex": [
       "0.322033898305085"
      ],
      "text/markdown": [
       "0.322033898305085"
      ],
      "text/plain": [
       "[1] 0.3220339"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tMcNemar's Chi-squared test with continuity correction\n",
       "\n",
       "data:  hfv$mortality_28day and non_hfv$mortality_28day\n",
       "McNemar's chi-squared = 5.6889, df = 1, p-value = 0.01707\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] 38 21 80 97\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.322033898305085"
      ],
      "text/latex": [
       "0.322033898305085"
      ],
      "text/markdown": [
       "0.322033898305085"
      ],
      "text/plain": [
       "[1] 0.3220339"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "0.177966101694915"
      ],
      "text/latex": [
       "0.177966101694915"
      ],
      "text/markdown": [
       "0.177966101694915"
      ],
      "text/plain": [
       "[1] 0.1779661"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<dl>\n",
       "\t<dt>$data</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>Disease1</th><th scope=col>Disease2</th><th scope=col>Total</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>38 </td><td> 80</td><td>118</td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td>21 </td><td> 97</td><td>118</td></tr>\n",
       "\t<tr><th scope=row>Total</th><td>59 </td><td>177</td><td>236</td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$measure</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>estimate</th><th scope=col>lower</th><th scope=col>upper</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>1.000000</td><td>      NA</td><td>      NA</td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td>2.180742</td><td>1.192347</td><td>4.077767</td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$p.value</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>midp.exact</th><th scope=col>fisher.exact</th><th scope=col>chi.square</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>        NA</td><td>        NA</td><td>       NA </td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td>0.01112853</td><td>0.01573759</td><td>0.0106006 </td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$correction</dt>\n",
       "\t\t<dd>FALSE</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description}\n",
       "\\item[\\$data] \\begin{tabular}{r|lll}\n",
       "  & Disease1 & Disease2 & Total\\\\\n",
       "\\hline\n",
       "\tExposed1 & 38  &  80 & 118\\\\\n",
       "\tExposed2 & 21  &  97 & 118\\\\\n",
       "\tTotal & 59  & 177 & 236\\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$measure] \\begin{tabular}{r|lll}\n",
       "  & estimate & lower & upper\\\\\n",
       "\\hline\n",
       "\tExposed1 & 1.000000 &       NA &       NA\\\\\n",
       "\tExposed2 & 2.180742 & 1.192347 & 4.077767\\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$p.value] \\begin{tabular}{r|lll}\n",
       "  & midp.exact & fisher.exact & chi.square\\\\\n",
       "\\hline\n",
       "\tExposed1 &         NA &         NA &        NA \\\\\n",
       "\tExposed2 & 0.01112853 & 0.01573759 & 0.0106006 \\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$correction] FALSE\n",
       "\\end{description}\n"
      ],
      "text/markdown": [
       "$data\n",
       ":   \n",
       "| <!--/--> | Disease1 | Disease2 | Total | \n",
       "|---|---|---|\n",
       "| Exposed1 | 38  |  80 | 118 | \n",
       "| Exposed2 | 21  |  97 | 118 | \n",
       "| Total | 59  | 177 | 236 | \n",
       "\n",
       "\n",
       "\n",
       "$measure\n",
       ":   \n",
       "| <!--/--> | estimate | lower | upper | \n",
       "|---|---|\n",
       "| Exposed1 | 1.000000 |       NA |       NA | \n",
       "| Exposed2 | 2.180742 | 1.192347 | 4.077767 | \n",
       "\n",
       "\n",
       "\n",
       "$p.value\n",
       ":   \n",
       "| <!--/--> | midp.exact | fisher.exact | chi.square | \n",
       "|---|---|\n",
       "| Exposed1 |         NA |         NA |        NA  | \n",
       "| Exposed2 | 0.01112853 | 0.01573759 | 0.0106006  | \n",
       "\n",
       "\n",
       "\n",
       "$correction\n",
       ":   FALSE\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "$data\n",
       "          Outcome\n",
       "Predictor  Disease1 Disease2 Total\n",
       "  Exposed1       38       80   118\n",
       "  Exposed2       21       97   118\n",
       "  Total          59      177   236\n",
       "\n",
       "$measure\n",
       "          odds ratio with 95% C.I.\n",
       "Predictor  estimate    lower    upper\n",
       "  Exposed1 1.000000       NA       NA\n",
       "  Exposed2 2.180742 1.192347 4.077767\n",
       "\n",
       "$p.value\n",
       "          two-sided\n",
       "Predictor  midp.exact fisher.exact chi.square\n",
       "  Exposed1         NA           NA         NA\n",
       "  Exposed2 0.01112853   0.01573759  0.0106006\n",
       "\n",
       "$correction\n",
       "[1] FALSE\n",
       "\n",
       "attr(,\"method\")\n",
       "[1] \"median-unbiased estimate & mid-p exact CI\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>ID</th><th scope=col>hospital_number</th><th scope=col>female</th><th scope=col>age</th><th scope=col>pim2</th><th scope=col>pelod</th><th scope=col>p_bloodculture</th><th scope=col>comorbidity</th><th scope=col>rf_pneumonia</th><th scope=col>rf_sepsis</th><th scope=col>⋯</th><th scope=col>mortality_28day</th><th scope=col>mortality_100day</th><th scope=col>ventilation_free_day</th><th scope=col>picu_free_days</th><th scope=col>fio2</th><th scope=col>map</th><th scope=col>amplitude</th><th scope=col>OI_cat</th><th scope=col>ICUduration</th><th scope=col>mort_100day</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>4</th><td>1006    </td><td>1       </td><td>1       </td><td>1.72    </td><td> 8.2    </td><td>12      </td><td>0       </td><td>0       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td> 7      </td><td> 0      </td><td>100.00  </td><td>23.0    </td><td>70.00000</td><td>4       </td><td>59      </td><td>0       </td></tr>\n",
       "\t<tr><th scope=row>6</th><td>1009    </td><td>1       </td><td>0       </td><td>0.70    </td><td> 2.8    </td><td>10      </td><td>0       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>1       </td><td>1       </td><td> 0      </td><td> 0      </td><td>100.00  </td><td>32.0    </td><td>43.66667</td><td>1       </td><td> 9      </td><td>1       </td></tr>\n",
       "\t<tr><th scope=row>8</th><td>1011    </td><td>1       </td><td>1       </td><td>1.31    </td><td>27.6    </td><td> 1      </td><td>1       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td> 0      </td><td> 0      </td><td>100.00  </td><td>30.0    </td><td>75.00000</td><td>4       </td><td>61      </td><td>0       </td></tr>\n",
       "\t<tr><th scope=row>18</th><td>1022    </td><td>1       </td><td>1       </td><td>0.28    </td><td> 9.0    </td><td>20      </td><td>0       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td>16      </td><td>15      </td><td> 58.75  </td><td>28.0    </td><td>50.00000</td><td>2       </td><td>14      </td><td>0       </td></tr>\n",
       "\t<tr><th scope=row>20</th><td>1024    </td><td>1       </td><td>0       </td><td>5.34    </td><td>75.6    </td><td>41      </td><td>1       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>1       </td><td>1       </td><td> 0      </td><td> 0      </td><td>100.00  </td><td>25.5    </td><td>62.00000</td><td>1       </td><td> 3      </td><td>1       </td></tr>\n",
       "\t<tr><th scope=row>27</th><td>1037    </td><td>1       </td><td>1       </td><td>0.28    </td><td> 8.9    </td><td>20      </td><td>0       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>0       </td><td>16      </td><td>15      </td><td> 67.00  </td><td>28.0    </td><td>53.00000</td><td>2       </td><td>14      </td><td>0       </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllllllllllllllllllllllllllllllllllll}\n",
       "  & ID & hospital\\_number & female & age & pim2 & pelod & p\\_bloodculture & comorbidity & rf\\_pneumonia & rf\\_sepsis & ⋯ & mortality\\_28day & mortality\\_100day & ventilation\\_free\\_day & picu\\_free\\_days & fio2 & map & amplitude & OI\\_cat & ICUduration & mort\\_100day\\\\\n",
       "\\hline\n",
       "\t4 & 1006     & 1        & 1        & 1.72     &  8.2     & 12       & 0        & 0        & 1        & 0        & ⋯        & 0        & 0        &  7       &  0       & 100.00   & 23.0     & 70.00000 & 4        & 59       & 0       \\\\\n",
       "\t6 & 1009     & 1        & 0        & 0.70     &  2.8     & 10       & 0        & 1        & 1        & 0        & ⋯        & 1        & 1        &  0       &  0       & 100.00   & 32.0     & 43.66667 & 1        &  9       & 1       \\\\\n",
       "\t8 & 1011     & 1        & 1        & 1.31     & 27.6     &  1       & 1        & 1        & 1        & 0        & ⋯        & 0        & 0        &  0       &  0       & 100.00   & 30.0     & 75.00000 & 4        & 61       & 0       \\\\\n",
       "\t18 & 1022     & 1        & 1        & 0.28     &  9.0     & 20       & 0        & 1        & 1        & 0        & ⋯        & 0        & 0        & 16       & 15       &  58.75   & 28.0     & 50.00000 & 2        & 14       & 0       \\\\\n",
       "\t20 & 1024     & 1        & 0        & 5.34     & 75.6     & 41       & 1        & 1        & 1        & 0        & ⋯        & 1        & 1        &  0       &  0       & 100.00   & 25.5     & 62.00000 & 1        &  3       & 1       \\\\\n",
       "\t27 & 1037     & 1        & 1        & 0.28     &  8.9     & 20       & 0        & 1        & 1        & 0        & ⋯        & 0        & 0        & 16       & 15       &  67.00   & 28.0     & 53.00000 & 2        & 14       & 0       \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | ID | hospital_number | female | age | pim2 | pelod | p_bloodculture | comorbidity | rf_pneumonia | rf_sepsis | ⋯ | mortality_28day | mortality_100day | ventilation_free_day | picu_free_days | fio2 | map | amplitude | OI_cat | ICUduration | mort_100day | \n",
       "|---|---|---|---|---|---|\n",
       "| 4 | 1006     | 1        | 1        | 1.72     |  8.2     | 12       | 0        | 0        | 1        | 0        | ⋯        | 0        | 0        |  7       |  0       | 100.00   | 23.0     | 70.00000 | 4        | 59       | 0        | \n",
       "| 6 | 1009     | 1        | 0        | 0.70     |  2.8     | 10       | 0        | 1        | 1        | 0        | ⋯        | 1        | 1        |  0       |  0       | 100.00   | 32.0     | 43.66667 | 1        |  9       | 1        | \n",
       "| 8 | 1011     | 1        | 1        | 1.31     | 27.6     |  1       | 1        | 1        | 1        | 0        | ⋯        | 0        | 0        |  0       |  0       | 100.00   | 30.0     | 75.00000 | 4        | 61       | 0        | \n",
       "| 18 | 1022     | 1        | 1        | 0.28     |  9.0     | 20       | 0        | 1        | 1        | 0        | ⋯        | 0        | 0        | 16       | 15       |  58.75   | 28.0     | 50.00000 | 2        | 14       | 0        | \n",
       "| 20 | 1024     | 1        | 0        | 5.34     | 75.6     | 41       | 1        | 1        | 1        | 0        | ⋯        | 1        | 1        |  0       |  0       | 100.00   | 25.5     | 62.00000 | 1        |  3       | 1        | \n",
       "| 27 | 1037     | 1        | 1        | 0.28     |  8.9     | 20       | 0        | 1        | 1        | 0        | ⋯        | 0        | 0        | 16       | 15       |  67.00   | 28.0     | 53.00000 | 2        | 14       | 0        | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "   ID   hospital_number female age  pim2 pelod p_bloodculture comorbidity\n",
       "4  1006 1               1      1.72  8.2 12    0              0          \n",
       "6  1009 1               0      0.70  2.8 10    0              1          \n",
       "8  1011 1               1      1.31 27.6  1    1              1          \n",
       "18 1022 1               1      0.28  9.0 20    0              1          \n",
       "20 1024 1               0      5.34 75.6 41    1              1          \n",
       "27 1037 1               1      0.28  8.9 20    0              1          \n",
       "   rf_pneumonia rf_sepsis ⋯ mortality_28day mortality_100day\n",
       "4  1            0         ⋯ 0               0               \n",
       "6  1            0         ⋯ 1               1               \n",
       "8  1            0         ⋯ 0               0               \n",
       "18 1            0         ⋯ 0               0               \n",
       "20 1            0         ⋯ 1               1               \n",
       "27 1            0         ⋯ 0               0               \n",
       "   ventilation_free_day picu_free_days fio2   map  amplitude OI_cat ICUduration\n",
       "4   7                    0             100.00 23.0 70.00000  4      59         \n",
       "6   0                    0             100.00 32.0 43.66667  1       9         \n",
       "8   0                    0             100.00 30.0 75.00000  4      61         \n",
       "18 16                   15              58.75 28.0 50.00000  2      14         \n",
       "20  0                    0             100.00 25.5 62.00000  1       3         \n",
       "27 16                   15              67.00 28.0 53.00000  2      14         \n",
       "   mort_100day\n",
       "4  0          \n",
       "6  1          \n",
       "8  0          \n",
       "18 0          \n",
       "20 1          \n",
       "27 0          "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                     Stratified by treatment\n",
      "                                      0                 1                \n",
      "  n                                    118               118             \n",
      "  female = 1 (%)                        60 ( 50.8)        63 ( 53.4)     \n",
      "  age (median [IQR])                   2.0 [0.8, 5.8]    2.3 [0.8, 5.5]  \n",
      "  pim2 (median [IQR])                  9.4 [4.1, 18.2]   8.0 [4.7, 27.6] \n",
      "  pelod (median [IQR])                10.5 [1.0, 13.8]  10.0 [1.2, 16.0] \n",
      "  p_bloodculture = 1 (%)                22 ( 18.6)        21 ( 17.8)     \n",
      "  MODS = 1 (%)                          54 ( 45.8)        54 ( 45.8)     \n",
      "  comorbidity = 1 (%)                   66 ( 55.9)        66 ( 55.9)     \n",
      "  rf_pneumonia = 1 (%)                 100 ( 84.7)       102 ( 86.4)     \n",
      "  rf_sepsis = 1 (%)                     43 ( 36.4)        32 ( 27.1)     \n",
      "  rf_aspiration = 1 (%)                  5 (  4.2)         4 (  3.4)     \n",
      "  rf_transfusion = 1 (%)                 1 (  0.8)         3 (  2.5)     \n",
      "  rf_trauma = 0 (%)                    118 (100.0)       118 (100.0)     \n",
      "  rf_drowning = 1 (%)                    2 (  1.7)         3 (  2.5)     \n",
      "  OI (median [IQR])                   18.0 [12.3, 29.0] 18.5 [12.0, 28.1]\n",
      "  OI_cat (%)                                                             \n",
      "     1                                  14 ( 11.9)        10 (  8.5)     \n",
      "     2                                  34 ( 28.8)        35 ( 29.7)     \n",
      "     3                                   2 (  1.7)         3 (  2.5)     \n",
      "     4                                  68 ( 57.6)        70 ( 59.3)     \n",
      "  ventilation_free_day (median [IQR])  5.5 [0.0, 18.8]   4.0 [0.0, 16.0] \n",
      "  picu_free_days (median [IQR])        1.5 [0.0, 15.8]   0.0 [0.0, 11.0] \n",
      "  mortality_28day = 1 (%)               21 ( 17.8)        38 ( 32.2)     \n",
      "  mort_100day = 1 (%)                   51 ( 43.2)        54 ( 45.8)     \n",
      "  died_in_ICU = 1 (%)                   50 ( 42.4)        48 ( 40.7)     \n",
      "                                     Stratified by treatment\n",
      "                                      p       test    SMD    \n",
      "  n                                                          \n",
      "  female = 1 (%)                       0.7944          0.0509\n",
      "  age (median [IQR])                   0.6994 nonnorm  0.0084\n",
      "  pim2 (median [IQR])                  0.4399 nonnorm  0.2461\n",
      "  pelod (median [IQR])                 0.7557 nonnorm  0.1259\n",
      "  p_bloodculture = 1 (%)               1.0000          0.0220\n",
      "  MODS = 1 (%)                         1.0000         <0.0001\n",
      "  comorbidity = 1 (%)                  1.0000         <0.0001\n",
      "  rf_pneumonia = 1 (%)                 0.8529          0.0483\n",
      "  rf_sepsis = 1 (%)                    0.1621          0.2012\n",
      "  rf_aspiration = 1 (%)                1.0000          0.0443\n",
      "  rf_transfusion = 1 (%)               0.6141          0.1316\n",
      "  rf_trauma = 0 (%)                    NA             <0.0001\n",
      "  rf_drowning = 1 (%)                  1.0000          0.0589\n",
      "  OI (median [IQR])                    0.8953 nonnorm  0.0164\n",
      "  OI_cat (%)                           0.8230          0.1244\n",
      "     1                                                       \n",
      "     2                                                       \n",
      "     3                                                       \n",
      "     4                                                       \n",
      "  ventilation_free_day (median [IQR])  0.2694 nonnorm  0.1501\n",
      "  picu_free_days (median [IQR])        0.0872 nonnorm  0.2607\n",
      "  mortality_28day = 1 (%)              0.0162          0.3374\n",
      "  mort_100day = 1 (%)                  0.7933          0.0512\n",
      "  died_in_ICU = 1 (%)                  0.8949          0.0344\n"
     ]
    }
   ],
   "source": [
    "mydata$mort_100day<-ifelse(kkh$Died.within.100.days==\"Yes\",1,0)\n",
    "\n",
    "matchedID_hfv<-mydata[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-mydata[m.out$index.control,\"ID\"]\n",
    "hfv<-mydata[match(matchedID_hfv,mydata$ID),]\n",
    "non_hfv<-mydata[match(matchedID_non_hfv,mydata$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "#y_treatment<-data$mortality_28day[data$treatment==1] \n",
