sepsis-3-main.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sepsis-3 in MIMIC-III\n",
    "\n",
    "This is the primary notebook for analyzing sepsis-3 in the MIMIC-III database. Before running this notebook, you'll need the `sepsis3-df.csv` file in the local directory: either by downloading it directly from PhysioNet or running the SQL scripts enclosed on the MIMIC-III database. See `sepsis-3-get-data.ipynb` for more detail."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import print_function\n",
    "\n",
    "# Import libraries\n",
    "import os\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt\n",
    "import subprocess\n",
    "import tableone\n",
    "from collections import OrderedDict\n",
    "from statsmodels.formula.api import logit\n",
    "from IPython.display import display, HTML # used to print out pretty pandas dataframes\n",
    "\n",
    "from sepsis_utils import sepsis_utils as su\n",
    "from sepsis_utils import roc_utils as ru\n",