    "#y_control<-data$mortality_28day[data$treatment==0] \n",
    "\n",
    "#calculate odds ratio for ICU mortality\n",
    "#y_treatment<-hfv[died_in_ICU] \n",
    "#y_control<-non_hfv$died_in_ICU\n",
    "\n",
    "########## calculate Odd's ratio here!\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day)\n",
    "b<-sum(non_hfv$mortality_28day)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    " print(c(a,b,c,d))\n",
    " a/(a+c)\n",
    " b/(b+d)\n",
    " oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Subgroup Analysis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 1 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=1,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=5.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 2 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=2,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=5.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 3 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=3,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=5.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 4 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=4,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=5.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 6 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=6,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=4.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=4.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 6 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=6,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=4.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=4.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 7 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=7,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=4.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 8 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=8,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=5.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 9 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=9,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=5.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 10 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=10,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=5.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=5.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "####################################################\n",
    "#Subgroup Analysis: Drop one out method on daily matching\n",
    "####################################################\n",
    "#unique(mydata$hospital_number)\n",
    "\n",
    "############### Drop hospital 11 out from data ##############\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "var_list<-c(\"age\",\"pim2\",\"pelod\",\"female\",\"p_bloodculture\",\"comorbidity\",\"MODS\",\"rf_pneumonia\",'rf_sepsis','rf_aspiration',\n",
    "            'rf_transfusion','rf_trauma','rf_drowning',\"OI\")\n",
    "\n",
    "\n",
    "gen_treat_data<-data.frame(mydata)\n",
    "df_not<-gen_treat_data[gen_treat_data$hospital_number!=11,]\n",
    "pop_size<-length(df_not[,1])\n",
    "#str(gen_treat_data)\n",
    "force_exact<-c(rep(0, 13),1)\n",
    "gen.out<-GenMatch(Tr=df_not$treatment,X=df_not[,var_list], pop.size=pop_size,replace=FALSE,exact=force_exact,distance.tolerance=4.0,print.level=0)\n",
    "m.out<-Match(Tr=df_not$treatment, X=df_not[,var_list], estimand = \"ATT\", M = 1,\n",
    "             exact =force_exact,distance.tolerance=4.0, caliper = NULL, replace=TRUE, ties=FALSE,\n",
    "             Weight.matrix = gen.out\n",
    "            )\n",
    "summary(m.out)\n",
    "\n",
    "matchedID_hfv<-df_not[m.out$index.treated,\"ID\"]\n",
    "matchedID_non_hfv<-df_not[m.out$index.control,\"ID\"]\n",
    "hfv<-df_not[match(matchedID_hfv,df_not$ID),]\n",
    "non_hfv<-df_not[match(matchedID_non_hfv,df_not$ID),]\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "#paired t-test for picu free days\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/length(matchedID_hfv)\n",
    "\n",
    "#calculate n% of 28 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 28day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "print(\"HFOV group 28day mortality n:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mortality_28day==1])/length(matchedID_hfv)\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "#y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "#y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "#table(y_treatment,y_control)\n",
    "mcnemar.test(hfv$mortality_28day, non_hfv$mortality_28day)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "########## calculate Odd's ratio here!\n",
    "a<-sum(hfv$mortality_28day==1)\n",
    "b<-sum(non_hfv$mortality_28day==1)\n",
    "c<-sum(hfv$mortality_28day==0)\n",
    "d<-sum(non_hfv$mortality_28day==0)\n",
    "# c<-length(hfv[hfv$died_in_ICU==0,1])\n",
    "# d<-length(non_hfv[non_hfv$died_in_ICU==0,1])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#Table1 for matching pairs\n",
    "\n",
    "PSmatching_data<-rbind(hfv,non_hfv)\n",
    "#PSmatching_data<-read.csv(\"PS_matching_data_significant_result.csv\")\n",
    "head(PSmatching_data)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "##### plot Odds Ratio for the drop one out subgroup analysis\n",
    "drop_out<-read.csv(\"gm_drop_out.csv\")\n",
    "drop_out\n",
    "df<-data.frame(drop_out[-1,])\n",
    "#df$group<-c(1,2,3,4,5,6,7,8,9,10,11)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "tiff(\"dropout_grey.tif\", units=\"in\", width=7, height=5.5, res=300)\n",
    "\n",
    "#png(\"dropout_grey.png\")\n",
    "p <- ggplot(df, aes(OR, Sequence))\n",
    "p + geom_point(size=4, shape=21, fill=\"black\",alpha=1) +\n",
    "  geom_errorbarh(mapping=aes(xmax = upper, xmin = lower),height=0.2, size=1, color=\"black\", alpha=0.6)+\n",
    "geom_vline(xintercept=1.0, color=\"black\",linetype=c)+\n",
    "coord_cartesian(xlim = c(0.8, 5.9), ylim = c(0,11)) +\n",
    "scale_x_discrete(limits=c(\"1\",\"2\",\"3\",\"4\",\"5\",\"6\"))+\n",
    "labs(x = \"Odds Ratio (95% CI)\",y =  \"Subgroup Analysis\" )+\n",
    "geom_text(aes(x=c(5.58,5.58,5.58,5.58,5.58,5.75,5.75,5.75,5.75,5.58),\n",
    "              y = c(10.35,9.3,8.3,7.3,6.2,5.2,4.2,3.3,2.2,1.2),label=sprintf(\"%0.3f\", round(p.value, digits = 3))),colour = \"grey32\",size=4, fontface = \"bold\")+\n",
    "annotate(\"text\", label = \"p-value=\", x = c(5.0,5.0,5.0,5.0,5.0,5.15,5.15,5.15,5.15,4.98), y = c(10.35,9.3,8.3,7.3,6.2,5.2,4.2,3.3,2.2,1.2), size = 4, colour = \"grey32\",fontface = \"bold\")+\n",
    "theme_bw()+\n",
    "scale_y_discrete(\n",
    "                    limits=c(\"1\",\"2\",\"3\",\"4\",\"5\",\"6\",\"7\",\"8\",\"9\",\"10\"))+\n",
    "theme(axis.text.y = element_text(size=18),axis.title.y = element_text(color=\"black\", size=14, face=\"bold\"),\n",
    "      axis.text.x = element_text(size=18),\n",
    "      axis.title.x = element_text(color=\"black\", size=12, face=\"bold\"),)\n",
    "\n",
    "dev.off()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PSM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "####################################################\n",
    "#Alternative Method: PS matching (average)\n",
    "####################################################\n",
    "psmodel <-glm(treatment~female+age+pim2+pelod+p_bloodculture+comorbidity+rf_pneumonia+rf_sepsis\n",
    "                 +rf_aspiration+rf_transfusion+rf_trauma+rf_drowning+MODS+OI_cat, data = mydata,family=binomial(link=\"logit\"))\n",
    "#show coefficients etc\n",
    "#summary(psmodel)\n",
    "#create propensity score\n",
    "pscore<-psmodel$fitted.values\n",
    "drops <- c(\"fio2\",\"map\",\"amplitude\")\n",
    "set.seed(19)\n",
    "mydata_wth_pscore<-data.frame(cbind(mydata[ , !(colnames(mydata) %in% drops)],pscore))\n",
    "\n",
    "m.out <- matchit(treatment~female+age+pim2+pelod+p_bloodculture+comorbidity+rf_pneumonia+rf_sepsis\n",
    "                 +rf_aspiration+rf_transfusion+rf_trauma+rf_drowning+MODS+OI_cat, data = mydata_wth_pscore,              \n",
    "                 method = \"nearest\", distance=mydata_wth_pscore$pscore,caliper=0.01\n",
    "                 ,ratio=1)\n",
    "\n",
    "\n",
    "matchedID_hfv<-match.data(m.out, \"treat\")$ID\n",
    "matchedID_non_hfv<-match.data(m.out, \"control\")$ID\n",
    "pairs<-data.frame(cbind(matchedID_hfv,matchedID_non_hfv))\n",
    "write.csv(pairs, file = \"PS_pair_ID.csv\")\n",
    "###################################\n",
    "#outcome analysis for PS matching\n",
    "y_treatment<-match.data(m.out)$mortality_28day[match.data(m.out, \"all\")$treatment==1] \n",
    "y_control<-match.data(m.out)$mortality_28day[match.data(m.out, \"all\")$treatment==0] \n",
    "#head(mydata)\n",
    "#colnames(mydata)\n",
    "#y_treatment<-mydata$mortality_28day[mydata$treatment==1] \n",
    "#y_control<-mydata$mortality_28day[mydata$treatment==0] \n",
    "length(y_treatment)\n",
    "length(y_control)\n",
    "#y_treatment\n",
    "#table(y_treatment,y_control)\n",
    "#McNemar test for PICU mortality\n",
    "mcnemar.test(y_treatment, y_control)\n",
    "\n",
    "########## calculate Odd's ratio here!\n",
    "a<-length(y_treatment[y_treatment==1])\n",
    "b<-length(y_control[y_control==1])\n",
    "c<-length(y_treatment[y_treatment==0])\n",
    "d<-length(y_control[y_control==0])\n",
    "\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "########## Tableone for PS 55 pairs ####################\n",
    "########## Note that age, pim2 and pelod are skewed data\n",
    "##### so we used IQR and Kruskal-Wallis test instead\n",
    "\n",
    "\n",
    "#Table1 for PS (average) pairs\n",
    "PSmatching_pairs<-read.csv(\"PS_pair_ID.csv\")\n",
    "mydata$mort_100day<-ifelse(kkh$Died.within.100.days==\"Yes\",1,0)\n",
    "PSmatching_data<-mydata[(mydata$ID %in% PSmatching_pairs$matchedID_hfv)| (mydata$ID %in% PSmatching_pairs$matchedID_non_hfv),]\n",
    "head(PSmatching_data)\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\")\n",
    "\n",
    "\n",
    "TableOne_daily<-CreateTableOne(var = xvars, factorVars =catvars,strata =\"treatment\", data=PSmatching_data)\n",
    "\n",
    "print(TableOne_daily, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 2, smd = TRUE)\n",
    "#non parametric KruskalWalli test for PICU free days\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## IPTW"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "Call:\n",
       "glm(formula = treatment ~ female + age + pim2 + pelod + p_bloodculture + \n",
       "    comorbidity + rf_pneumonia + rf_sepsis + rf_aspiration + \n",
       "    rf_transfusion + rf_drowning + MODS + factor(OI_cat), family = binomial(link = \"logit\"), \n",
       "    data = mydata)\n",
       "\n",
       "Deviance Residuals: \n",
       "    Min       1Q   Median       3Q      Max  \n",
       "-1.9023  -0.8269  -0.4074   0.8808   2.7528  \n",
       "\n",
       "Coefficients:\n",
       "                 Estimate Std. Error z value Pr(>|z|)    \n",
       "(Intercept)     -3.099726   0.618075  -5.015 5.30e-07 ***\n",
       "female           0.392725   0.282944   1.388   0.1651    \n",
       "age             -0.007870   0.030752  -0.256   0.7980    \n",
       "pim2             0.005246   0.008290   0.633   0.5268    \n",
       "pelod            0.026776   0.017062   1.569   0.1166    \n",
       "p_bloodculture   0.431948   0.406125   1.064   0.2875    \n",
       "comorbidity      0.093346   0.289559   0.322   0.7472    \n",
       "rf_pneumonia     0.509078   0.445919   1.142   0.2536    \n",
       "rf_sepsis       -0.725136   0.366442  -1.979   0.0478 *  \n",
       "rf_aspiration   -0.060818   0.692903  -0.088   0.9301    \n",
       "rf_transfusion   2.296029   1.135860   2.021   0.0432 *  \n",
       "rf_drowning     -0.139380   0.955147  -0.146   0.8840    \n",
       "MODS            -0.196494   0.311507  -0.631   0.5282    \n",
       "factor(OI_cat)2  1.801127   0.419753   4.291 1.78e-05 ***\n",
       "factor(OI_cat)3  0.101830   0.729003   0.140   0.8889    \n",
       "factor(OI_cat)4  3.045389   0.423836   7.185 6.71e-13 ***\n",
       "---\n",
       "Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1\n",
       "\n",
       "(Dispersion parameter for binomial family taken to be 1)\n",
       "\n",
       "    Null deviance: 432.95  on 327  degrees of freedom\n",
       "Residual deviance: 334.65  on 312  degrees of freedom\n",
       "AIC: 366.65\n",
       "\n",
       "Number of Fisher Scoring iterations: 5\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[1]]\n",
      "[1] 0.812311\n",
      "\n",
      "                                     Stratified by treatment\n",
      "                                      0                 1                \n",
      "  n                                   326.5             362.4            \n",
      "  female = 1 (%)                      159.5 (48.8)      139.4 (38.5)     \n",
      "  age (median [IQR])                    1.8 [0.5, 5.8]    1.8 [0.3, 5.9] \n",
      "  pim2 (median [IQR])                   9.2 [4.1, 18.5]   6.8 [2.7, 26.6]\n",
      "  pelod (median [IQR])                 10.0 [1.0, 13.0]  10.0 [2.0, 15.1]\n",
      "  p_bloodculture = 1 (%)               51.3 (15.7)       46.1 (12.7)     \n",
      "  MODS = 1 (%)                        137.5 (42.1)      151.2 (41.7)     \n",
      "  comorbidity = 1 (%)                 163.8 (50.2)      162.7 (44.9)     \n",
      "  rf_pneumonia = 1 (%)                268.5 (82.2)      255.5 (70.5)     \n",
      "  rf_sepsis = 1 (%)                    93.1 (28.5)      139.8 (38.6)     \n",
      "  rf_aspiration = 1 (%)                13.8 ( 4.2)        9.1 ( 2.5)     \n",
      "  rf_transfusion = 1 (%)                3.1 ( 1.0)        4.5 ( 1.2)     \n",
      "  rf_trauma = 1 (%)                     5.3 ( 1.6)        0.0 ( 0.0)     \n",
      "  rf_drowning = 1 (%)                  11.1 ( 3.4)        7.5 ( 2.1)     \n",
      "  OI (median [IQR])                    10.5 [6.0, 18.0]  10.4 [6.5, 19.0]\n",
      "  OI_cat (%)                                                             \n",
      "     1                                 94.8 (29.0)      138.2 (38.1)     \n",
      "     2                                 91.8 (28.1)       87.5 (24.2)     \n",
      "     3                                 27.1 ( 8.3)       22.7 ( 6.3)     \n",
      "     4                                112.8 (34.6)      113.9 (31.4)     \n",
      "  ventilation_free_day (median [IQR])  15.0 [0.0, 22.0]   6.6 [0.0, 18.9]\n",
      "  picu_free_days (median [IQR])        10.0 [0.0, 19.0]   0.0 [0.0, 15.0]\n",
      "  mortality_28day = 1 (%)              62.3 (19.1)      120.5 (33.3)     \n",
      "  mort_100day = 1 (%)                 103.7 (31.8)      164.0 (45.3)     \n",
      "  died_in_ICU = 1 (%)                  97.5 (29.9)      151.1 (41.7)     \n",
      "                                     Stratified by treatment\n",
      "                                      p       test    SMD    \n",
      "  n                                                          \n",
      "  female = 1 (%)                       0.2198          0.2101\n",
      "  age (median [IQR])                   0.9052 nonnorm  0.1142\n",
      "  pim2 (median [IQR])                  0.6992 nonnorm  0.0059\n",
      "  pelod (median [IQR])                 0.6948 nonnorm  0.0233\n",
      "  p_bloodculture = 1 (%)               0.5163          0.0860\n",
      "  MODS = 1 (%)                         0.9657          0.0080\n",
      "  comorbidity = 1 (%)                  0.5631          0.1057\n",
      "  rf_pneumonia = 1 (%)                 0.2259          0.2786\n",
      "  rf_sepsis = 1 (%)                    0.3084          0.2143\n",
      "  rf_aspiration = 1 (%)                0.3940          0.0952\n",
      "  rf_transfusion = 1 (%)               0.7946          0.0259\n",
      "  rf_trauma = 1 (%)                    0.0407          0.1814\n",
      "  rf_drowning = 1 (%)                  0.4892          0.0816\n",
      "  OI (median [IQR])                    0.7972 nonnorm  0.1400\n",
      "  OI_cat (%)                           0.6705          0.1994\n",
      "     1                                                       \n",
      "     2                                                       \n",
      "     3                                                       \n",
      "     4                                                       \n",
      "  ventilation_free_day (median [IQR])  0.0516 nonnorm  0.3392\n",
      "  picu_free_days (median [IQR])        0.1289 nonnorm  0.3516\n",
      "  mortality_28day = 1 (%)              0.0708          0.3273\n",
      "  mort_100day = 1 (%)                  0.1339          0.2803\n",
      "  died_in_ICU = 1 (%)                  0.1797          0.2489\n"
     ]
    }
   ],
   "source": [
    "#To get table 1 for the weighted cohort\n",
    "\n",
    "#fit a propensity score model with logistic regression\n",
    "psmodel <-glm(treatment~female+age+pim2+pelod+p_bloodculture+comorbidity+rf_pneumonia+rf_sepsis\n",
    "                 +rf_aspiration+rf_transfusion+rf_drowning+MODS+factor(OI_cat)\n",
    "                 , data = mydata,family=binomial(link=\"logit\"))\n",
    "#show coefficients etc\n",
    "summary(psmodel)\n",
    "\n",
    "\n",
    "xvars<-c(\"female\",\"age\",\"pim2\",\"pelod\",\"p_bloodculture\",\"MODS\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \"rf_trauma\",\"rf_drowning\",\n",
    "         \"OI\",\"OI_cat\",\"ventilation_free_day\",\"picu_free_days\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "catvars<-c(\"female\",\"p_bloodculture\",\"comorbidity\",\"rf_pneumonia\",\"rf_sepsis\",\"rf_aspiration\" , \"rf_transfusion\", \n",
    "           \"rf_drowning\",\"rf_trauma\",\"MODS\",\"OI_cat\",\"mortality_28day\",\"mort_100day\",\"died_in_ICU\")\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "#create propensity score\n",
    "pscore<-psmodel$fitted.values\n",
    "label<-mydata$treatment\n",
    "perf<- ROCR::prediction(pscore, label)\n",
    "psm_auc<-ROCR::performance(ROCR::prediction(pscore, label), \"auc\")@y.values[1]\n",
    "print(psm_auc)\n",
    "\n",
    "\n",
    "mydata$weight<-ifelse(treatment==1,1/pscore, 1/(1-pscore))\n",
    "weighteddata<-svydesign(ids=~1, data= mydata, weights = ~weight)\n",
    "weighted_table<-svyCreateTableOne(var = xvars, factorVars = catvars, strata =\"treatment\",\n",
    "                                data=weighteddata)\n",
    "#print(weighted_table, smd=TRUE)\n",
    "print(weighted_table, nonnormal = c(\"age\",\"pim2\",\"pelod\",\"OI\",\"ventilation_free_day\",\"picu_free_days\"), contDigits = 1, catDigits = 1, pDigits = 4, smd = TRUE,testName=TRUE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\n",
       "\tMcNemar's Chi-squared test with continuity correction\n",
       "\n",
       "data:  matrix(c(a, b, c, d), ncol = 2, nrow = 2)\n",
       "McNemar's chi-squared = 104.86, df = 1, p-value < 2.2e-16\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<dl>\n",
       "\t<dt>$data</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>Disease1</th><th scope=col>Disease2</th><th scope=col>Total</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>121</td><td>242</td><td>363</td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td> 62</td><td>264</td><td>326</td></tr>\n",
       "\t<tr><th scope=row>Total</th><td>183</td><td>506</td><td>689</td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$measure</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>estimate</th><th scope=col>lower</th><th scope=col>upper</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>1.000000</td><td>     NA </td><td>      NA</td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td>2.124587</td><td>1.49803 </td><td>3.036639</td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$p.value</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>midp.exact</th><th scope=col>fisher.exact</th><th scope=col>chi.square</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>          NA</td><td>          NA</td><td>          NA</td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td>2.009576e-05</td><td>2.195544e-05</td><td>2.159173e-05</td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$correction</dt>\n",
       "\t\t<dd>FALSE</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description}\n",
       "\\item[\\$data] \\begin{tabular}{r|lll}\n",
       "  & Disease1 & Disease2 & Total\\\\\n",
       "\\hline\n",
       "\tExposed1 & 121 & 242 & 363\\\\\n",
       "\tExposed2 &  62 & 264 & 326\\\\\n",
       "\tTotal & 183 & 506 & 689\\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$measure] \\begin{tabular}{r|lll}\n",
       "  & estimate & lower & upper\\\\\n",
       "\\hline\n",
       "\tExposed1 & 1.000000 &      NA  &       NA\\\\\n",
       "\tExposed2 & 2.124587 & 1.49803  & 3.036639\\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$p.value] \\begin{tabular}{r|lll}\n",
       "  & midp.exact & fisher.exact & chi.square\\\\\n",
       "\\hline\n",
       "\tExposed1 &           NA &           NA &           NA\\\\\n",
       "\tExposed2 & 2.009576e-05 & 2.195544e-05 & 2.159173e-05\\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$correction] FALSE\n",
       "\\end{description}\n"
      ],
      "text/markdown": [
       "$data\n",
       ":   \n",
       "| <!--/--> | Disease1 | Disease2 | Total | \n",
       "|---|---|---|\n",
       "| Exposed1 | 121 | 242 | 363 | \n",
       "| Exposed2 |  62 | 264 | 326 | \n",
       "| Total | 183 | 506 | 689 | \n",
       "\n",
       "\n",
       "\n",
       "$measure\n",
       ":   \n",
       "| <!--/--> | estimate | lower | upper | \n",
       "|---|---|\n",
       "| Exposed1 | 1.000000 |      NA  |       NA | \n",
       "| Exposed2 | 2.124587 | 1.49803  | 3.036639 | \n",
       "\n",
       "\n",
       "\n",
       "$p.value\n",
       ":   \n",
       "| <!--/--> | midp.exact | fisher.exact | chi.square | \n",
       "|---|---|\n",
       "| Exposed1 |           NA |           NA |           NA | \n",
       "| Exposed2 | 2.009576e-05 | 2.195544e-05 | 2.159173e-05 | \n",
       "\n",
       "\n",
       "\n",
       "$correction\n",
       ":   FALSE\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "$data\n",
       "          Outcome\n",
       "Predictor  Disease1 Disease2 Total\n",
       "  Exposed1      121      242   363\n",
       "  Exposed2       62      264   326\n",
       "  Total         183      506   689\n",
       "\n",
       "$measure\n",
       "          odds ratio with 95% C.I.\n",
       "Predictor  estimate   lower    upper\n",
       "  Exposed1 1.000000      NA       NA\n",
       "  Exposed2 2.124587 1.49803 3.036639\n",
       "\n",
       "$p.value\n",
       "          two-sided\n",
       "Predictor    midp.exact fisher.exact   chi.square\n",
       "  Exposed1           NA           NA           NA\n",
       "  Exposed2 2.009576e-05 2.195544e-05 2.159173e-05\n",
       "\n",
       "$correction\n",
       "[1] FALSE\n",
       "\n",
       "attr(,\"method\")\n",
       "[1] \"median-unbiased estimate & mid-p exact CI\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tDesign-based t-test\n",
       "\n",
       "data:  ventilation_free_day ~ treatment\n",
       "t = -1.6581, df = 326, p-value = 0.09825\n",
       "alternative hypothesis: true difference in mean is not equal to 0\n",
       "95 percent confidence interval:\n",
       " -7.2310135  0.6031956\n",
       "sample estimates:\n",
       "difference in mean \n",
       "         -3.313909 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tDesign-based t-test\n",
       "\n",
       "data:  picu_free_days ~ treatment\n",
       "t = -1.5097, df = 326, p-value = 0.1321\n",
       "alternative hypothesis: true difference in mean is not equal to 0\n",
       "95 percent confidence interval:\n",
       " -7.263533  0.942545\n",
       "sample estimates:\n",
       "difference in mean \n",
       "         -3.160494 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tDesign-based KruskalWallis test\n",
       "\n",
       "data:  ventilation_free_day ~ treatment\n",
       "t = -1.9536, df = 326, p-value = 0.05161\n",
       "alternative hypothesis: true difference in mean rank score is not equal to 0\n",
       "sample estimates:\n",
       "difference in mean rank score \n",
       "                   -0.1055206 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tDesign-based KruskalWallis test\n",
       "\n",
       "data:  picu_free_days ~ treatment\n",
       "t = -1.5222, df = 326, p-value = 0.1289\n",
       "alternative hypothesis: true difference in mean rank score is not equal to 0\n",
       "sample estimates:\n",
       "difference in mean rank score \n",
       "                  -0.08964588 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#IPTW outcome anlaysis\n",
    "\n",
    "#Mcnemar test for 28-day mortality\n",
    "a<-120.5\n",
    "b<-62.3\n",
    "#c<-(362.4-120.5)\n",
    "c<-241.9\n",
    "#d<-(326.5-62.3)\n",
    "d<-264.2\n",
    "\n",
    "mcnemar.test(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "# odds ratio for 28-day mortality\n",
    "a<-121\n",
    "b<-62\n",
    "#c<-(362.4-120.5)\n",
    "c<-242\n",
    "#d<-(326.5-62.3)\n",
    "d<-264\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "svyttest(ventilation_free_day~treatment, weighteddata)\n",
    "#paired t-test for picu free days\n",
    "svyttest(picu_free_days~treatment, weighteddata)\n",
    "\n",
    "# non parametric KruskalWalli test for  ventilation free day\n",
    "svyranktest(ventilation_free_day~treatment, weighteddata, test=\"KruskalWallis\")\n",
    "\n",
    "# non parametric KruskalWalli test for  PICU free day\n",
    "svyranktest(picu_free_days~treatment, weighteddata, test=\"KruskalWallis\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Marginal Structural Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "df3<-df2\n",
    "df3$event_7<-ifelse(df3$end_day==7,df3$mortality_28day, df3$event)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Call:\n",
       "coxph(formula = Surv(start_day, end_day, event) ~ HFV_indicator + \n",
       "    cluster(ID), data = df3, weights = time_model$weights.trunc)\n",
       "\n",
       "  n= 2162, number of events= 34 \n",
       "\n",
       "                 coef exp(coef) se(coef) robust se    z Pr(>|z|)\n",
       "HFV_indicator -0.6243    0.5356   0.4150    0.5203 -1.2     0.23\n",
       "\n",
       "              exp(coef) exp(-coef) lower .95 upper .95\n",
       "HFV_indicator    0.5356      1.867    0.1932     1.485\n",
       "\n",
       "Concordance= 0.556  (se = 0.038 )\n",
       "Rsquare= 0.001   (max possible= 0.189 )\n",
       "Likelihood ratio test= 2.56  on 1 df,   p=0.1\n",
       "Wald test            = 1.44  on 1 df,   p=0.2\n",
       "Score (logrank) test = 2.34  on 1 df,   p=0.1,   Robust = 1.95  p=0.2\n",
       "\n",
       "  (Note: the likelihood ratio and score tests assume independence of\n",
       "     observations within a cluster, the Wald and robust score tests do not)."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "                  rho chisq     p\n",
       "HFV_indicator -0.0558 0.143 0.706"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "## with weight but without confounding adjustment\n",
    "MSM<-coxph(Surv(start_day, end_day, event) ~ HFV_indicator\n",
    "           +cluster(ID), data = df3, weights = time_model$weights.trunc)\n",
    "\n",
    "summary(MSM)\n",
    "\n",
    "fit0<-cox.zph(MSM, global=TRUE)\n",
    "cox.zph(MSM, global=TRUE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. \n",
       " 0.3681  0.7117  0.8676  0.9793  1.0308  4.1047 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "Call:\n",
       "coxph(formula = Surv(start_day, end_day, event) ~ HFV_indicator + \n",
       "    factor(female) + age + pim2 + pelod + factor(p_bloodculture) + \n",
       "    factor(MODS) + factor(comorbidity) + cluster(ID), data = df3, \n",
       "    weights = time_model$weights.trunc)\n",
       "\n",
       "  n= 2162, number of events= 34 \n",
       "\n",
       "                            coef exp(coef) se(coef) robust se      z Pr(>|z|)\n",
       "HFV_indicator            0.29194   1.33902  0.37709   0.57527  0.507   0.6118\n",
       "factor(female)1         -0.12679   0.88092  0.33090   0.44091 -0.288   0.7737\n",
       "age                      0.01876   1.01894  0.03145   0.03533  0.531   0.5954\n",
       "pim2                    -0.01349   0.98660  0.00861   0.01059 -1.273   0.2030\n",
       "pelod                    0.07136   1.07396  0.01704   0.01725  4.137 3.51e-05\n",
       "factor(p_bloodculture)1  0.89457   2.44629  0.35907   0.38859  2.302   0.0213\n",
       "factor(MODS)1            0.68364   1.98107  0.39736   0.57819  1.182   0.2371\n",
       "factor(comorbidity)1     0.95796   2.60637  0.37519   0.46511  2.060   0.0394\n",
       "                           \n",
       "HFV_indicator              \n",
       "factor(female)1            \n",
       "age                        \n",
       "pim2                       \n",
       "pelod                   ***\n",
       "factor(p_bloodculture)1 *  \n",
       "factor(MODS)1              \n",
       "factor(comorbidity)1    *  \n",
       "---\n",
       "Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1\n",
       "\n",
       "                        exp(coef) exp(-coef) lower .95 upper .95\n",
       "HFV_indicator              1.3390     0.7468    0.4336     4.135\n",
       "factor(female)1            0.8809     1.1352    0.3712     2.090\n",
       "age                        1.0189     0.9814    0.9508     1.092\n",
       "pim2                       0.9866     1.0136    0.9663     1.007\n",
       "pelod                      1.0740     0.9311    1.0383     1.111\n",
       "factor(p_bloodculture)1    2.4463     0.4088    1.1422     5.239\n",
       "factor(MODS)1              1.9811     0.5048    0.6379     6.153\n",
       "factor(comorbidity)1       2.6064     0.3837    1.0475     6.485\n",
       "\n",
       "Concordance= 0.808  (se = 0.048 )\n",
       "Rsquare= 0.023   (max possible= 0.186 )\n",
       "Likelihood ratio test= 49.52  on 8 df,   p=5e-08\n",
       "Wald test            = 66.15  on 8 df,   p=3e-11\n",
       "Score (logrank) test = 63.73  on 8 df,   p=9e-11,   Robust = 23.3  p=0.003\n",
       "\n",
       "  (Note: the likelihood ratio and score tests assume independence of\n",
       "     observations within a cluster, the Wald and robust score tests do not)."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "                            rho    chisq      p\n",
       "HFV_indicator           -0.1426  3.77405 0.0521\n",
       "factor(female)1          0.0405  0.15919 0.6899\n",
       "age                     -0.1078  0.81512 0.3666\n",
       "pim2                    -0.0516  0.29199 0.5890\n",
       "pelod                   -0.0122  0.00692 0.9337\n",
       "factor(p_bloodculture)1 -0.2261  2.59700 0.1071\n",
       "factor(MODS)1            0.2033  6.42048 0.0113\n",
       "factor(comorbidity)1    -0.1470  3.88102 0.0488\n",
       "GLOBAL                       NA 11.50730 0.1746"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAADAFBMVEUAAAABAQECAgIDAwME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\nTkSuQmCC",
      "text/plain": [
       "Plot with title “”"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# stabilized weights for MSM Cox PH model for 7 days\n",
    "time_model <- ipwtm(\n",
    "  exposure = HFV_indicator,\n",
    "  family = \"survival\",\n",
    "  numerator = ~1,\n",
    "  denominator = ~ age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(rf_pneumonia)+factor(rf_sepsis)\n",
    "                 +factor(rf_aspiration)+factor(rf_transfusion)+factor(rf_drowning)+factor(MODS)+factor(OI_cat),\n",
    "  \n",
    "  id = ID,\n",
    "  tstart = start_day,\n",
    "  timevar = end_day,\n",
    "  type = \"first\",\n",
    "  data = df3, trunc=0.01)\n",
    "summary(time_model$weights.trunc)\n",
    "\n",
    "\n",
    "######################plot (stabilized) inverse probability weights######################\n",
    "\n",
    "#tiff(\"sw_plot.tiff\", width = 4, height = 4,  units = 'in', res = 300)\n",
    "#png(filename=\"sw_plot.png\")\n",
    "ipwplot(weights = time_model$weights.trunc, timevar = df2$end_day, xlab=\"Days in PICU\", ylab=\"Log of Stabilized Weights\",\n",
    "   binwidth = 1, ylim=c(-2,2))\n",
    "#dev.off()\n",
    "\n",
    "#MSM Cox PH model with adjustment for covariates\n",
    "# MSM<-coxph(Surv(start_day, end_day, event_7) ~ HFV_indicator+factor(female)+pim2+pelod+factor(p_bloodculture)+factor(comorbidity)\n",
    "#            +factor(rf_pneumonia)+factor(rf_sepsis)+factor(rf_aspiration)+factor(rf_transfusion)+factor(rf_drowning)+factor(MODS)\n",
    "#            +factor(OI_cat)+cluster(ID)+age:end_day+pim2:end_day+pelod:end_day, data = df3, weights = time_model$weights.trunc)\n",
    "\n",
    "MSM<-coxph(Surv(start_day, end_day, event) ~ HFV_indicator+factor(female)+age+pim2+pelod+factor(p_bloodculture) +factor(MODS)+factor(comorbidity)\n",
    "           +cluster(ID), data = df3, weights = time_model$weights.trunc)\n",
    "\n",
    "summary(MSM)\n",
    "\n",
    "fit0<-cox.zph(MSM, global=TRUE)\n",
    "cox.zph(MSM, global=TRUE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning message in fitter(X, Y, strats, offset, init, control, weights = weights, :\n",
      "“Loglik converged before variable  11 ; beta may be infinite. ”"
     ]
    },
    {
     "data": {
      "text/plain": [
       "                         rho    chisq        p\n",
       "HFV_indicator        0.17065 3.56e+00 5.91e-02\n",
       "female              -0.11480 3.52e+00 6.08e-02\n",
       "age                  0.09304 8.66e-01 3.52e-01\n",
       "pim2                -0.19556 9.08e+00 2.59e-03\n",
       "pelod                0.02630 8.91e-02 7.65e-01\n",
       "p_bloodculture      -0.12243 1.38e+00 2.40e-01\n",
       "comorbidity         -0.08322 1.20e+00 2.74e-01\n",
       "rf_pneumonia         0.22290 1.42e+01 1.67e-04\n",
       "rf_sepsis            0.01670 2.13e-02 8.84e-01\n",
       "rf_aspiration       -0.18491 8.33e+00 3.90e-03\n",
       "rf_transfusion      -0.21963 4.03e+00 4.47e-02\n",
       "rf_drowning         -0.01107 8.38e-03 9.27e-01\n",
       "MODS                 0.22079 1.89e+01 1.40e-05\n",
       "factor(OI_cat)2      0.04033 1.16e-01 7.33e-01\n",
       "factor(OI_cat)3     -0.22963 1.25e+01 4.05e-04\n",
       "factor(OI_cat)4     -0.36164 1.79e+01 2.35e-05\n",
       "age:end_day         -0.15472 2.51e+00 1.13e-01\n",
       "comorbidity:end_day  0.07478 1.01e+00 3.15e-01\n",
       "pelod:end_day       -0.00226 7.01e-04 9.79e-01\n",
       "GLOBAL                    NA 5.03e+01 1.19e-04"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>png:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{png:} 2"
      ],
      "text/markdown": [
       "**png:** 2"
      ],
      "text/plain": [
       "png \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>png:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{png:} 2"
      ],
      "text/markdown": [
       "**png:** 2"
      ],
      "text/plain": [
       "png \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>png:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{png:} 2"
      ],
      "text/markdown": [
       "**png:** 2"
      ],
      "text/plain": [
       "png \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>png:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{png:} 2"
      ],
      "text/markdown": [
       "**png:** 2"
      ],
      "text/plain": [
       "png \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<strong>png:</strong> 2"
      ],
      "text/latex": [
       "\\textbf{png:} 2"
      ],
      "text/markdown": [
       "**png:** 2"
      ],
      "text/plain": [
       "png \n",
       "  2 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "###### Check for PH assumption of MSM\n",
    "#####      redisual plots for MSM\n",
    "\n",
    "time_model <- ipwtm(\n",
    "  exposure = HFV_indicator,\n",
    "  family = \"survival\",\n",
    "\n",
    "  numerator = ~ female+age+pim2+pelod+p_bloodculture+comorbidity+rf_pneumonia+rf_sepsis\n",
    "                 +rf_aspiration+rf_transfusion+rf_drowning+MODS+factor(OI_cat),\n",
    "  denominator = ~1,\n",
    "  id = ID,\n",
    "  tstart = start_day,\n",
    "  timevar = end_day,\n",
    "  type = \"first\",\n",
    "  data = df2, trunc=0.01)\n",
    "\n",
    "\n",
    "#MSM Cox PH model with adjustment for covariates\n",
    "MSM<-coxph(Surv(start_day, end_day, event) ~ HFV_indicator+female+age+pim2+pelod+p_bloodculture+comorbidity+rf_pneumonia+rf_sepsis\n",
    "                 +rf_aspiration+rf_transfusion+rf_drowning+MODS+factor(OI_cat)+cluster(ID)+age:end_day+comorbidity:end_day+pelod:end_day,\n",
    "   data = df2, weights = time_model$weights.trunc)\n",
    "\n",
    "\n",
    "fit0<-cox.zph(MSM, global=TRUE)\n",
    "cox.zph(MSM, global=TRUE)\n",
    "\n",
    "vars<-c(\"HFOV\",\"Gender(Female)\",\"Age\",\"PIM2\",\"PELOD\",\"Bacteremia\",\"Comorbidity\",\"Pneumonia Risk\",\"Sepsis Risk\",\"Aspiration Risk\",\n",
    "        \"Transfusion Risk\",\"Drowning Risk\",\"Multiorgan dysfunction\",\"Oxygenation index\",\"Cat2\",\"Cat3\",\"Age:Time\",\"Comorbidity:Time\",\"PELOD:Time\")\n",
    "\n",
    "\n",
    "png(\"rplot0.png\",width = 600, height = 600)\n",
    "p1<-plot(fit0, ylab=vars, var=1,col=\"red\")\n",
    "dev.off()\n",
    "\n",
    "\n",
    "png(\"rplot1.png\",width = 600, height = 600)\n",
    "layout(matrix(c(1,2,3,4), 2, 2, byrow = TRUE), \n",
    "   widths=c(1,1,1,1), heights=c(1,1,1,1))\n",
    "#p1<-plot(fit0, ylab=vars, var=1,col=\"red\")\n",
    "\n",
    "p2<-plot(fit0, ylab=vars, var=2,col=\"red\")\n",
    "p3<-plot(fit0, ylab=vars, var=3,col=\"red\")\n",
    "p4<-plot(fit0, ylab=vars, var=4,col=\"red\")\n",
    "p5<-plot(fit0, ylab=vars, var=5,col=\"red\")\n",
    "dev.off()\n",
    "png(\"rplot2.png\",width = 600, height = 600)\n",
    "layout(matrix(c(1,2,3,4), 2, 2, byrow = TRUE), \n",
    "   widths=c(1,1,1,1), heights=c(1,1,1,1))\n",
    "p6<-plot(fit0, ylab=vars, var=6,col=\"red\")\n",
    "p7<-plot(fit0, ylab=vars, var=7,col=\"red\")\n",
    "p8<-plot(fit0, ylab=vars, var=8,col=\"red\")\n",
    "p9<-plot(fit0, ylab=vars, var=9,col=\"red\")\n",
    "dev.off()\n",
    "png(\"rplot3.png\",width = 600, height = 600)\n",
    "layout(matrix(c(1,2,3,4), 2, 2, byrow = TRUE), \n",
    "   widths=c(1,1,1,1), heights=c(1,1,1,1))\n",
    "p10<-plot(fit0, ylab=vars, var=10,col=\"red\")\n",
    "p11<-plot(fit0, ylab=vars, var=11,col=\"red\")\n",
    "p12<-plot(fit0, ylab=vars, var=12,col=\"red\")\n",
    "p13<-plot(fit0, ylab=vars, var=13,col=\"red\")\n",
    "dev.off()\n",
    "png(\"rplot4.png\",width = 600, height = 600)\n",
    "layout(matrix(c(1,2,3,4), 2, 2, byrow = TRUE), \n",
    "   widths=c(1,1,1,1), heights=c(1,1,1,1))\n",
    "p14<-plot(fit0, ylab=vars, var=14,col=\"red\")\n",
    "#p15<-plot(fit0, ylab=vars, var=15,col=\"red\")\n",
    "#p16<-plot(fit0, ylab=vars, var=16,col=\"red\")\n",
    "p17<-plot(fit0, ylab=vars, var=17,col=\"red\")\n",
    "p18<-plot(fit0, ylab=vars, var=18,col=\"red\")\n",
    "p19<-plot(fit0, ylab=vars, var=19,col=\"red\")\n",
    "dev.off()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PSM on a daily basis"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAADAFBMVEUAAAABAQECAgIDAwME\nBAQFBQUGBgYHBwcICAgJCQkKCgoLCwsMDAwNDQ0ODg4PDw8QEBARERESEhITExMUFBQVFRUW\nFhYXFxcYGBgZGRkaGhobGxscHBwdHR0eHh4fHx8gICAhISEiIiIjIyMkJCQlJSUmJiYnJyco\nKCgpKSkqKiorKyssLCwtLS0uLi4vLy8wMDAxMTEyMjIzMzM0NDQ1NTU2NjY3Nzc4ODg5OTk6\nOjo7Ozs8PDw9PT0+Pj4/Pz9AQEBBQUFCQkJDQ0NERERFRUVGRkZHR0dISEhJSUlKSkpLS0tM\nTExNTU1OTk5PT09QUFBRUVFSUlJTU1NUVFRVVVVWVlZXV1dYWFhZWVlaWlpbW1tcXFxdXV1e\nXl5fX19gYGBhYWFiYmJjY2NkZGRlZWVmZmZnZ2doaGhpaWlqampra2tsbGxtbW1ubm5vb29w\ncHBxcXFycnJzc3N0dHR1dXV2dnZ3d3d4eHh5eXl6enp7e3t8fHx9fX1+fn5/f3+AgICBgYGC\ngoKDg4OEhISFhYWGhoaHh4eIiIiJiYmKioqLi4uMjIyNjY2Ojo6Pj4+QkJCRkZGSkpKTk5OU\nlJSVlZWWlpaXl5eYmJiZmZmampqbm5ucnJydnZ2enp6fn5+goKChoaGioqKjo6OkpKSlpaWm\npqanp6eoqKipqamqqqqrq6usrKytra2urq6vr6+wsLCxsbGysrKzs7O0tLS1tbW2tra3t7e4\nuLi5ubm6urq7u7u8vLy9vb2+vr6/v7/AwMDBwcHCwsLDw8PExMTFxcXGxsbHx8fIyMjJycnK\nysrLy8vMzMzNzc3Ozs7Pz8/Q0NDR0dHS0tLT09PU1NTV1dXW1tbX19fY2NjZ2dna2trb29vc\n3Nzd3d3e3t7f39/g4ODh4eHi4uLj4+Pk5OTl5eXm5ubn5+fo6Ojp6enq6urr6+vs7Ozt7e3u\n7u7v7+/w8PDx8fHy8vLz8/P09PT19fX29vb39/f4+Pj5+fn6+vr7+/v8/Pz9/f3+/v7////i\nsF19AAAACXBIWXMAABJ0AAASdAHeZh94AAAgAElEQVR4nO3de4BUZd3A8R8it+WysWgoi4hy\nUbyByFVKULlkWhqgeUPEwny1QrNS883srVBffTVTX+8Zr2Vp3i9Z8KphGSj2qhSoKd7yjmlA\nyGV357znzOzMzjnLHJ7D/OY8s89+P3/s3J4zz3N258tcdtgRD0DZxPYCABcQEqCAkAAFhAQo\nICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQo\nICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQo\nICRAASEBCggJxho/2mB7CVWLkGDsL3KR7SVULUIydYuI/CF3dJDIBM+7UaSLrcWsPb1/x863\nRM992F/i6uJLl8wY1m3o5+/PaEyZuWN6X6mbfME6jStzDyGZMgqpwR90awqL+Zo/j9wYPTcf\nUv7S8ztI1uGbtn2Z+aENR+auS3Z5uezVu4iQTLUK6Zbu3euig9IKaX//Bn3OU9Fz8yE1X7rY\nP/mJg+r9r+du+zLzQ+eL9DlFpuwpcqjKHZxrCMlUq5C2JK2Q9hA5r/W5+ZCaL50qst8H3qZj\nRTqu3+Zl5ocOEXnOf470QU+RVWUu3kmEZKrkQ7umu6bu1mXgIbds9rwv5h79LPHP/ef8I3bt\nOfZLv2/e+o3jd6g56A/L/Nu0f+IHIrt6d+w3xPM23fjpAZ3rx12+xj93nsjYj8/at+teVzRt\nuGBQ591O+nvx9MXX1zzN+S2XvjijT7fRd+RCKlzaT+S24DL/xJOhXSkss2Uh3p+O3rnzwKN+\nl723KVpVfuimDtI7+2LDYSJ3e18SmfjunL5d9v7Rx17kW9BeEZKpUiE15Z88DF9bFNL/9Ws+\n998agy2W7RQc335+S0g/DW7DG8c2jxq2JhvS0Nzp0yZmD/p+0DJ76PpahbSwNnvGUaGQ1vtf\nnvYvDA7vCu1KKKTsQrwfNV/9lxu80KryQzO9RVYUXrXzQ9pv9+wF+74f+Ra0V4RkKgipa/cs\nKQ7pmuDmdOIk/3n9Sd7LS/wT31+2zlsd3O6HTApG/sgf8/HO/pEBO8h2hZDqPhHcfs8Nbnyf\nHZCLYl5wU6zbJXuL7Dioq//13wuTh6/v5WX+JnOWvZm/9N3g/IE7ZLdcXbi08S9/+Uvwe5/n\n/HPDT6cKyyws5E/+GQefPt7/eoEXWlVh6OEi3U8pCsmfcNY+/tfZkW9Be0VIpm6RYi0hfV7k\nOP/gOv+JfVPhGcV5/uF1nveO/7S/+4eed6lIhwWZpoulEJJ0/dbtD3h75V4HOEFkSi6kC5sy\nl/kHu73ivberyOGFySPXFzwLKnpg91V/Jfd5TT/MhRS5tOlzfoKN4X3JLzO/kMyEbLUZ/3SX\n98Kryg99vnNw7UOueDfYPgjp4PVek3+4/euRb0E7RUimSoU0yv+3+a713tp77rlnQ+FmN0Lk\niGCjp/zT93vBTXNmcHJaS0jBg62m66691r9bafCfdwzPhlTn3+Df8i+8yr9wrv+UqTB55PrC\nqWR6iMwLDsdsIaRNx4l0+t/IvhSHFCzkFf8weBy5cXv/aVV4VYXXJZbmHstt/91NuZCe88/7\nsEbk0si3oJ0iJFNFz5H2Lg4p+0ubLtOueD64pPlm19RNZEF26MDgltbYSeT24NRthZC6NP/T\n/dFd35ncQ/Ih7e+fs9o/9ZB/eEZRSJHri4T0pr/FE8GRy1qHtGq0f59zZ3RfikLKLmRh0T8R\n3wuvquUFvo9/fXh2wBcyQUj9sudNFjk18i1opwjJVKmQPpjUfAscs6xws/u7f7A4O3SiyFxv\nlX9yaXDqyUJIA7KXfjS7Y/5pejakA7xcSA974ZAi1xcJ6TH/0uzzpV+3Cun2XiK7P9NqX4pC\nyi7kv4tC+mp4VcWvlP9FTpjqn/xNENKY7DlzRA6NfAvaKUIyVSokL/Pnbw3N3ox6vpe/2a1t\nfgSW/d3o2dm7jEeDU48WvfwdbBn8Ez/pkj9+b2shRa4vEtKf/UtfyC+xOKRNX/fPOPKj1vtS\nFFJ2ITeL9FrW7LXwqsIhXdS4f/Bsyg9pSPaco3KPWYu/Be0UIZkqGVLg+YuDG9JNhZtdvcj3\ng/PXdQ/ODR6Z/Tg4eWk4pL/5ox/0ss3EhxS5vkhIayT3CyPv6+GQMrP80/+5pfchREN6XKRD\n4W1E4VU1D71cZGL290hninw5CKlT8JyqcUDLOvLfgnaKkEyVCGnlxIkTgxvvRj+Wa7M3u5/6\npz4r8sng1bX/8E/7D3cO8P8FX+95/6gPh/SQf+m7/h3HoK2GFLm+yKt2+4iM9J/lv9kzHFKw\n4pZX0Ivll5lfyPvb5V50eG/43nsvDa+qeeh9It3eD0I6ROQ/si82nONl78nkvsi3oJ0iJFMl\nQtrQWWTaoy8t+mb2hawm/9nF7I82e7/xTw277fdn+QfjM9kXGWTs/1wzRMIhLfNPH/7Qb6Zu\n9TlS9PoiIf2Pf/aEBdfsKqGQMsP9k91yv/laGN6X/DLzC/G+4j+2u/2t+0eL7NMUXlXz0I/7\niBz4M/nq6f65y3O/R5p7xzn+Zfs3Rb4F7RQhmSr10O7CwtP0E/2b+LDgyBLPOy1/Zt2L/pjG\n0bkTvcMhbdwtd3atyE5NsSFFri8S0ub9c5fVhEJaXfQKwsORnWleZiGk9/o0D9xxRWRV+aF3\nb5e/rtOyL3/vlHtBovNjrb4F7RMhmSoVUubivtnbUJevBu87u6tTLqSmS3plzz3kpewW62YE\nJ6YtCofkPRn8bqbTf94lwXOo2JAi1xcOyXvvsOCyfW4IhbQ0JqTmZRYW4q3MpX7EC9FVFfbo\nt3tkR/S8oDEb0sSfBufvFLytMPItaJ8IqXwbH1lw8VX3vZM78dL1l9wevAXVW7/slst+80Z+\nTOaFBVctaXjUv0cJbbr+oUvvNXyhK3x9Uc/+5KqnkrxltLDMwgL/9utLblu5pVXlhzY88V05\n4uHsRkFI3hvXzv9V/n/5hb4F7RIhpeGis8++OTj8RvZdN21W8ZtWJ1pdSfUhpDTM8h/SXfDE\n0m90yL9BIW2/PKHYvG28FkIqjZDS8MF++ScrZ9tZwDlSbNdyr46QWiGkVDT8YsqQrn1Gzn7W\n9kJ0EFIrhAQoICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQ\nEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQ\nEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQ\nEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQoICSkYdMjixJ63PaSkyEkpOH+\nDr2T+YS8aHvNiRAS0nB374QbvCcrKrKQSiEkpIGQAAWEBCggJEABIQEKCAlQQEiAAkICFBAS\noICQAAWEBCggJEABIQEKCAlQQEiAAkICFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkICFBAS\noICQAAWEBCggJEABIQEKCAlQQEiAAkICFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkICFBAS\noICQAAWEBCggJEABIQEKCAlQQEiAAkICFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkICFBAS\noICQAAWEBCggJEABIQEKCAlQQEiAAkICFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkICFBAS\noICQAAWEBCggpFjrX1+TUVoInEZIJT1y0tBaEakZfNZyxfXATYRUQmauSO0BU6ZPGVUnMqdR\ndU1wDyGVcKWMXtyQPda4dLJcrLcgOImQShhbv6FwvGHEYJ3FwFmEVELPmUUn5nXWWAocRkgl\njO/fco/UOHKQzmLgLEIq4eqW50hPTpb5eguCkwiphMxpIrWjps6YNqZOZFaD6prgHkIqaeXp\nQ3qISLdB855VXA/cREix1r3GOxtggpC2Yt1Tf+O3sdgqQirh/GuDry8f5j+463r2WsUFwUmE\nVGq7sf6Xd3eQISfO2VtGbNJcEhxESKW2C0I6RS7wH9c1zZfvay4JDiKkUtsFIQ0Z1hQcz+w1\nJnLppp9eX3DNeWUtEE4gpFLbBSHVHJc7cXyPyKVv7LV7Qb3wwA+EVGq7IKQR43MnDh4eM/IJ\nQgIhldxuh3Nvemz+dvcFx38rX44ZSUggpJJ26yiBHT2v6diOXVfGjCQkEFJpm1ctvO7bMyd6\nXoMMfTxuICGBkAw0rYh/lxAhgZAUEBIISQEhgZAUEBIISQEhgZBKqg2LGUlIIKSSbhwlMnB4\nQcxIQgIhldYwTe4xGkhIIKQYDxASjBFSSW91v99oHCGBkBQQEghJASGBkBQQEghJASGBkBQQ\nEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQ\nEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQ\nEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQ\nEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQEghJASGBkBQQ\nEghJASGBkBQQEghJASGlLbNn76Ruq/SaCKlshJS2Jvn+HckMnV/pNRFS2QgpbU2yOOEW4wmp\nXITkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDc\nQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkC\nQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDc40hIv1uVzGuZiqzcECG5x4mQ\nXpbEflmRlRsiJPc4EdIKuTHhPdLAGyqyckOE5B5HQro/4RZDCAmqCMkCQnIPIVlASO4hJAsI\nyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQ\nLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTk\nHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDcQ0gW\nEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkCQnIP\nIVlQXkjrX1+T2doYQkobIVmw7SE9ctLQWhGpGXzW8thxhJQ2QrJgW0PKzBWpPWDK9Cmj6kTm\nNMaMJKS0EZIFxSH9dI35dlfK6MUN2WONSyfLxTEjCSlthGRBcUjS9diHNhtuN7Z+Q+F4w4jB\nMSMJKW2EZEFxSNdMEPnkvKe3+vJBoOfMohPzOseMJKS0EZIF4edIr87fR2TY/Ne3vt34/i33\nSI0jB8WMJKS0EZIFrV5seO6cASKTbv7nVra7uuU50pOTJe7nQEhpIyQLWoW04YHZHUWk67wN\nWxpekDlNpHbU1BnTxtSJzGqIGUlIaSMkC8Ih/evO43qKjL/8+ev3lrO2suXK04f08IvrNmje\ns7HjCClthGRBcUi/mN5N5MAr/h4c37xnvcHW617jnQ3Vh5AsCL38LRN+/Gb+xPRxBlvzFqFq\nREgWFId05Zslh20BbxGqVoRkQfg50sqF/pef/NVgO94iVL0IyYLikDJfkQnBWfK1uDByeItQ\n9SIkC4pDul7GPegfPDBBrt/qdvFvEco8vqjgx4SUshRC+mhRQhf2SjhDWw7poIG5ODbuNnKr\n28W/RWhVFymysexVIokUQjpXkuqQcIa2HFKv2c1HTuqx1e14i1D1SiGkbx6RcINz2lNIe0xs\nPjJx6Fa34y1C1YuQLCgOaW6H27OHd8vJW92OtwhVL0KyoDikDwbIwT+48aLPSt93DLbkLULV\nipAsCP0e6bVZHYLnhZ97wXRr3iJUjQjJgsi7vz9Y8qvH3tKdgZDSRkgW8Hft3ENIFoRCuvPY\nQ5spzkBIaSMkC4pDukmke58cxRkIKW2EZEFxSHt1/73RHz4J1IbFjCSktBGSBcUhdTnNfLsb\nR4kMHF4QM5KQ0kZIFhSH1P+MBBs2TJN7jAYSUtqSh7TXtIuTOWhCwhnaVUjfq1+dYMsHCKlK\nJQ+pZ68Dkqnpm3CGdhVSw8lDFzz//uqAwZZvdTfbU0JK2zaENDXhBqMJKaI4pNrawpveFWcg\npLQRkgXFyXylheIMhJQ2QrKAdza4h5AsCIeUeXXJ+57x75LMEFLaCMmCUEh/2FPkHm/QN9Zq\nzkBIaSMkC4pDWlHTfYYf0n6y59b+hH4ShJQ2QrKgOKQTOyx91Q8pc7mcrTgDIaWNkCwoDqnf\nod6r2d+yHriH4gyElDZCsqA4pJo5zSEd311xBkJKGyFZUBzSmE/lQmrof4DiDISUNkKyoDik\nH8pDQUgbjpbzFGcgpLQRkgWh99p9evuJctgxdbJv/Kf1JUNIaSMkC0K/R9p4eX8R6fPddZoz\nEFLaCMmC6FuE1v71H8ozEFLaCMkC3mvnHkKyoDikE1oozkBIaSMkC8KfIdtsQNynSyRFSGkj\nJAtCr9oFNr/zwMhJ6xVnIKS0EZIFW3qO9M9dz1ScgZDSRkgWbPHFhq/urDgDIaWNkCzYYkhz\nuyrOQEhpIyQLthBS06KafRVnIKS0EZIFxSF1z+kk8jPFGQgpbYRkQXFIhzebda/mDISUNkKy\ngHc2uIeQLCAk9xCSBcUh9Q1TmoGQ0kZIFhSHdMZ4kb4j+4nsNiGgNAMhpY2QLCgOaWXvg5f7\nBysO3uV1xRkIKW2EZEFxSDP75f5H39p+RyvOQEhpIyQLikPa6ZjmI8fUK85ASGkjJAuKQxow\ntvnIWN5r15YRkgXFIX1RfpE9/IV8TnEGQkobIVlQHNIrveXI6++//kjp8oziDISUNkKyIPQL\n2T9Pyv4H2b1+pzkDIaWNkCyIfD7S8jsuW/BEo+oMhJQ2QrKADxpzDyFZwAeNuYeQLOCDxtxD\nSBbwQWPuISQL+KAx9xCSBXzQmHsIyQI+aMw9hGQBHzTmHkKygA8acw8hWcAHjbmHkCzgg8bc\nQ0gWFId01S8rMQMhpY2QLAj9pdVdKzEDIaWNkCwoDunrsrQCMxBS2gjJguKQmi7seeXTqz8K\nKM5ASGkjJAuKQ+rTZ7v8h18qzkBIaSMkC4qTmd1CcQZCShshWcDf/nYPIVmQD+mMGys1AyGl\njZAsyIckM4KvV83Wn4GQ0kZIFoRDmlGBR3qElDZCsoCQ3ENIFhCSewjJAkJyDyFZQEjuISQL\nCMk9hGRBIaT6L/rq5Ys5ijMQUtoIyYJCSGGKMxBS2gjJgnwyy8IUZyCktBGSBbzXzj2EZAEh\nuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCS\nBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDc\nQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkC\nQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4h\nJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEh\nuYeQLCAk9xCSBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCS\nBYTkHkKygJDcQ0gWEJJ7CMkCQnIPIVlASO4hJAsIyT2EZAEhuYeQLCAk9xCSBYTkHkKygJDc\nQ0gWEJJ7CMmC8kJa//qazNbGEFKZ3j58cjKHyj0JpyCksm17SI+cNLRWRGoGn7U8dhwhlWmx\nfPucRL4lVyWcgpDKtq0hZeaK1B4wZfqUUXUicxpjRhJSmRZLU7INNhNS+rY1pCtl9OKG7LHG\npZPl4piRhFQmQjLTNkMaW7+hcLxhxOCYkYRUJkIy0zZD6jmz6MS8zjEjCalMhGSmbYY0vn/L\nPVLjyEExIwmpTIRkpm2GdHXLc6QnJ8v8mJGEVCZCMtM2Q8qcJlI7auqMaWPqRGY1xIwkpDIR\nkpm2GZLnrTx9SA8R6TZo3rOtL3x7VcGvCak8hGSmrYYUWPfalt/Z8LIU21jWHO0eIZlpyyH5\nKT31ty39NvYN7pHUEJKZthnS+dcGX18+zL/D6Xr22riRPEcqEyGZaZshyVj/y7s7yJAT5+wt\nI+JSIaQyEZKZHjW9E4p7Q05SZYV0ilzgP65rmi/fjxlJSGUiJDOdDrojmUknJ5whTlkhDRmW\n/RFn9hoTM5KQykRIZjrNSrjByVUTUs1xuRPH94gZSUhlIiQzbTikEeNzJw4eHjOSkMpESGba\naEg7nHvTY/O3uy84/lv5csxIQioTIZlpmyHt1jH7u9YdPa/p2I5dV8aMJKQyEZKZthmSt3nV\nwuu+PXOi5zXI0MfjBhJSmQjJTBsNqaBpRfzfPyGkMhGSmbYe0tYQUpkIyQwhIRYhmSEkxCIk\nM4SEWIRkhpAQi5DMEBJiEZIZQkIsQjJDSIhFSGYICbEIyQwhIRYhmSEkxCIkM4SEWIRkhpAQ\ni5DMEBJiEZIZQkIsQjJDSIhFSGYICbEIyQwhIRYhmSEkxCIkM4SEWIRkhpAQi5DMEBJiEZIZ\nQkIsQjJDSIhFSGYICbEIyQwhIRYhmSEkxCIkM4SEWIRkhpAQi5DMEBJiEZIZQkIsQjJDSIhF\nSGYICbEIyQwhIRYhmSEkxCIkM4SEWIRkhpAQi5DMEBJiEZIZQkIsQjJDSIhFSGYICbEIyQwh\nIRYhmSEkxCIkM4SEWIRkhpAQi5DMEBJiEZIZQkIsQjJDSIhFSGYICbEIyQwhIRYhmSEkxCIk\nM4SEWIRkhpAQi5DMEBJiEZIZQkIsQjJDSIhFSGYICbEIyQwhIRYhmSEkxCIkM4SEWIRkhpAQ\ni5DMEBJiEZIZQkIsQjJDSIhFSGYIqX1Z/+iiZC4nJCOE1L5cJok1JpuBkMwQUps2f3zCDX4i\nDck2ICQzhNSmEZIRQooipDBCMkJIUYQURkhGCCmKkMIIyQghRRFSGCEZIaQoQgojJCOEFEVI\nYYRkhJCiSof04KlJXVnx1VYeIRkhpKjSIZ08NGFHBw2p+Gorj5CMEFJUTEhJd+QGQjJBSGYI\nqU0jJCOEFEVIYYRkhJCiCCmMkIwQUhQhhRGSEUKKIqQwQjJCSFGEFEZIRggpipDCCMkIIUUR\nUhghGSGkKEIKIyQjhBRFSGGEZISQoggpjJCMEFIUIYURkhFCiiKkMEIyQkhRhBRGSEYIKYqQ\nwgjJCCFFEVIYIRkhpChCCiMkI4QURUhhhGSEkKIIKYyQjBBSlGJI1w54OqEXy129PkIyQkhR\niiHNSfzJQh0+Knf56gjJCCFFKYY0q9OHySyR98pdvjpCMkJIUaohJdxgBSEZIaSyEVLaCMkI\nIUURUhghGSGkKEIKIyQjhBRFSGGEZISQoggpjJCMEFIUIYURkhFCiiKkMEIyQkhRhBRGSEYI\nKYqQwgjJCCFFEVIYIRkhpChCCiMkI4QURUhhhGSEkKLaVkhvJP0fT0+vTzgDIRkhpKi2FdJO\nif/L0/kJZyAkI4QU1bZC6n1rwv/yNO2bCWcgJCOEFNXGQro74QZHEJIJQiqb4yENSPxYsE/C\nGQjJDCEZq8aQ+u6xKJlRPRPOQEhmCMlYVYY0OuEGUwnJBCFFEVIYIRkhpChCCiMkI4QURUhh\nhGSEkKJshvRnmXVqMl1uTjgFIRkhpLLZDOl+mXZ0MnJOwikIyQghlc1uSCsSbtGBkEwQUhQh\nhRGSEUKKIqQwQjJCSFGEFEZIRggpipDCCMkIIUURUhghGSGkKEIKIyQjhBRFSGGEZISQoggp\njJCMEFIUIYURkhFCiiKkMEIyQkhRhBRGSEYIKYqQwgjJCCFFEVIYIRkhpChCCiMkI4QURUhh\nhGSEkKIIKYyQjBBSFCGFEZIRQooipDBCMkJIUeWFtP71NZmtjSGkMEIy0n5CeuSkobUiUjP4\nrOWx4wgpjJCMtJeQMnNFag+YMn3KqDqROY0xIwkpjJCMtJeQrpTRi3M/38alk+XimJGEFEZI\nRtpLSGPrNxSON4wYHDOSkMIIyUh7CannzKIT8zpHLn1lx94FPWVziev4UufeyXSWhBv0kNqE\nW0jXhBtst33CDTp1SLhBTdK97i01CTfo0CnhBttvl3CDrkl3olZ6JNxCEt+cvrSNN/4t2daQ\nxvdvuUdqHDkocmnTYy2fB7Tw56Wu4+2Enyy06MGbk27x30k3uOHhhBv8/M6EG9y7IOEGC69L\nuMGi6xYm3GDBvQk3uPPnCTd4+IaEGyT/0d38YNIt3t7GG/+WbGtIV7c8R3pysszXWxDQFm3z\nq3anidSOmjpj2pg6kVkJH8QDrtn23yOtPH1IDxHpNmjes4rrAdqk8t7ZsO61rb+zAWgHKv9e\nO6AdICRAASEBCggJUEBIgAJCAhQQEqCAkAAFhAQoICRAASEBCggJUEBIgAJCAhQQEqCAkAAF\nhAQoqMaQMtsJUHnjFG+01RhSk9zwtC29LrU29YDzrU293xnWpv70LGtTf+54xRttdYa02Nrc\nve+2NvWQG6xNPd7e31M74pvWpq6Kv7RaSYSUNkIqGyGFEVLKCKlyCClthFQ2QgojpJQRUuUQ\nUtoIqWyEFEZIKSOkyiGktBN9Wh0AAAg2SURBVBFS2QgpjJBSRkiVQ0hpI6SyVWNImU5LrM3d\n90FrU+/9M2tTT7zM2tTTv2Nt6lNPVbyyagzJe8XeZ8W81mht6jdKfdZu5b39sbWp319rbeoP\nP1S8sqoMCWhrCAlQQEiAAkICFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkICFBASoICQAAWE\nBCiotpBurC0+tVvuj52fb2HqzDUH9hx/VRr/MarVTCntdat53d/lnIr8oKsspIbRxTu5ueMO\nYwPXpj+1d5rUz+gnc1OYODpTWnvdag/d3+Wsyvygqyqktx/6jBTv5Co519bUK2X0Bu/jUfJo\nxWduNVNKe91qXvd3OVCpH3RVhdTdv3cv3smFcpOtqU/P/gGWxXJSxWduNVNKe91qXvd3OVCp\nH3RVhXT/PfcMLN7J6+QxW1MPqW3wv26uHVrxmVvNlNJet5rX/V0OVOoHXVUh+YYX7+S35aKR\nNXt8+T0LU/cYlT04oHbLYxW1mimlvW41r/u73KwiP+iqDmmmdBh33J6yw6rUp14nU7OHU6TS\nf1+n9Uzp7HWred3f5byK/KCrOqSJO97reU3/Lp9NferXZEb2cLq8XuFZW8+Uzl63mtf9Xc6r\nyA+6GkJqfMn3du748NZ3sQ2DZV3aU6+RadnDKbKmwlOXmqmSex1oNW8Ku1xq6maV3uW8ivyg\nqyGk1cHv4o7MHd9CSN6J8lTaU2e6jskejqqp2O8nm6cuOVMF9zrQat4UdrnU1HkV3uW8ivyg\nqyGkjb/2PZE7XryTTQ1N2cM58mLaU3u71wV/crWxbnClZi5MHZ0phb3OarWHld/lUjOltcvN\nKvKDroaQihXv5Ao5Njho2qdrKn9HOPT9nSdP+l+XypkVnzY6U1p73WoP3d/lZhX5QVdpSB+/\n+pZ/tzu04+/8rxfJN1Kf2ntGpjZ6DZNlecWnLZop1b2OztsOdrlZRX7QVRrSIhnuf328q0w+\nYR8Zkc7fWQ9N7Z0go84cIbNTmLdlpnT3OjpvO9jlnIr8oKs6JO+vp+zbfdT3NtqYevMPBnba\nfX5DCvO2zJTuXrea1/1dzqnID7raQgLaJEICFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkIC\nFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkICFBASoICQAAWEBCggJEABIQEKCAlQQEiAAkIC\nFBASoICQAAWEBCggJEABIQEKCAlQQEiOeknW2V5Cu0JIVWRCX//LTbt0e8nz1nxrUJfdT3xj\nW6/prTOHyb7z3ldcG+IRUhUJQnqn445ff99bu4/0P+5A6fFC6PIH5NatXUVuyHO9646Wk3vu\n+GalVoooQqoiQUh/lPn+sYvlCw2et0Amhi43DmlSnzf9h3bPbX9KZdaJ1gipigQh/UGu8o+N\nlOCj670DO4Se6JiG1NDllOxzpInDKrRQtEJIVeH56fX1xzznh3SC+J736nbOnn2sLC8aNC24\nbLU3u0/TmT382hp+OLam/qQVwSWrjt+jS/1Ry/JDNnU+uvBiQ5/ZL560yy4z/addXtPNo3vV\nffo3ae9b+0BI1eAP3WXc0Tv3GtDXW3yenHDLP71nXgzOburb4aOiUQvnydxbNvghXSgDbvM2\nfloOnHNIh+5/9LzlXTseNnvfDr3eyA8Z3/HeQkgT6/odPUZqn/K8C+QTRx7cucNjFnbQfYRU\nBZr2l9s8b+0kKTy0y519pkwPjcs9tJu9Xc+l/sHlHR/0vy6pGZbxviZ3+0d/LLfkhzzbXQbL\nzzcE2/SRg9cFz7YmeZneA/8VXP/stHarXSGkKvCkHBkcLA+H9M7RUh9+2a05JPlxcKL/xFcD\n0+VV7/e3NvhnPCxXFJ5GrTq7t8gnvr4uCOmvwRmfkec3d9x1k1/nMy+ltVvtCiFVgVvl+uxh\n36KQMtf0kk+9Fh6XD+lp/+tayVsSXPTBIz/aoygk/1mXXLCrHOqH1C978idyv/+Ma7cL/7gx\nlT1qfwipClzq38oDI1tC+uCz8smbGiPj8iGt9r/+VQ67J+d97/1TdpHuBx0dCsl/jrR+kvzd\n63NA9uRdcrW36b/2E+l5yjsp7VX7QkhV4FfN90j9CiF9PE6O+KjVuHxIwSX/kGPyZ2c+Jf/2\n+0ZvSSGkZSc8l33V7pd+oM33SFfLncHBm7/4jOzbVPk9an8IqQr8nxwVHKxoeY70XTlzCzf3\n4pC8PnXBo7TMfoMyb8sXgjMWFUL6o9yQDelaWeE/R8q+QP45Wf63sx8Ijn1GXkllp9oZQqoC\nmbHBq3brDimE1Niv97+2MO4BuckrhPRdOcZ/6HeFfMP7UCZnPO/DSXJZ85B/dB/zLz+kjfv1\nbvJDOmRd8CxsXOZ1Gbbev68bvv36dPeufSCkavCnnjLumH49JuVDekVqx+a8XTzsf2W/89bl\nQ1q3r/Q/foLsudbzpspRl3ztk4fJsDubh/xI9vqmXDJcrvPvuPp/sv6YcdJziZeZIf2PO7Fe\nvmVnHx1HSFXhhen9d5r5zBn5kB4tvCT3avGoDZ/v0ucf+ZC8j88dWTPkrOD4B6fW9zzo1sy8\n2lOah2R+Pq6XbD8qeFrUZ8Kr03fuNz149+u67+1Z02f8Ap4iVQIhOeol+TB72GeC5YW0E4Tk\nqMJbhAgpFYRU3a7t0+I723IFhJQOQqpu/3qnxZptuQJCSgchAQoICVBASIACQgIUEBKggJAA\nBYQEKCAkQAEhAQoICVBASIACQgIUEBKggJAABYQEKCAkQAEhAQoICVBASIACQgIUEBKggJAA\nBYQEKCAkQAEhAQoICVBASICC/wfLWzfKVrRfeAAAAABJRU5ErkJggg==",
      "text/plain": [
       "Plot with title “Histogram of df2_treat$ps”"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAMAAADKOT/pAAADAFBMVEUAAAABAQECAgIDAwME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\nkAADhAQYICTAACEBBggJMEBIgIH/B+n47dxBln4UAAAAAElFTkSuQmCC",
      "text/plain": [
       "Plot with title “Histogram of df2_control$ps”"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df2$weight_trunc<-time_model$weights.trunc\n",
    "####################################################\n",
    "#Alternative Method: PS matching on a daily basis\n",
    "####################################################\n",
    "\n",
    "########## Prepare Dataset for matching ####################\n",
    "df2_treat<-df2[df2$treatment==1,]\n",
    "df2_control<-df2[df2$treatment==0,]\n",
    "\n",
    "df2_treat$ps<-log(1/df2_treat$weight_trunc)\n",
    "hist(df2_treat$ps)\n",
    "\n",
    "df2_control$ps<-log(1/df2_control$weight_trunc)\n",
    "hist(df2_control$ps)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# ########### Match for Day 2 hfov ####################\n",
    "# treat_day<-df2_treat[df2_treat$end_day==2 & df2_treat$HFV_indicator==1 & !(df2_treat$ID %in% pairs_total$matchedID_hfv),]\n",
    "# control_day<-df2_control[df2_control$end_day==2 & !(df2_control$ID %in% pairs_total$matchedID_non_hfv),]\n",
    "# #nrow(treat_day)\n",
    "# #nrow(control_day)\n",
    "# df_day2<-rbind(treat_day,control_day)\n",
    "# #nrow(df_day2)\n",
    "\n",
    "# m.out <- matchit(treatment~age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(MODS)+factor(OI_cat), data = df_day2,              \n",
    "#                  method = \"nearest\",\n",
    "#                  distance=\"linear.logit\",\n",
    "#                  discard=\"both\",\n",
    "#                  m.order=\"smallest\",\n",
    "#                  calclosest=TRUE)\n",
    "# #summary(m.out)\n",
    "# matchedID_hfv<-match.data(m.out, \"treat\")$ID\n",
    "# matchedID_non_hfv<-match.data(m.out, \"control\")$ID\n",
    "# pairs<-data.frame(cbind(matchedID_hfv,matchedID_non_hfv))\n",
    "# pairs_total<-rbind(pairs_total, pairs)\n",
    "\n",
    "# #write.csv(pairs, file = \"pair_ID.csv\")\n",
    "# nrow(pairs)\n",
    "# #nrow(pairs_total)\n",
    "\n",
    "# ########## Match for Day 3 hfov ####################\n",
    "\n",
    "# treat_day<-df2_treat[df2_treat$end_day==3 & df2_treat$HFV_indicator==1 & !(df2_treat$ID %in% pairs_total$matchedID_hfv),]\n",
    "# control_day<-df2_control[df2_control$end_day==3 & !(df2_control$ID %in% pairs_total$matchedID_non_hfv),]\n",
    "# #nrow(treat_day)\n",
    "# #nrow(control_day)\n",
    "# df_day3<-rbind(treat_day,control_day)\n",
    "# #nrow(df_day3)\n",
    "\n",
    "# m.out <- matchit(treatment~age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(rf_pneumonia)+factor(rf_sepsis)\n",
    "#                  +factor(rf_aspiration)\n",
    "#                  #+factor(rf_transfusion)\n",
    "#                  +factor(rf_drowning)+factor(MODS)+factor(OI_cat), data = df_day3,              \n",
    "#                  method = \"nearest\",\n",
    "#                  distance=\"linear.logit\",\n",
    "#                  discard=\"both\",\n",
    "#                  m.order=\"smallest\",\n",
    "#                  calclosest=TRUE)\n",
    "# #summary(m.out)\n",
    "# matchedID_hfv<-match.data(m.out, \"treat\")$ID\n",
    "# matchedID_non_hfv<-match.data(m.out, \"control\")$ID\n",
    "# pairs<-data.frame(cbind(matchedID_hfv,matchedID_non_hfv))\n",
    "# pairs_total<-rbind(pairs_total, pairs)\n",
    "\n",
    "\n",
    "# nrow(pairs)\n",
    "# #nrow(pairs_total)\n",
    "\n",
    "# ########## Match for Day 4 hfov ####################\n",
    "\n",
    "# treat_day<-df2_treat[df2_treat$end_day==4 & df2_treat$HFV_indicator==1 & !(df2_treat$ID %in% pairs_total$matchedID_hfv),]\n",
    "# control_day<-df2_control[df2_control$end_day==4 & !(df2_control$ID %in% pairs_total$matchedID_non_hfv),]\n",
    "# #nrow(treat_day)\n",
    "# #nrow(control_day)\n",
    "# df_day4<-rbind(treat_day,control_day)\n",
    "# #nrow(df_day4)\n",
    "\n",
    "# m.out <- matchit(treatment~age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(rf_pneumonia)+factor(rf_sepsis)\n",
    "#                  +factor(rf_aspiration)\n",
    "#                  #+factor(rf_transfusion)\n",
    "#                  +factor(rf_drowning)+factor(MODS)+factor(OI_cat),  data = df_day4,              \n",
    "#                  method = \"nearest\",\n",
    "#                  distance=\"linear.logit\",\n",
    "#                  discard=\"both\",\n",
    "#                  m.order=\"smallest\",\n",
    "#                  calclosest=TRUE)\n",
    "# #summary(m.out)\n",
    "# matchedID_hfv<-match.data(m.out, \"treat\")$ID\n",
    "# matchedID_non_hfv<-match.data(m.out, \"control\")$ID\n",
    "# pairs<-data.frame(cbind(matchedID_hfv,matchedID_non_hfv))\n",
    "# pairs_total<-rbind(pairs_total, pairs)\n",
    "\n",
    "\n",
    "# nrow(pairs)\n",
    "# #nrow(pairs_total)\n",
    "\n",
    "# ########## Match for Day 5 hfov ####################\n",
    "\n",
    "# treat_day<-df2_treat[df2_treat$end_day==5 & df2_treat$HFV_indicator==1 & !(df2_treat$ID %in% pairs_total$matchedID_hfv),]\n",
    "# control_day<-df2_control[df2_control$end_day==5 & !(df2_control$ID %in% pairs_total$matchedID_non_hfv),]\n",
    "# #nrow(treat_day)\n",
    "# #nrow(control_day)\n",
    "# df_day5<-rbind(treat_day,control_day)\n",
    "# #nrow(df_day5)\n",
    "\n",
    "# m.out <- matchit(treatment~age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(rf_pneumonia)+factor(rf_sepsis)\n",
    "#                  +factor(rf_aspiration)\n",
    "#                  #+factor(rf_transfusion)\n",
    "#                  +factor(rf_drowning)+factor(MODS)+factor(OI_cat),  data = df_day5,              \n",
    "#                  method = \"nearest\",\n",
    "#                  distance=\"linear.logit\",\n",
    "#                  discard=\"both\",\n",
    "#                  m.order=\"smallest\",\n",
    "#                  calclosest=TRUE)\n",
    "# #summary(m.out)\n",
    "# matchedID_hfv<-match.data(m.out, \"treat\")$ID\n",
    "# matchedID_non_hfv<-match.data(m.out, \"control\")$ID\n",
    "# pairs<-data.frame(cbind(matchedID_hfv,matchedID_non_hfv))\n",
    "# pairs_total<-rbind(pairs_total, pairs)\n",
    "\n",
    "\n",
    "# nrow(pairs)\n",
    "# #nrow(pairs_total)\n",
    "\n",
    "# ########### Match for Day 6 hfov ####################\n",
    "# treat_day<-df2_treat[df2_treat$end_day==6 & df2_treat$HFV_indicator==1 & !(df2_treat$ID %in% pairs_total$matchedID_hfv),]\n",
    "# control_day<-df2_control[df2_control$end_day==6 & !(df2_control$ID %in% pairs_total$matchedID_non_hfv),]\n",
    "# #nrow(treat_day)\n",
    "# #nrow(control_day)\n",
    "# df_day6<-rbind(treat_day,control_day)\n",
    "# #nrow(df_day6)\n",
    "\n",
    "# m.out <- matchit(treatment~age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(rf_pneumonia)+factor(rf_sepsis)\n",
    "#                  +factor(rf_aspiration)\n",
    "#                  #+factor(rf_transfusion)\n",
    "#                  +factor(rf_drowning)+factor(MODS)+factor(OI_cat), data = df_day6,              \n",
    "#                  method = \"nearest\",\n",
    "#                  distance=\"linear.logit\",\n",
    "#                  discard=\"both\",\n",
    "#                  m.order=\"smallest\",\n",
    "#                  calclosest=TRUE)\n",
    "# #summary(m.out)\n",
    "# matchedID_hfv<-match.data(m.out, \"treat\")$ID\n",
    "# matchedID_non_hfv<-match.data(m.out, \"control\")$ID\n",
    "# pairs<-data.frame(cbind(matchedID_hfv,matchedID_non_hfv))\n",
    "# pairs_total<-rbind(pairs_total, pairs)\n",
    "\n",
    "# #write.csv(pairs, file = \"pair_ID.csv\")\n",
    "# nrow(pairs)\n",
    "# #nrow(pairs_total)\n",
    "\n",
    "# ########## Match for Day 7 hfov ####################\n",
    "# treat_day<-df2_treat[df2_treat$end_day==7 & df2_treat$HFV_indicator==1 & !(df2_treat$ID %in% pairs_total$matchedID_hfv),]\n",
    "# control_day<-df2_control[df2_control$end_day==7 & !(df2_control$ID %in% pairs_total$matchedID_non_hfv),]\n",
    "# #nrow(treat_day)\n",
    "# #nrow(control_day)\n",
    "# df_day7<-rbind(treat_day,control_day)\n",
    "# #nrow(df_day7)\n",
    "\n",
    "# m.out <- matchit(treatment~age+pim2+pelod+factor(female)+factor(p_bloodculture)+factor(comorbidity)+factor(rf_pneumonia)+factor(rf_sepsis)\n",
    "#                  +factor(rf_aspiration)\n",
    "#                  #+factor(rf_transfusion)\n",
    "#                  +factor(rf_drowning)+factor(MODS)+factor(OI_cat),  data = df_day7,              \n",
    "#                  method = \"nearest\",\n",
    "#                  distance=\"linear.logit\",\n",
    "#                  discard=\"both\",\n",
    "#                  m.order=\"smallest\",\n",
    "#                  calclosest=TRUE)\n",
    "# #summary(m.out)\n",
    "# matchedID_hfv<-match.data(m.out, \"treat\")$ID\n",
    "# matchedID_non_hfv<-match.data(m.out, \"control\")$ID\n",
    "# pairs<-data.frame(cbind(matchedID_hfv,matchedID_non_hfv))\n",
    "# pairs_total<-rbind(pairs_total, pairs)\n",
    "\n",
    "\n",
    "# nrow(pairs)\n",
    "# nrow(pairs_total)\n",
    "\n",
    "\n",
    "# ####################################################\n",
    "# #Alternative Method: PS matching on a daily basis\n",
    "# ####################################################\n",
    "\n",
    "# ########## Outcome analysis for daily pairs ####################\n",
    "\n",
    "# #daily_pair_ID <- read.csv(\"daily_pair_ID.csv\")\n",
    "# daily_pair_ID<- pairs_total\n",
    "# mydata$mort_100day<-ifelse(kkh$Died.within.100.days==\"Yes\",1,0)\n",
    "# sensitivity_data<-mydata[(mydata$ID %in% daily_pair_ID$matchedID_hfv)| (mydata$ID %in% daily_pair_ID$matchedID_non_hfv),]\n",
    "# head(sensitivity_data)\n",
    "# pair_number<-length(sensitivity_data[,1])\n",
    "\n",
    "# #pairwise difference for continuous data\n",
    "# hfv<-sensitivity_data[sensitivity_data$ID %in% daily_pair_ID$matchedID_hfv,]\n",
    "# non_hfv<-sensitivity_data[sensitivity_data$ID %in% daily_pair_ID$matchedID_non_hfv,]\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "# #paired t-test for ventilator free days\n",
    "# print(\"paired t-test for ventilator free days difference between HFOV less Non-HFOV\")\n",
    "# diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "# t.test(diffy)\n",
    "\n",
    "# #paired t-test for picu free days\n",
    "# print(\"paired t-test for picu free days difference between HFOV less Non-HFOV\")\n",
    "# diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "# t.test(diffy1)\n",
    "\n",
    "# #McNemar test for 100 day mortality\n",
    "# print(\"100 day mortality difference between HFOV less Non-HFOV\")\n",
    "# mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "# #calculate n% of 100 day mortality for matched cohort\n",
    "# print(\"Non-HFOV group 100day mortality n:\")\n",
    "# length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "# print(\"%:\")\n",
    "# length(non_hfv$ID[non_hfv$mort_100day==1])/pair_number\n",
    "# print(\"HFOV group 100day mortality n:\")\n",
    "# length(hfv$ID[hfv$mort_100day==1])\n",
    "# print(\"%:\")\n",
    "# length(hfv$ID[hfv$mort_100day==1])/pair_number\n",
    "\n",
    "# #McNemar test for 28-day mortality\n",
    "# y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "# y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "# table(y_treatment,y_control)\n",
    "# print(\"28 day mortality difference between HFOV less Non-HFOV\")\n",
    "# mcnemar.test(y_treatment, y_control)\n",
    "\n",
    "# #calculate odds ratio for 28-day mortality\n",
    "# y_treatment<-sensitivity_data$mortality_28day[sensitivity_data$treatment==1] \n",
    "# y_control<-sensitivity_data$mortality_28day[sensitivity_data$treatment==0] \n",
    "\n",
    "# ########## calculate Odd's ratio here!\n",
    "# print(\"Odds ratio\")\n",
    "# a<-length(y_treatment[y_treatment==1])\n",
    "# b<-length(y_control[y_control==1])\n",
    "# c<-length(y_treatment[y_treatment==0])\n",
    "# d<-length(y_control[y_control==0])\n",
    "# print(c(a,b,c,d))\n",
    "# a/(a+c)\n",
    "# b/(b+d)\n",
    "\n",
    "# oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>ID</th><th scope=col>hospital_number</th><th scope=col>female</th><th scope=col>age</th><th scope=col>pim2</th><th scope=col>pelod</th><th scope=col>p_bloodculture</th><th scope=col>comorbidity</th><th scope=col>rf_pneumonia</th><th scope=col>rf_sepsis</th><th scope=col>⋯</th><th scope=col>mortality_100day</th><th scope=col>ventilation_free_day</th><th scope=col>picu_free_days</th><th scope=col>fio2</th><th scope=col>map</th><th scope=col>amplitude</th><th scope=col>OI_cat</th><th scope=col>ICUduration</th><th scope=col>mort_100day</th><th scope=col>weight</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>3</th><td>1005    </td><td>1       </td><td>1       </td><td>0.10    </td><td> 0.6    </td><td> 1      </td><td>0       </td><td>0       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>21      </td><td>20      </td><td>   NA   </td><td>NA      </td><td>NA      </td><td>2       </td><td> 9      </td><td>0       </td><td>1.692353</td></tr>\n",
       "\t<tr><th scope=row>10</th><td>1014    </td><td>1       </td><td>1       </td><td>5.08    </td><td>18.2    </td><td> 3      </td><td>0       </td><td>0       </td><td>1       </td><td>1       </td><td>⋯       </td><td>1       </td><td> 0      </td><td> 0      </td><td>   NA   </td><td>NA      </td><td>NA      </td><td>1       </td><td>62      </td><td>1       </td><td>1.050604</td></tr>\n",
       "\t<tr><th scope=row>11</th><td>1015    </td><td>1       </td><td>1       </td><td>5.08    </td><td>18.2    </td><td> 3      </td><td>1       </td><td>0       </td><td>1       </td><td>1       </td><td>⋯       </td><td>1       </td><td> 0      </td><td> 0      </td><td>   NA   </td><td>NA      </td><td>NA      </td><td>4       </td><td>62      </td><td>1       </td><td>2.638227</td></tr>\n",
       "\t<tr><th scope=row>16</th><td>1020    </td><td>1       </td><td>0       </td><td>2.78    </td><td>41.8    </td><td>21      </td><td>0       </td><td>0       </td><td>0       </td><td>0       </td><td>⋯       </td><td>0       </td><td>24      </td><td>23      </td><td>   NA   </td><td>NA      </td><td>NA      </td><td>1       </td><td> 6      </td><td>0       </td><td>1.068845</td></tr>\n",
       "\t<tr><th scope=row>18</th><td>1022    </td><td>1       </td><td>1       </td><td>0.28    </td><td> 9.0    </td><td>20      </td><td>0       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>16      </td><td>15      </td><td>58.75   </td><td>28      </td><td>50      </td><td>2       </td><td>14      </td><td>0       </td><td>1.757998</td></tr>\n",
       "\t<tr><th scope=row>27</th><td>1037    </td><td>1       </td><td>1       </td><td>0.28    </td><td> 8.9    </td><td>20      </td><td>0       </td><td>1       </td><td>1       </td><td>0       </td><td>⋯       </td><td>0       </td><td>16      </td><td>15      </td><td>67.00   </td><td>28      </td><td>53      </td><td>2       </td><td>14      </td><td>0       </td><td>1.758396</td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lllllllllllllllllllllllllllllllllllllll}\n",
       "  & ID & hospital\\_number & female & age & pim2 & pelod & p\\_bloodculture & comorbidity & rf\\_pneumonia & rf\\_sepsis & ⋯ & mortality\\_100day & ventilation\\_free\\_day & picu\\_free\\_days & fio2 & map & amplitude & OI\\_cat & ICUduration & mort\\_100day & weight\\\\\n",
       "\\hline\n",
       "\t3 & 1005     & 1        & 1        & 0.10     &  0.6     &  1       & 0        & 0        & 1        & 0        & ⋯        & 0        & 21       & 20       &    NA    & NA       & NA       & 2        &  9       & 0        & 1.692353\\\\\n",
       "\t10 & 1014     & 1        & 1        & 5.08     & 18.2     &  3       & 0        & 0        & 1        & 1        & ⋯        & 1        &  0       &  0       &    NA    & NA       & NA       & 1        & 62       & 1        & 1.050604\\\\\n",
       "\t11 & 1015     & 1        & 1        & 5.08     & 18.2     &  3       & 1        & 0        & 1        & 1        & ⋯        & 1        &  0       &  0       &    NA    & NA       & NA       & 4        & 62       & 1        & 2.638227\\\\\n",
       "\t16 & 1020     & 1        & 0        & 2.78     & 41.8     & 21       & 0        & 0        & 0        & 0        & ⋯        & 0        & 24       & 23       &    NA    & NA       & NA       & 1        &  6       & 0        & 1.068845\\\\\n",
       "\t18 & 1022     & 1        & 1        & 0.28     &  9.0     & 20       & 0        & 1        & 1        & 0        & ⋯        & 0        & 16       & 15       & 58.75    & 28       & 50       & 2        & 14       & 0        & 1.757998\\\\\n",
       "\t27 & 1037     & 1        & 1        & 0.28     &  8.9     & 20       & 0        & 1        & 1        & 0        & ⋯        & 0        & 16       & 15       & 67.00    & 28       & 53       & 2        & 14       & 0        & 1.758396\\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | ID | hospital_number | female | age | pim2 | pelod | p_bloodculture | comorbidity | rf_pneumonia | rf_sepsis | ⋯ | mortality_100day | ventilation_free_day | picu_free_days | fio2 | map | amplitude | OI_cat | ICUduration | mort_100day | weight | \n",
       "|---|---|---|---|---|---|\n",
       "| 3 | 1005     | 1        | 1        | 0.10     |  0.6     |  1       | 0        | 0        | 1        | 0        | ⋯        | 0        | 21       | 20       |    NA    | NA       | NA       | 2        |  9       | 0        | 1.692353 | \n",
       "| 10 | 1014     | 1        | 1        | 5.08     | 18.2     |  3       | 0        | 0        | 1        | 1        | ⋯        | 1        |  0       |  0       |    NA    | NA       | NA       | 1        | 62       | 1        | 1.050604 | \n",
       "| 11 | 1015     | 1        | 1        | 5.08     | 18.2     |  3       | 1        | 0        | 1        | 1        | ⋯        | 1        |  0       |  0       |    NA    | NA       | NA       | 4        | 62       | 1        | 2.638227 | \n",
       "| 16 | 1020     | 1        | 0        | 2.78     | 41.8     | 21       | 0        | 0        | 0        | 0        | ⋯        | 0        | 24       | 23       |    NA    | NA       | NA       | 1        |  6       | 0        | 1.068845 | \n",
       "| 18 | 1022     | 1        | 1        | 0.28     |  9.0     | 20       | 0        | 1        | 1        | 0        | ⋯        | 0        | 16       | 15       | 58.75    | 28       | 50       | 2        | 14       | 0        | 1.757998 | \n",
       "| 27 | 1037     | 1        | 1        | 0.28     |  8.9     | 20       | 0        | 1        | 1        | 0        | ⋯        | 0        | 16       | 15       | 67.00    | 28       | 53       | 2        | 14       | 0        | 1.758396 | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "   ID   hospital_number female age  pim2 pelod p_bloodculture comorbidity\n",
       "3  1005 1               1      0.10  0.6  1    0              0          \n",
       "10 1014 1               1      5.08 18.2  3    0              0          \n",
       "11 1015 1               1      5.08 18.2  3    1              0          \n",
       "16 1020 1               0      2.78 41.8 21    0              0          \n",
       "18 1022 1               1      0.28  9.0 20    0              1          \n",
       "27 1037 1               1      0.28  8.9 20    0              1          \n",
       "   rf_pneumonia rf_sepsis ⋯ mortality_100day ventilation_free_day\n",
       "3  1            0         ⋯ 0                21                  \n",
       "10 1            1         ⋯ 1                 0                  \n",
       "11 1            1         ⋯ 1                 0                  \n",
       "16 0            0         ⋯ 0                24                  \n",
       "18 1            0         ⋯ 0                16                  \n",
       "27 1            0         ⋯ 0                16                  \n",
       "   picu_free_days fio2  map amplitude OI_cat ICUduration mort_100day weight  \n",
       "3  20                NA NA  NA        2       9          0           1.692353\n",
       "10  0                NA NA  NA        1      62          1           1.050604\n",
       "11  0                NA NA  NA        4      62          1           2.638227\n",
       "16 23                NA NA  NA        1       6          0           1.068845\n",
       "18 15             58.75 28  50        2      14          0           1.757998\n",
       "27 15             67.00 28  53        2      14          0           1.758396"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"paired t-test for ventilator free days difference between HFOV less Non-HFOV\"\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tOne Sample t-test\n",
       "\n",
       "data:  diffy\n",
       "t = -2.4547, df = 38, p-value = 0.01879\n",
       "alternative hypothesis: true mean is not equal to 0\n",
       "95 percent confidence interval:\n",
       " -9.5913158 -0.9215048\n",
       "sample estimates:\n",
       "mean of x \n",
       " -5.25641 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"paired t-test for picu free days difference between HFOV less Non-HFOV\"\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tOne Sample t-test\n",
       "\n",
       "data:  diffy1\n",
       "t = -1.9139, df = 38, p-value = 0.06318\n",
       "alternative hypothesis: true mean is not equal to 0\n",
       "95 percent confidence interval:\n",
       " -7.3866771  0.2071899\n",
       "sample estimates:\n",
       "mean of x \n",
       "-3.589744 \n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"100 day mortality difference between HFOV less Non-HFOV\"\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tMcNemar's Chi-squared test with continuity correction\n",
       "\n",
       "data:  hfv$mort_100day and non_hfv$mort_100day\n",
       "McNemar's chi-squared = 0.94118, df = 1, p-value = 0.332\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"Non-HFOV group 100day mortality n:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "9"
      ],
      "text/latex": [
       "9"
      ],
      "text/markdown": [
       "9"
      ],
      "text/plain": [
       "[1] 9"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"%:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.115384615384615"
      ],
      "text/latex": [
       "0.115384615384615"
      ],
      "text/markdown": [
       "0.115384615384615"
      ],
      "text/plain": [
       "[1] 0.1153846"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"HFOV group 100day mortality n:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "14"
      ],
      "text/latex": [
       "14"
      ],
      "text/markdown": [
       "14"
      ],
      "text/plain": [
       "[1] 14"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"%:\"\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.179487179487179"
      ],
      "text/latex": [
       "0.179487179487179"
      ],
      "text/markdown": [
       "0.179487179487179"
      ],
      "text/plain": [
       "[1] 0.1794872"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "           y_control\n",
       "y_treatment  0  1\n",
       "          0 25  5\n",
       "          1  7  2"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"28 day mortality difference between HFOV less Non-HFOV\"\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "\n",
       "\tMcNemar's Chi-squared test with continuity correction\n",
       "\n",
       "data:  y_treatment and y_control\n",
       "McNemar's chi-squared = 0.083333, df = 1, p-value = 0.7728\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1] \"Odds ratio\"\n",
      "[1]  9  7 30 32\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "0.230769230769231"
      ],
      "text/latex": [
       "0.230769230769231"
      ],
      "text/markdown": [
       "0.230769230769231"
      ],
      "text/plain": [
       "[1] 0.2307692"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "0.179487179487179"
      ],
      "text/latex": [
       "0.179487179487179"
      ],
      "text/markdown": [
       "0.179487179487179"
      ],
      "text/plain": [
       "[1] 0.1794872"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<dl>\n",
       "\t<dt>$data</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>Disease1</th><th scope=col>Disease2</th><th scope=col>Total</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td> 9</td><td>30</td><td>39</td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td> 7</td><td>32</td><td>39</td></tr>\n",
       "\t<tr><th scope=row>Total</th><td>16</td><td>62</td><td>78</td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$measure</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>estimate</th><th scope=col>lower</th><th scope=col>upper</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>1.000000 </td><td>       NA</td><td>      NA </td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td>1.360465 </td><td>0.4434294</td><td>4.331214 </td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$p.value</dt>\n",
       "\t\t<dd><table>\n",
       "<thead><tr><th></th><th scope=col>midp.exact</th><th scope=col>fisher.exact</th><th scope=col>chi.square</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>Exposed1</th><td>       NA</td><td>       NA</td><td>      NA </td></tr>\n",
       "\t<tr><th scope=row>Exposed2</th><td>0.5903969</td><td>0.7799148</td><td>0.574922 </td></tr>\n",
       "</tbody>\n",
       "</table>\n",
       "</dd>\n",
       "\t<dt>$correction</dt>\n",
       "\t\t<dd>FALSE</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description}\n",
       "\\item[\\$data] \\begin{tabular}{r|lll}\n",
       "  & Disease1 & Disease2 & Total\\\\\n",
       "\\hline\n",
       "\tExposed1 &  9 & 30 & 39\\\\\n",
       "\tExposed2 &  7 & 32 & 39\\\\\n",
       "\tTotal & 16 & 62 & 78\\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$measure] \\begin{tabular}{r|lll}\n",
       "  & estimate & lower & upper\\\\\n",
       "\\hline\n",
       "\tExposed1 & 1.000000  &        NA &       NA \\\\\n",
       "\tExposed2 & 1.360465  & 0.4434294 & 4.331214 \\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$p.value] \\begin{tabular}{r|lll}\n",
       "  & midp.exact & fisher.exact & chi.square\\\\\n",
       "\\hline\n",
       "\tExposed1 &        NA &        NA &       NA \\\\\n",
       "\tExposed2 & 0.5903969 & 0.7799148 & 0.574922 \\\\\n",
       "\\end{tabular}\n",
       "\n",
       "\\item[\\$correction] FALSE\n",
       "\\end{description}\n"
      ],
      "text/markdown": [
       "$data\n",
       ":   \n",
       "| <!--/--> | Disease1 | Disease2 | Total | \n",
       "|---|---|---|\n",
       "| Exposed1 |  9 | 30 | 39 | \n",
       "| Exposed2 |  7 | 32 | 39 | \n",
       "| Total | 16 | 62 | 78 | \n",
       "\n",
       "\n",
       "\n",
       "$measure\n",
       ":   \n",
       "| <!--/--> | estimate | lower | upper | \n",
       "|---|---|\n",
       "| Exposed1 | 1.000000  |        NA |       NA  | \n",
       "| Exposed2 | 1.360465  | 0.4434294 | 4.331214  | \n",
       "\n",
       "\n",
       "\n",
       "$p.value\n",
       ":   \n",
       "| <!--/--> | midp.exact | fisher.exact | chi.square | \n",
       "|---|---|\n",
       "| Exposed1 |        NA |        NA |       NA  | \n",
       "| Exposed2 | 0.5903969 | 0.7799148 | 0.574922  | \n",
       "\n",
       "\n",
       "\n",
       "$correction\n",
       ":   FALSE\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "$data\n",
       "          Outcome\n",
       "Predictor  Disease1 Disease2 Total\n",
       "  Exposed1        9       30    39\n",
       "  Exposed2        7       32    39\n",
       "  Total          16       62    78\n",
       "\n",
       "$measure\n",
       "          odds ratio with 95% C.I.\n",
       "Predictor  estimate     lower    upper\n",
       "  Exposed1 1.000000        NA       NA\n",
       "  Exposed2 1.360465 0.4434294 4.331214\n",
       "\n",
       "$p.value\n",
       "          two-sided\n",
       "Predictor  midp.exact fisher.exact chi.square\n",
       "  Exposed1         NA           NA         NA\n",
       "  Exposed2  0.5903969    0.7799148   0.574922\n",
       "\n",
       "$correction\n",
       "[1] FALSE\n",
       "\n",
       "attr(,\"method\")\n",
       "[1] \"median-unbiased estimate & mid-p exact CI\""
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "####################################################\n",
    "#Alternative Method: PS matching on a daily basis\n",
    "####################################################\n",
    "\n",
    "########## Outcome analysis for daily pairs ####################\n",
    "\n",
    "daily_pair_ID <- read.csv(\"daily_pair_ID.csv\")\n",
    "\n",
    "mydata$mort_100day<-ifelse(kkh$Died.within.100.days==\"Yes\",1,0)\n",
    "sensitivity_data<-mydata[(mydata$ID %in% daily_pair_ID$matchedID_hfv)| (mydata$ID %in% daily_pair_ID$matchedID_non_hfv),]\n",
    "head(sensitivity_data)\n",
    "pair_number<-length(sensitivity_data[,1])\n",
    "\n",
    "#pairwise difference for continuous data\n",
    "hfv<-sensitivity_data[sensitivity_data$ID %in% daily_pair_ID$matchedID_hfv,]\n",
    "non_hfv<-sensitivity_data[sensitivity_data$ID %in% daily_pair_ID$matchedID_non_hfv,]\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "#paired t-test for ventilator free days\n",
    "print(\"paired t-test for ventilator free days difference between HFOV less Non-HFOV\")\n",
    "diffy<-hfv$ventilation_free_day-non_hfv$ventilation_free_day\n",
    "t.test(diffy)\n",
    "\n",
    "#paired t-test for picu free days\n",
    "print(\"paired t-test for picu free days difference between HFOV less Non-HFOV\")\n",
    "diffy1<-hfv$picu_free_days-non_hfv$picu_free_days\n",
    "t.test(diffy1)\n",
    "\n",
    "#McNemar test for 100 day mortality\n",
    "print(\"100 day mortality difference between HFOV less Non-HFOV\")\n",
    "mcnemar.test(hfv$mort_100day, non_hfv$mort_100day)\n",
    "\n",
    "#calculate n% of 100 day mortality for matched cohort\n",
    "print(\"Non-HFOV group 100day mortality n:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(non_hfv$ID[non_hfv$mort_100day==1])/pair_number\n",
    "print(\"HFOV group 100day mortality n:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])\n",
    "print(\"%:\")\n",
    "length(hfv$ID[hfv$mort_100day==1])/pair_number\n",
    "\n",
    "#McNemar test for 28-day mortality\n",
    "y_treatment<-hfv$mortality_28day[hfv$treatment==1] \n",
    "y_control<-non_hfv$mortality_28day[non_hfv$treatment==0] \n",
    "table(y_treatment,y_control)\n",
    "print(\"28 day mortality difference between HFOV less Non-HFOV\")\n",
    "mcnemar.test(y_treatment, y_control)\n",
    "\n",
    "#calculate odds ratio for 28-day mortality\n",
    "y_treatment<-sensitivity_data$mortality_28day[sensitivity_data$treatment==1] \n",
    "y_control<-sensitivity_data$mortality_28day[sensitivity_data$treatment==0] \n",
    "\n",
    "########## calculate Odd's ratio here!\n",
    "print(\"Odds ratio\")\n",
    "a<-length(y_treatment[y_treatment==1])\n",
    "b<-length(y_control[y_control==1])\n",
    "c<-length(y_treatment[y_treatment==0])\n",
    "d<-length(y_control[y_control==0])\n",
    "print(c(a,b,c,d))\n",
    "a/(a+c)\n",
    "b/(b+d)\n",
    "\n",
    "oddsratio(matrix(c(a,b,c,d), ncol=2, nrow=2))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## multi-center effects"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Call:\n",
       "coxph(formula = Surv(mortality_28day) ~ treatment + treatment:factor(hospital_number) + \n",
       "    female + age + pim2 + pelod + p_bloodculture + comorbidity + \n",
       "    rf_pneumonia + rf_sepsis + rf_aspiration + rf_transfusion + \n",
       "    rf_drowning + MODS + factor(OI_cat) + strata(hospital_number), \n",
       "    data = mydata)\n",
       "\n",
       "  n= 328, number of events= 328 \n",
       "\n",
       "                                         coef exp(coef)  se(coef)      z\n",
       "treatment                           -0.303536  0.738204  0.270010 -1.124\n",
       "female                              -0.082235  0.921056  0.126627 -0.649\n",
       "age                                 -0.001981  0.998021  0.013980 -0.142\n",
       "pim2                                -0.002125  0.997877  0.003991 -0.533\n",
       "pelod                               -0.009811  0.990237  0.007837 -1.252\n",
       "p_bloodculture                      -0.277258  0.757859  0.187753 -1.477\n",
       "comorbidity                         -0.135601  0.873191  0.145430 -0.932\n",
       "rf_pneumonia                         0.026487  1.026841  0.206589  0.128\n",
       "rf_sepsis                           -0.122007  0.885142  0.178647 -0.683\n",
       "rf_aspiration                       -0.018230  0.981935  0.306461 -0.059\n",
       "rf_transfusion                       0.039397  1.040184  0.480209  0.082\n",
       "rf_drowning                          0.066122  1.068357  0.387719  0.171\n",
       "MODS                                -0.437731  0.645499  0.145418 -3.010\n",
       "factor(OI_cat)2                      0.266494  1.305380  0.171152  1.557\n",
       "factor(OI_cat)3                      0.062374  1.064360  0.235946  0.264\n",
       "factor(OI_cat)4                      0.285292  1.330150  0.181521  1.572\n",
       "treatment:factor(hospital_number)2   0.294629  1.342628  0.501601  0.587\n",
       "treatment:factor(hospital_number)3   0.831359  2.296438  0.870318  0.955\n",
       "treatment:factor(hospital_number)4   0.645468  1.906880  0.862845  0.748\n",
       "treatment:factor(hospital_number)6  -0.168285  0.845113  0.384002 -0.438\n",
       "treatment:factor(hospital_number)7  -0.814926  0.442672  0.629489 -1.295\n",
       "treatment:factor(hospital_number)8   0.047546  1.048694  0.540373  0.088\n",
       "treatment:factor(hospital_number)9  -0.124314  0.883102  0.446177 -0.279\n",
       "treatment:factor(hospital_number)10 -1.412978  0.243417  1.197239 -1.180\n",
       "treatment:factor(hospital_number)11 -0.102708  0.902390  0.583940 -0.176\n",
       "                                    Pr(>|z|)   \n",
       "treatment                            0.26094   \n",
       "female                               0.51606   \n",
       "age                                  0.88730   \n",
       "pim2                                 0.59437   \n",
       "pelod                                0.21061   \n",
       "p_bloodculture                       0.13975   \n",
       "comorbidity                          0.35112   \n",
       "rf_pneumonia                         0.89798   \n",
       "rf_sepsis                            0.49464   \n",
       "rf_aspiration                        0.95257   \n",
       "rf_transfusion                       0.93461   \n",
       "rf_drowning                          0.86458   \n",
       "MODS                                 0.00261 **\n",
       "factor(OI_cat)2                      0.11946   \n",
       "factor(OI_cat)3                      0.79151   \n",
       "factor(OI_cat)4                      0.11603   \n",
       "treatment:factor(hospital_number)2   0.55695   \n",
       "treatment:factor(hospital_number)3   0.33946   \n",
       "treatment:factor(hospital_number)4   0.45442   \n",
       "treatment:factor(hospital_number)6   0.66121   \n",
       "treatment:factor(hospital_number)7   0.19546   \n",
       "treatment:factor(hospital_number)8   0.92989   \n",
       "treatment:factor(hospital_number)9   0.78054   \n",
       "treatment:factor(hospital_number)10  0.23792   \n",
       "treatment:factor(hospital_number)11  0.86038   \n",
       "---\n",
       "Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1\n",
       "\n",
       "                                    exp(coef) exp(-coef) lower .95 upper .95\n",
       "treatment                              0.7382     1.3546   0.43485    1.2532\n",
       "female                                 0.9211     1.0857   0.71862    1.1805\n",
       "age                                    0.9980     1.0020   0.97105    1.0257\n",
       "pim2                                   0.9979     1.0021   0.99010    1.0057\n",
       "pelod                                  0.9902     1.0099   0.97514    1.0056\n",
       "p_bloodculture                         0.7579     1.3195   0.52453    1.0950\n",
       "comorbidity                            0.8732     1.1452   0.65663    1.1612\n",
       "rf_pneumonia                           1.0268     0.9739   0.68494    1.5394\n",
       "rf_sepsis                              0.8851     1.1298   0.62366    1.2563\n",
       "rf_aspiration                          0.9819     1.0184   0.53855    1.7904\n",
       "rf_transfusion                         1.0402     0.9614   0.40584    2.6660\n",
       "rf_drowning                            1.0684     0.9360   0.49968    2.2843\n",
       "MODS                                   0.6455     1.5492   0.48542    0.8584\n",
       "factor(OI_cat)2                        1.3054     0.7661   0.93337    1.8257\n",
       "factor(OI_cat)3                        1.0644     0.9395   0.67027    1.6902\n",
       "factor(OI_cat)4                        1.3301     0.7518   0.93194    1.8985\n",
       "treatment:factor(hospital_number)2     1.3426     0.7448   0.50233    3.5886\n",
       "treatment:factor(hospital_number)3     2.2964     0.4355   0.41710   12.6436\n",
       "treatment:factor(hospital_number)4     1.9069     0.5244   0.35145   10.3461\n",
       "treatment:factor(hospital_number)6     0.8451     1.1833   0.39815    1.7938\n",
       "treatment:factor(hospital_number)7     0.4427     2.2590   0.12890    1.5202\n",
       "treatment:factor(hospital_number)8     1.0487     0.9536   0.36365    3.0242\n",
       "treatment:factor(hospital_number)9     0.8831     1.1324   0.36832    2.1174\n",
       "treatment:factor(hospital_number)10    0.2434     4.1082   0.02329    2.5436\n",
       "treatment:factor(hospital_number)11    0.9024     1.1082   0.28731    2.8343\n",
       "\n",
       "Concordance= 0.834  (se = 0.18 )\n",
       "Rsquare= 0.121   (max possible= 0.997 )\n",
       "Likelihood ratio test= 42.45  on 25 df,   p=0.02\n",
       "Wald test            = 39.91  on 25 df,   p=0.03\n",
       "Score (logrank) test = 41.43  on 25 df,   p=0.02\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "                                          rho    chisq     p\n",
       "treatment                           -0.006496 1.55e-02 0.901\n",
       "female                               0.029285 3.29e-01 0.566\n",
       "age                                 -0.012622 5.34e-02 0.817\n",
       "pim2                                -0.000180 1.15e-05 0.997\n",
       "pelod                                0.055480 1.05e+00 0.307\n",
       "p_bloodculture                       0.034915 4.80e-01 0.488\n",
       "comorbidity                          0.018367 1.25e-01 0.724\n",
       "rf_pneumonia                         0.001628 9.48e-04 0.975\n",
       "rf_sepsis                           -0.003312 4.30e-03 0.948\n",
       "rf_aspiration                       -0.009059 2.86e-02 0.866\n",
       "rf_transfusion                       0.005168 9.02e-03 0.924\n",
       "rf_drowning                         -0.009832 3.45e-02 0.853\n",
       "MODS                                 0.055718 1.03e+00 0.311\n",
       "factor(OI_cat)2                     -0.036873 4.70e-01 0.493\n",
       "factor(OI_cat)3                     -0.002379 2.00e-03 0.964\n",
       "factor(OI_cat)4                     -0.048797 8.46e-01 0.358\n",
       "treatment:factor(hospital_number)2  -0.008738 2.71e-02 0.869\n",
       "treatment:factor(hospital_number)3  -0.057971 1.04e+00 0.308\n",
       "treatment:factor(hospital_number)4   0.000616 1.30e-04 0.991\n",
       "treatment:factor(hospital_number)6   0.044222 7.25e-01 0.395\n",
       "treatment:factor(hospital_number)7   0.034064 3.94e-01 0.530\n",
       "treatment:factor(hospital_number)8   0.022835 1.79e-01 0.673\n",
       "treatment:factor(hospital_number)9   0.023583 2.01e-01 0.654\n",
       "treatment:factor(hospital_number)10  0.000426 6.51e-05 0.994\n",
       "treatment:factor(hospital_number)11  0.009431 2.96e-02 0.863\n",
       "GLOBAL                                     NA 8.99e+00 0.999"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "################################################################################\n",
    "### Stratified Cox model with interaction terms to study the multi-center effects\n",
    "################################################################################\n",
    "Strata_cox_model<-coxph(Surv(mortality_28day)~treatment+treatment:factor(hospital_number)+female+age+pim2+pelod+p_bloodculture+comorbidity+rf_pneumonia+rf_sepsis\n",
    "                 +rf_aspiration+rf_transfusion+rf_drowning+MODS+factor(OI_cat)+strata(hospital_number), data=mydata)\n",
    "summary(Strata_cox_model)\n",
    "cox.zph(Strata_cox_model, global=TRUE)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Calculate HOFV and CMV setting for the original cohort"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>fio2</th><th scope=col>map</th><th scope=col>amplitude</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>4</th><td>100.00  </td><td>23.0    </td><td>70.00000</td></tr>\n",
       "\t<tr><th scope=row>6</th><td>100.00  </td><td>32.0    </td><td>43.66667</td></tr>\n",
       "\t<tr><th scope=row>8</th><td>100.00  </td><td>30.0    </td><td>75.00000</td></tr>\n",
       "\t<tr><th scope=row>15</th><td> 71.00  </td><td>30.0    </td><td>55.80000</td></tr>\n",
       "\t<tr><th scope=row>18</th><td> 58.75  </td><td>28.0    </td><td>50.00000</td></tr>\n",
       "\t<tr><th scope=row>20</th><td>100.00  </td><td>25.5    </td><td>62.00000</td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|lll}\n",
       "  & fio2 & map & amplitude\\\\\n",
       "\\hline\n",
       "\t4 & 100.00   & 23.0     & 70.00000\\\\\n",
       "\t6 & 100.00   & 32.0     & 43.66667\\\\\n",
       "\t8 & 100.00   & 30.0     & 75.00000\\\\\n",
       "\t15 &  71.00   & 30.0     & 55.80000\\\\\n",
       "\t18 &  58.75   & 28.0     & 50.00000\\\\\n",
       "\t20 & 100.00   & 25.5     & 62.00000\\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | fio2 | map | amplitude | \n",
       "|---|---|---|---|---|---|\n",
       "| 4 | 100.00   | 23.0     | 70.00000 | \n",
       "| 6 | 100.00   | 32.0     | 43.66667 | \n",
       "| 8 | 100.00   | 30.0     | 75.00000 | \n",
       "| 15 |  71.00   | 30.0     | 55.80000 | \n",
       "| 18 |  58.75   | 28.0     | 50.00000 | \n",
       "| 20 | 100.00   | 25.5     | 62.00000 | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "   fio2   map  amplitude\n",
       "4  100.00 23.0 70.00000 \n",
       "6  100.00 32.0 43.66667 \n",
       "8  100.00 30.0 75.00000 \n",
       "15  71.00 30.0 55.80000 \n",
       "18  58.75 28.0 50.00000 \n",
       "20 100.00 25.5 62.00000 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<dl class=dl-horizontal>\n",
       "\t<dt>map</dt>\n",
       "\t\t<dd>24.9224041206195</dd>\n",
       "\t<dt>amplitude</dt>\n",
       "\t\t<dd>55.2147529637383</dd>\n",
       "\t<dt>fio2</dt>\n",
       "\t\t<dd>82.9314535204386</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description*}\n",
       "\\item[map] 24.9224041206195\n",
       "\\item[amplitude] 55.2147529637383\n",
       "\\item[fio2] 82.9314535204386\n",
       "\\end{description*}\n"
      ],
      "text/markdown": [
       "map\n",
       ":   24.9224041206195amplitude\n",
       ":   55.2147529637383fio2\n",
       ":   82.9314535204386\n",
       "\n"
      ],
      "text/plain": [
       "      map amplitude      fio2 \n",
       " 24.92240  55.21475  82.93145 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "5.65389301806335"
      ],
      "text/latex": [
       "5.65389301806335"
      ],
      "text/markdown": [
       "5.65389301806335"
      ],
      "text/plain": [
       "[1] 5.653893"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "13.8445500287911"
      ],
      "text/latex": [
       "13.8445500287911"
      ],
      "text/markdown": [
       "13.8445500287911"
      ],
      "text/plain": [
       "[1] 13.84455"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "17.8983817294144"
      ],
      "text/latex": [
       "17.8983817294144"
      ],
      "text/markdown": [
       "17.8983817294144"
      ],
      "text/plain": [
       "[1] 17.89838"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hfov_setting<-data.frame(mydata[mydata$treatment==1,c(\"fio2\",\"map\",\"amplitude\")])\n",
    "\n",
    "head(hfov_setting)\n",
    "colMeans(hfov_setting[,c(\"map\",\"amplitude\",\"fio2\")], na.rm=TRUE)\n",
    "\n",
    "map<-subset(hfov_setting$map, (!is.na(hfov_setting[,\"map\"])))\n",
    "amplitude<-subset(hfov_setting$amplitude, (!is.na(hfov_setting[,\"amplitude\"])))\n",
    "fio2<-subset(hfov_setting$fio2, (!is.na(hfov_setting[,\"fio2\"])))\n",
    "\n",
    "sd(map) \n",
    "sd(amplitude) \n",
    "sd(fio2) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th scope=col>StudyID</th><th scope=col>D1_HFV</th><th scope=col>D2_HFV</th><th scope=col>D3_HFV</th><th scope=col>D4_HFV</th><th scope=col>D5_HFV</th><th scope=col>D6_HFV</th><th scope=col>D7_HFV</th><th scope=col>HFOV_days</th><th scope=col>exclude</th><th scope=col>⋯</th><th scope=col>d3_amplitude</th><th scope=col>d4_amplitude</th><th scope=col>d5_amplitude</th><th scope=col>d6_amplitude</th><th scope=col>d7_amplitude</th><th scope=col>cmv_tidal</th><th scope=col>cmv_fio2</th><th scope=col>cmv_pip</th><th scope=col>cmv_peep</th><th scope=col>cmv_map</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><td>1001     </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>include  </td><td>⋯        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td> 19.00000</td><td>65.00000 </td><td>25.00000 </td><td>9.000000 </td><td>14.50000 </td></tr>\n",
       "\t<tr><td>1003     </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>include  </td><td>⋯        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td> 83.50000</td><td>28.57143 </td><td>15.50000 </td><td>3.714286 </td><td>13.06667 </td></tr>\n",
       "\t<tr><td>1005     </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>include  </td><td>⋯        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td> 44.71429</td><td>33.57143 </td><td>22.28572 </td><td>1.000000 </td><td>18.00000 </td></tr>\n",
       "\t<tr><td>1008     </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>include  </td><td>⋯        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>231.50000</td><td>34.00000 </td><td>15.25000 </td><td>6.000000 </td><td>20.00000 </td></tr>\n",
       "\t<tr><td>1010     </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>include  </td><td>⋯        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td> 26.75000</td><td>42.80000 </td><td>17.50000 </td><td>7.200000 </td><td>12.00000 </td></tr>\n",
       "\t<tr><td>1012     </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>include  </td><td>⋯        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td>0        </td><td> 51.20000</td><td>31.83333 </td><td>19.00000 </td><td>6.833333 </td><td>10.48000 </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll}\n",
       " StudyID & D1\\_HFV & D2\\_HFV & D3\\_HFV & D4\\_HFV & D5\\_HFV & D6\\_HFV & D7\\_HFV & HFOV\\_days & exclude & ⋯ & d3\\_amplitude & d4\\_amplitude & d5\\_amplitude & d6\\_amplitude & d7\\_amplitude & cmv\\_tidal & cmv\\_fio2 & cmv\\_pip & cmv\\_peep & cmv\\_map\\\\\n",
       "\\hline\n",
       "\t 1001      & 0         & 0         & 0         & 0         & 0         & 0         & 0         & 0         & include   & ⋯         & 0         & 0         & 0         & 0         & 0         &  19.00000 & 65.00000  & 25.00000  & 9.000000  & 14.50000 \\\\\n",
       "\t 1003      & 0         & 0         & 0         & 0         & 0         & 0         & 0         & 0         & include   & ⋯         & 0         & 0         & 0         & 0         & 0         &  83.50000 & 28.57143  & 15.50000  & 3.714286  & 13.06667 \\\\\n",
       "\t 1005      & 0         & 0         & 0         & 0         & 0         & 0         & 0         & 0         & include   & ⋯         & 0         & 0         & 0         & 0         & 0         &  44.71429 & 33.57143  & 22.28572  & 1.000000  & 18.00000 \\\\\n",
       "\t 1008      & 0         & 0         & 0         & 0         & 0         & 0         & 0         & 0         & include   & ⋯         & 0         & 0         & 0         & 0         & 0         & 231.50000 & 34.00000  & 15.25000  & 6.000000  & 20.00000 \\\\\n",
       "\t 1010      & 0         & 0         & 0         & 0         & 0         & 0         & 0         & 0         & include   & ⋯         & 0         & 0         & 0         & 0         & 0         &  26.75000 & 42.80000  & 17.50000  & 7.200000  & 12.00000 \\\\\n",
       "\t 1012      & 0         & 0         & 0         & 0         & 0         & 0         & 0         & 0         & include   & ⋯         & 0         & 0         & 0         & 0         & 0         &  51.20000 & 31.83333  & 19.00000  & 6.833333  & 10.48000 \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "StudyID | D1_HFV | D2_HFV | D3_HFV | D4_HFV | D5_HFV | D6_HFV | D7_HFV | HFOV_days | exclude | ⋯ | d3_amplitude | d4_amplitude | d5_amplitude | d6_amplitude | d7_amplitude | cmv_tidal | cmv_fio2 | cmv_pip | cmv_peep | cmv_map | \n",
       "|---|---|---|---|---|---|\n",
       "| 1001      | 0         | 0         | 0         | 0         | 0         | 0         | 0         | 0         | include   | ⋯         | 0         | 0         | 0         | 0         | 0         |  19.00000 | 65.00000  | 25.00000  | 9.000000  | 14.50000  | \n",
       "| 1003      | 0         | 0         | 0         | 0         | 0         | 0         | 0         | 0         | include   | ⋯         | 0         | 0         | 0         | 0         | 0         |  83.50000 | 28.57143  | 15.50000  | 3.714286  | 13.06667  | \n",
       "| 1005      | 0         | 0         | 0         | 0         | 0         | 0         | 0         | 0         | include   | ⋯         | 0         | 0         | 0         | 0         | 0         |  44.71429 | 33.57143  | 22.28572  | 1.000000  | 18.00000  | \n",
       "| 1008      | 0         | 0         | 0         | 0         | 0         | 0         | 0         | 0         | include   | ⋯         | 0         | 0         | 0         | 0         | 0         | 231.50000 | 34.00000  | 15.25000  | 6.000000  | 20.00000  | \n",
       "| 1010      | 0         | 0         | 0         | 0         | 0         | 0         | 0         | 0         | include   | ⋯         | 0         | 0         | 0         | 0         | 0         |  26.75000 | 42.80000  | 17.50000  | 7.200000  | 12.00000  | \n",
       "| 1012      | 0         | 0         | 0         | 0         | 0         | 0         | 0         | 0         | include   | ⋯         | 0         | 0         | 0         | 0         | 0         |  51.20000 | 31.83333  | 19.00000  | 6.833333  | 10.48000  | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  StudyID D1_HFV D2_HFV D3_HFV D4_HFV D5_HFV D6_HFV D7_HFV HFOV_days exclude ⋯\n",
       "1 1001    0      0      0      0      0      0      0      0         include ⋯\n",
       "2 1003    0      0      0      0      0      0      0      0         include ⋯\n",
       "3 1005    0      0      0      0      0      0      0      0         include ⋯\n",
       "4 1008    0      0      0      0      0      0      0      0         include ⋯\n",
       "5 1010    0      0      0      0      0      0      0      0         include ⋯\n",
       "6 1012    0      0      0      0      0      0      0      0         include ⋯\n",
       "  d3_amplitude d4_amplitude d5_amplitude d6_amplitude d7_amplitude cmv_tidal\n",
       "1 0            0            0            0            0             19.00000\n",
       "2 0            0            0            0            0             83.50000\n",
       "3 0            0            0            0            0             44.71429\n",
       "4 0            0            0            0            0            231.50000\n",
       "5 0            0            0            0            0             26.75000\n",
       "6 0            0            0            0            0             51.20000\n",
       "  cmv_fio2 cmv_pip  cmv_peep cmv_map \n",
       "1 65.00000 25.00000 9.000000 14.50000\n",
       "2 28.57143 15.50000 3.714286 13.06667\n",
       "3 33.57143 22.28572 1.000000 18.00000\n",
       "4 34.00000 15.25000 6.000000 20.00000\n",
       "5 42.80000 17.50000 7.200000 12.00000\n",
       "6 31.83333 19.00000 6.833333 10.48000"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cmv_data <- read.csv(\"cmv_data_1125.csv\")\n",
    "head(cmv_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<ol class=list-inline>\n",
       "\t<li>Pressure controlled ventilation</li>\n",
       "\t<li>Airway pressure release ventilation</li>\n",
       "\t<li>Volume controlled ventilation</li>\n",
       "</ol>\n",
       "\n",
       "<details>\n",
       "\t<summary style=display:list-item;cursor:pointer>\n",
       "\t\t<strong>Levels</strong>:\n",
       "\t</summary>\n",
       "\t<ol class=list-inline>\n",
       "\t\t<li>'Airway pressure release ventilation'</li>\n",
       "\t\t<li>'Pressure controlled ventilation'</li>\n",
       "\t\t<li>'Volume controlled ventilation'</li>\n",
       "\t</ol>\n",
       "</details>"
      ],
      "text/latex": [
       "\\begin{enumerate*}\n",
       "\\item Pressure controlled ventilation\n",
       "\\item Airway pressure release ventilation\n",
       "\\item Volume controlled ventilation\n",
       "\\end{enumerate*}\n",
       "\n",
       "\\emph{Levels}: \\begin{enumerate*}\n",
       "\\item 'Airway pressure release ventilation'\n",
       "\\item 'Pressure controlled ventilation'\n",
       "\\item 'Volume controlled ventilation'\n",
       "\\end{enumerate*}\n"
      ],
      "text/markdown": [
       "1. Pressure controlled ventilation\n",
       "2. Airway pressure release ventilation\n",
       "3. Volume controlled ventilation\n",
       "\n",
       "\n",
       "\n",
       "**Levels**: 1. 'Airway pressure release ventilation'\n",
       "2. 'Pressure controlled ventilation'\n",
       "3. 'Volume controlled ventilation'\n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "[1] Pressure controlled ventilation     Airway pressure release ventilation\n",
       "[3] Volume controlled ventilation      \n",
       "3 Levels: Airway pressure release ventilation ..."
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#colnames(cmv_data)\n",
    "cmv<-cmv_data[,c(\"cmv_tidal\",\"cmv_fio2\",\"cmv_pip\",\"cmv_peep\",\"cmv_map\",\"D2_ventilation_mode\",\"Weight_kg\")]\n",
    "unique(cmv$D2_ventilation_mode)\n",
    "cmv_setting<-cmv[cmv$D2_ventilation_mode==\"Pressure controlled ventilation\" |cmv$D2_ventilation_mode==\"Volume controlled ventilation\"  ,]\n",
    "aprv_setting<-cmv[cmv$D2_ventilation_mode==\"Airway pressure release ventilation\" ,]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>cmv_tidal</th><th scope=col>cmv_fio2</th><th scope=col>cmv_pip</th><th scope=col>cmv_peep</th><th scope=col>cmv_map</th><th scope=col>D2_ventilation_mode</th><th scope=col>Weight_kg</th><th scope=col>cmv_tidal_kg</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>2</th><td> 83.50000                          </td><td>28.57143                           </td><td>15.50000                           </td><td>3.714286                           </td><td>13.06667                           </td><td>Airway pressure release ventilation</td><td> 8.80                              </td><td> 9.488636                          </td></tr>\n",
       "\t<tr><th scope=row>3</th><td> 44.71429                          </td><td>33.57143                           </td><td>22.28572                           </td><td>1.000000                           </td><td>18.00000                           </td><td>Airway pressure release ventilation</td><td> 4.69                              </td><td> 9.533963                          </td></tr>\n",
       "\t<tr><th scope=row>4</th><td>231.50000                          </td><td>34.00000                           </td><td>15.25000                           </td><td>6.000000                           </td><td>20.00000                           </td><td>Airway pressure release ventilation</td><td>42.00                              </td><td> 5.511905                          </td></tr>\n",
       "\t<tr><th scope=row>7</th><td>257.71429                          </td><td>40.14286                           </td><td>20.42857                           </td><td>1.428571                           </td><td>18.28572                           </td><td>Airway pressure release ventilation</td><td>19.00                              </td><td>13.563910                          </td></tr>\n",
       "\t<tr><th scope=row>8</th><td>258.71429                          </td><td>58.57143                           </td><td>30.00000                           </td><td>0.000000                           </td><td>27.42857                           </td><td>Airway pressure release ventilation</td><td>19.00                              </td><td>13.616542                          </td></tr>\n",
       "\t<tr><th scope=row>9</th><td>225.00000                          </td><td>70.00000                           </td><td>29.33333                           </td><td>2.000000                           </td><td>25.16667                           </td><td>Airway pressure release ventilation</td><td>19.00                              </td><td>11.842105                          </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllllll}\n",
       "  & cmv\\_tidal & cmv\\_fio2 & cmv\\_pip & cmv\\_peep & cmv\\_map & D2\\_ventilation\\_mode & Weight\\_kg & cmv\\_tidal\\_kg\\\\\n",
       "\\hline\n",
       "\t2 &  83.50000                           & 28.57143                            & 15.50000                            & 3.714286                            & 13.06667                            & Airway pressure release ventilation &  8.80                               &  9.488636                          \\\\\n",
       "\t3 &  44.71429                           & 33.57143                            & 22.28572                            & 1.000000                            & 18.00000                            & Airway pressure release ventilation &  4.69                               &  9.533963                          \\\\\n",
       "\t4 & 231.50000                           & 34.00000                            & 15.25000                            & 6.000000                            & 20.00000                            & Airway pressure release ventilation & 42.00                               &  5.511905                          \\\\\n",
       "\t7 & 257.71429                           & 40.14286                            & 20.42857                            & 1.428571                            & 18.28572                            & Airway pressure release ventilation & 19.00                               & 13.563910                          \\\\\n",
       "\t8 & 258.71429                           & 58.57143                            & 30.00000                            & 0.000000                            & 27.42857                            & Airway pressure release ventilation & 19.00                               & 13.616542                          \\\\\n",
       "\t9 & 225.00000                           & 70.00000                            & 29.33333                            & 2.000000                            & 25.16667                            & Airway pressure release ventilation & 19.00                               & 11.842105                          \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | cmv_tidal | cmv_fio2 | cmv_pip | cmv_peep | cmv_map | D2_ventilation_mode | Weight_kg | cmv_tidal_kg | \n",
       "|---|---|---|---|---|---|\n",
       "| 2 |  83.50000                           | 28.57143                            | 15.50000                            | 3.714286                            | 13.06667                            | Airway pressure release ventilation |  8.80                               |  9.488636                           | \n",
       "| 3 |  44.71429                           | 33.57143                            | 22.28572                            | 1.000000                            | 18.00000                            | Airway pressure release ventilation |  4.69                               |  9.533963                           | \n",
       "| 4 | 231.50000                           | 34.00000                            | 15.25000                            | 6.000000                            | 20.00000                            | Airway pressure release ventilation | 42.00                               |  5.511905                           | \n",
       "| 7 | 257.71429                           | 40.14286                            | 20.42857                            | 1.428571                            | 18.28572                            | Airway pressure release ventilation | 19.00                               | 13.563910                           | \n",
       "| 8 | 258.71429                           | 58.57143                            | 30.00000                            | 0.000000                            | 27.42857                            | Airway pressure release ventilation | 19.00                               | 13.616542                           | \n",
       "| 9 | 225.00000                           | 70.00000                            | 29.33333                            | 2.000000                            | 25.16667                            | Airway pressure release ventilation | 19.00                               | 11.842105                           | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "  cmv_tidal cmv_fio2 cmv_pip  cmv_peep cmv_map \n",
       "2  83.50000 28.57143 15.50000 3.714286 13.06667\n",
       "3  44.71429 33.57143 22.28572 1.000000 18.00000\n",
       "4 231.50000 34.00000 15.25000 6.000000 20.00000\n",
       "7 257.71429 40.14286 20.42857 1.428571 18.28572\n",
       "8 258.71429 58.57143 30.00000 0.000000 27.42857\n",
       "9 225.00000 70.00000 29.33333 2.000000 25.16667\n",
       "  D2_ventilation_mode                 Weight_kg cmv_tidal_kg\n",
       "2 Airway pressure release ventilation  8.80      9.488636   \n",
       "3 Airway pressure release ventilation  4.69      9.533963   \n",
       "4 Airway pressure release ventilation 42.00      5.511905   \n",
       "7 Airway pressure release ventilation 19.00     13.563910   \n",
       "8 Airway pressure release ventilation 19.00     13.616542   \n",
       "9 Airway pressure release ventilation 19.00     11.842105   "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "aprv_setting$cmv_tidal_kg<- aprv_setting$cmv_tidal/aprv_setting$Weight_kg\n",
    "head(aprv_setting)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<dl class=dl-horizontal>\n",
       "\t<dt>cmv_tidal_kg</dt>\n",
       "\t\t<dd>10.0669426222409</dd>\n",
       "\t<dt>cmv_fio2</dt>\n",
       "\t\t<dd>61.0319397163415</dd>\n",
       "\t<dt>cmv_pip</dt>\n",
       "\t\t<dd>25.9035423321951</dd>\n",
       "\t<dt>cmv_peep</dt>\n",
       "\t\t<dd>0.999651564365854</dd>\n",
       "\t<dt>cmv_map</dt>\n",
       "\t\t<dd>23.2675957565854</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description*}\n",
       "\\item[cmv\\textbackslash{}\\_tidal\\textbackslash{}\\_kg] 10.0669426222409\n",
       "\\item[cmv\\textbackslash{}\\_fio2] 61.0319397163415\n",
       "\\item[cmv\\textbackslash{}\\_pip] 25.9035423321951\n",
       "\\item[cmv\\textbackslash{}\\_peep] 0.999651564365854\n",
       "\\item[cmv\\textbackslash{}\\_map] 23.2675957565854\n",
       "\\end{description*}\n"
      ],
      "text/markdown": [
       "cmv_tidal_kg\n",
       ":   10.0669426222409cmv_fio2\n",
       ":   61.0319397163415cmv_pip\n",
       ":   25.9035423321951cmv_peep\n",
       ":   0.999651564365854cmv_map\n",
       ":   23.2675957565854\n",
       "\n"
      ],
      "text/plain": [
       "cmv_tidal_kg     cmv_fio2      cmv_pip     cmv_peep      cmv_map \n",
       "  10.0669426   61.0319397   25.9035423    0.9996516   23.2675958 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<dl class=dl-horizontal>\n",
       "\t<dt>cmv_tidal_kg</dt>\n",
       "\t\t<dd>3.30765454306125</dd>\n",
       "\t<dt>cmv_fio2</dt>\n",
       "\t\t<dd>23.0879317697911</dd>\n",
       "\t<dt>cmv_pip</dt>\n",
       "\t\t<dd>6.62194307154047</dd>\n",
       "\t<dt>cmv_peep</dt>\n",
       "\t\t<dd>1.9619245143465</dd>\n",
       "\t<dt>cmv_map</dt>\n",
       "\t\t<dd>6.09287166650766</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description*}\n",
       "\\item[cmv\\textbackslash{}\\_tidal\\textbackslash{}\\_kg] 3.30765454306125\n",
       "\\item[cmv\\textbackslash{}\\_fio2] 23.0879317697911\n",
       "\\item[cmv\\textbackslash{}\\_pip] 6.62194307154047\n",
       "\\item[cmv\\textbackslash{}\\_peep] 1.9619245143465\n",
       "\\item[cmv\\textbackslash{}\\_map] 6.09287166650766\n",
       "\\end{description*}\n"
      ],
      "text/markdown": [
       "cmv_tidal_kg\n",
       ":   3.30765454306125cmv_fio2\n",
       ":   23.0879317697911cmv_pip\n",
       ":   6.62194307154047cmv_peep\n",
       ":   1.9619245143465cmv_map\n",
       ":   6.09287166650766\n",
       "\n"
      ],
      "text/plain": [
       "cmv_tidal_kg     cmv_fio2      cmv_pip     cmv_peep      cmv_map \n",
       "    3.307655    23.087932     6.621943     1.961925     6.092872 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#APRV setting\n",
    "colMeans(aprv_setting[,c(\"cmv_tidal_kg\",\"cmv_fio2\",\"cmv_pip\",\"cmv_peep\",\"cmv_map\")], na.rm=TRUE)\n",
    "apply(aprv_setting[,c(\"cmv_tidal_kg\",\"cmv_fio2\",\"cmv_pip\",\"cmv_peep\",\"cmv_map\")], 2, sd, na.rm = TRUE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table>\n",
       "<thead><tr><th></th><th scope=col>cmv_tidal</th><th scope=col>cmv_fio2</th><th scope=col>cmv_pip</th><th scope=col>cmv_peep</th><th scope=col>cmv_map</th><th scope=col>D2_ventilation_mode</th><th scope=col>Weight_kg</th><th scope=col>cmv_tidal_kg</th></tr></thead>\n",
       "<tbody>\n",
       "\t<tr><th scope=row>1</th><td> 19.00000                      </td><td>65.00000                       </td><td>25.00000                       </td><td>9.000000                       </td><td>14.50000                       </td><td>Pressure controlled ventilation</td><td> 2.5                           </td><td> 7.600000                      </td></tr>\n",
       "\t<tr><th scope=row>5</th><td> 26.75000                      </td><td>42.80000                       </td><td>17.50000                       </td><td>7.200000                       </td><td>12.00000                       </td><td>Pressure controlled ventilation</td><td> 2.7                           </td><td> 9.907407                      </td></tr>\n",
       "\t<tr><th scope=row>6</th><td> 51.20000                      </td><td>31.83333                       </td><td>19.00000                       </td><td>6.833333                       </td><td>10.48000                       </td><td>Pressure controlled ventilation</td><td> 4.2                           </td><td>12.190476                      </td></tr>\n",
       "\t<tr><th scope=row>10</th><td>350.00000                      </td><td>41.50000                       </td><td>24.00000                       </td><td>9.333333                       </td><td>15.00000                       </td><td>Pressure controlled ventilation</td><td>30.0                           </td><td>11.666667                      </td></tr>\n",
       "\t<tr><th scope=row>11</th><td> 40.71429                      </td><td>35.00000                       </td><td>20.14286                       </td><td>6.857143                       </td><td>12.14286                       </td><td>Pressure controlled ventilation</td><td> 2.6                           </td><td>15.659341                      </td></tr>\n",
       "\t<tr><th scope=row>12</th><td>120.00000                      </td><td>40.16667                       </td><td>23.00000                       </td><td>7.400000                       </td><td>13.66000                       </td><td>Volume controlled ventilation  </td><td>12.0                           </td><td>10.000000                      </td></tr>\n",
       "</tbody>\n",
       "</table>\n"
      ],
      "text/latex": [
       "\\begin{tabular}{r|llllllll}\n",
       "  & cmv\\_tidal & cmv\\_fio2 & cmv\\_pip & cmv\\_peep & cmv\\_map & D2\\_ventilation\\_mode & Weight\\_kg & cmv\\_tidal\\_kg\\\\\n",
       "\\hline\n",
       "\t1 &  19.00000                       & 65.00000                        & 25.00000                        & 9.000000                        & 14.50000                        & Pressure controlled ventilation &  2.5                            &  7.600000                      \\\\\n",
       "\t5 &  26.75000                       & 42.80000                        & 17.50000                        & 7.200000                        & 12.00000                        & Pressure controlled ventilation &  2.7                            &  9.907407                      \\\\\n",
       "\t6 &  51.20000                       & 31.83333                        & 19.00000                        & 6.833333                        & 10.48000                        & Pressure controlled ventilation &  4.2                            & 12.190476                      \\\\\n",
       "\t10 & 350.00000                       & 41.50000                        & 24.00000                        & 9.333333                        & 15.00000                        & Pressure controlled ventilation & 30.0                            & 11.666667                      \\\\\n",
       "\t11 &  40.71429                       & 35.00000                        & 20.14286                        & 6.857143                        & 12.14286                        & Pressure controlled ventilation &  2.6                            & 15.659341                      \\\\\n",
       "\t12 & 120.00000                       & 40.16667                        & 23.00000                        & 7.400000                        & 13.66000                        & Volume controlled ventilation   & 12.0                            & 10.000000                      \\\\\n",
       "\\end{tabular}\n"
      ],
      "text/markdown": [
       "\n",
       "| <!--/--> | cmv_tidal | cmv_fio2 | cmv_pip | cmv_peep | cmv_map | D2_ventilation_mode | Weight_kg | cmv_tidal_kg | \n",
       "|---|---|---|---|---|---|\n",
       "| 1 |  19.00000                       | 65.00000                        | 25.00000                        | 9.000000                        | 14.50000                        | Pressure controlled ventilation |  2.5                            |  7.600000                       | \n",
       "| 5 |  26.75000                       | 42.80000                        | 17.50000                        | 7.200000                        | 12.00000                        | Pressure controlled ventilation |  2.7                            |  9.907407                       | \n",
       "| 6 |  51.20000                       | 31.83333                        | 19.00000                        | 6.833333                        | 10.48000                        | Pressure controlled ventilation |  4.2                            | 12.190476                       | \n",
       "| 10 | 350.00000                       | 41.50000                        | 24.00000                        | 9.333333                        | 15.00000                        | Pressure controlled ventilation | 30.0                            | 11.666667                       | \n",
       "| 11 |  40.71429                       | 35.00000                        | 20.14286                        | 6.857143                        | 12.14286                        | Pressure controlled ventilation |  2.6                            | 15.659341                       | \n",
       "| 12 | 120.00000                       | 40.16667                        | 23.00000                        | 7.400000                        | 13.66000                        | Volume controlled ventilation   | 12.0                            | 10.000000                       | \n",
       "\n",
       "\n"
      ],
      "text/plain": [
       "   cmv_tidal cmv_fio2 cmv_pip  cmv_peep cmv_map \n",
       "1   19.00000 65.00000 25.00000 9.000000 14.50000\n",
       "5   26.75000 42.80000 17.50000 7.200000 12.00000\n",
       "6   51.20000 31.83333 19.00000 6.833333 10.48000\n",
       "10 350.00000 41.50000 24.00000 9.333333 15.00000\n",
       "11  40.71429 35.00000 20.14286 6.857143 12.14286\n",
       "12 120.00000 40.16667 23.00000 7.400000 13.66000\n",
       "   D2_ventilation_mode             Weight_kg cmv_tidal_kg\n",
       "1  Pressure controlled ventilation  2.5       7.600000   \n",
       "5  Pressure controlled ventilation  2.7       9.907407   \n",
       "6  Pressure controlled ventilation  4.2      12.190476   \n",
       "10 Pressure controlled ventilation 30.0      11.666667   \n",
       "11 Pressure controlled ventilation  2.6      15.659341   \n",
       "12 Volume controlled ventilation   12.0      10.000000   "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#CMV setting\n",
    "cmv_setting$cmv_tidal_kg<- cmv_setting$cmv_tidal/cmv_setting$Weight_kg\n",
    "head(cmv_setting)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<dl class=dl-horizontal>\n",
       "\t<dt>cmv_tidal_kg</dt>\n",
       "\t\t<dd>3.66465510089467</dd>\n",
       "\t<dt>cmv_fio2</dt>\n",
       "\t\t<dd>55.288528130303</dd>\n",
       "\t<dt>cmv_pip</dt>\n",
       "\t\t<dd>20.5784126801212</dd>\n",
       "\t<dt>cmv_peep</dt>\n",
       "\t\t<dd>5.99565657557576</dd>\n",
       "\t<dt>cmv_map</dt>\n",
       "\t\t<dd>11.8775974214364</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description*}\n",
       "\\item[cmv\\textbackslash{}\\_tidal\\textbackslash{}\\_kg] 3.66465510089467\n",
       "\\item[cmv\\textbackslash{}\\_fio2] 55.288528130303\n",
       "\\item[cmv\\textbackslash{}\\_pip] 20.5784126801212\n",
       "\\item[cmv\\textbackslash{}\\_peep] 5.99565657557576\n",
       "\\item[cmv\\textbackslash{}\\_map] 11.8775974214364\n",
       "\\end{description*}\n"
      ],
      "text/markdown": [
       "cmv_tidal_kg\n",
       ":   3.66465510089467cmv_fio2\n",
       ":   55.288528130303cmv_pip\n",
       ":   20.5784126801212cmv_peep\n",
       ":   5.99565657557576cmv_map\n",
       ":   11.8775974214364\n",
       "\n"
      ],
      "text/plain": [
       "cmv_tidal_kg     cmv_fio2      cmv_pip     cmv_peep      cmv_map \n",
       "    3.664655    55.288528    20.578413     5.995657    11.877597 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<dl class=dl-horizontal>\n",
       "\t<dt>cmv_tidal_kg</dt>\n",
       "\t\t<dd>4.65910143044025</dd>\n",
       "\t<dt>cmv_fio2</dt>\n",
       "\t\t<dd>19.1820301500779</dd>\n",
       "\t<dt>cmv_pip</dt>\n",
       "\t\t<dd>10.1090505477355</dd>\n",
       "\t<dt>cmv_peep</dt>\n",
       "\t\t<dd>3.38272009161348</dd>\n",
       "\t<dt>cmv_map</dt>\n",
       "\t\t<dd>7.02230066692817</dd>\n",
       "</dl>\n"
      ],
      "text/latex": [
       "\\begin{description*}\n",
       "\\item[cmv\\textbackslash{}\\_tidal\\textbackslash{}\\_kg] 4.65910143044025\n",
       "\\item[cmv\\textbackslash{}\\_fio2] 19.1820301500779\n",
       "\\item[cmv\\textbackslash{}\\_pip] 10.1090505477355\n",
       "\\item[cmv\\textbackslash{}\\_peep] 3.38272009161348\n",
       "\\item[cmv\\textbackslash{}\\_map] 7.02230066692817\n",
       "\\end{description*}\n"
      ],
      "text/markdown": [
       "cmv_tidal_kg\n",
       ":   4.65910143044025cmv_fio2\n",
       ":   19.1820301500779cmv_pip\n",
       ":   10.1090505477355cmv_peep\n",
       ":   3.38272009161348cmv_map\n",
       ":   7.02230066692817\n",
       "\n"
      ],
      "text/plain": [
       "cmv_tidal_kg     cmv_fio2      cmv_pip     cmv_peep      cmv_map \n",
       "    4.659101    19.182030    10.109051     3.382720     7.022301 "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#CMV setting\n",
    "colMeans(cmv_setting[,c(\"cmv_tidal_kg\",\"cmv_fio2\",\"cmv_pip\",\"cmv_peep\",\"cmv_map\")], na.rm=TRUE)\n",
    "apply(cmv_setting[,c(\"cmv_tidal_kg\",\"cmv_fio2\",\"cmv_pip\",\"cmv_peep\",\"cmv_map\")], 2, sd, na.rm = TRUE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "p_map<-ggplot(data=hfov_setting, aes(hfov_setting$map,..density..))+\n",
    "        geom_histogram(color=\"black\", fill=\"lightblue\", bins=6)+xlab(\"MAP\")+ylab(\"Probability Density\")+\n",
    "        theme_bw(base_size = 16)+\n",
    "         theme(plot.background = element_blank(),#axis.title.y=element_blank(),\n",
    "           panel.grid.major = element_blank(),\n",
    "           panel.grid.minor = element_blank() )+\n",
    "         theme(panel.border= element_blank())+\n",
    "         theme(axis.line.x = element_line(color=\"black\", size = 0.5),axis.line.y = element_line(color=\"black\", size = 0.5))\n",
    "\n",
    "p_fio2<-ggplot(data=hfov_setting, aes(hfov_setting$fio2,..density..))+ylim(0, 0.04)+\n",
    "        geom_histogram(color=\"black\", fill=\"lightblue\", bins=6)+xlab(\"FiO2\")+\n",
    "        theme_bw(base_size = 16)+\n",
    "        theme(plot.background = element_blank(),\n",
    "           panel.grid.major = element_blank(),\n",
    "           panel.grid.minor = element_blank(),axis.title.y=element_blank())+\n",
    "\n",
    "         theme(panel.border= element_blank())+\n",
    "         theme(axis.line.x = element_line(color=\"black\", size = 0.5),axis.line.y = element_line(color=\"black\", size = 0.5))\n",
    "p_amplitue<-ggplot(data=hfov_setting, aes(hfov_setting$amplitude,..density..))+ylim(0, 0.04)+\n",
    "        geom_histogram(color=\"black\", fill=\"lightblue\", bins=6)+xlab(\"Amplitude\")+\n",
    "        theme_bw(base_size = 16)+\n",
    "        theme(plot.background = element_blank(),\n",
    "           panel.grid.major = element_blank(),\n",
    "           panel.grid.minor = element_blank(),axis.title.y=element_blank())+theme(panel.border= element_blank())+\n",
    "         theme(axis.line.x = element_line(color=\"black\", size = 0.5),axis.line.y = element_line(color=\"black\", size = 0.5))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "p_map<-ggplot(data=hfv, aes(hfv$map,..density..))+\n",
    "        geom_histogram(color=\"black\", fill=\"lightblue\", bins=6)+xlab(\"MAP\")+\n",
    "        theme_bw(base_size = 16)+\n",
    "         theme(plot.background = element_blank(),axis.title.y=element_blank(),\n",
    "           panel.grid.major = element_blank(),\n",
    "           panel.grid.minor = element_blank() )+\n",
    "         theme(panel.border= element_blank())+\n",
    "         theme(axis.line.x = element_line(color=\"black\", size = 0.5),axis.line.y = element_line(color=\"black\", size = 0.5))\n",
    "\n",
    "p_fio2<-ggplot(data=hfv, aes(hfv$fio2,..density..))+ylim(0, 0.04)+\n",
    "        geom_histogram(color=\"black\", fill=\"lightblue\", bins=6)+xlab(\"FiO2\")+\n",
    "        theme_bw(base_size = 16)+\n",
    "        theme(plot.background = element_blank(),\n",
    "           panel.grid.major = element_blank(),\n",
    "           panel.grid.minor = element_blank(),axis.title.y=element_blank())+\n",
    "\n",
    "         theme(panel.border= element_blank())+\n",
    "         theme(axis.line.x = element_line(color=\"black\", size = 0.5),axis.line.y = element_line(color=\"black\", size = 0.5))\n",
    "p_amplitue<-ggplot(data=hfv, aes(hfv$amplitude,..density..))+ylim(0, 0.04)+\n",
    "        geom_histogram(color=\"black\", fill=\"lightblue\", bins=6)+xlab(\"Amplitude\")+\n",
    "        theme_bw(base_size = 16)+\n",
    "        theme(plot.background = element_blank(),\n",
    "           panel.grid.major = element_blank(),\n",
    "           panel.grid.minor = element_blank(),axis.title.y=element_blank() )+\n",
    "         theme(panel.border= element_blank())+\n",
    "         theme(axis.line.x = element_line(color=\"black\", size = 0.5),axis.line.y = element_line(color=\"black\", size = 0.5))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Warning message:\n",
      "“Removed 1 rows containing non-finite values (stat_bin).”Warning message:\n",
      "“Removed 1 rows containing non-finite values (stat_bin).”Warning message:\n",
      "“Removed 2 rows containing non-finite values (stat_bin).”Warning message:\n",
      "“Removed 2 rows containing non-finite values (stat_bin).”"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94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\nHQAAAICLQMcOAAAAwEWgYwcAAADgIv4CqTT1LwQ5ERwAAAAASUVORK5CYII=",
      "text/plain": [
       "plot without title"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "grid.arrange(p_map,p_fio2, p_amplitue,p_amplitue, ncol=2,\n",
    "             left=textGrob(\"Probability Density\",gp=gpar(fontsize=20,font=3),rot=90))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "R",
   "language": "R",
   "name": "ir"
  },
  "language_info": {
   "codemirror_mode": "r",
   "file_extension": ".r",
   "mimetype": "text/x-r-source",
   "name": "R",
   "pygments_lexer": "r",
   "version": "3.5.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}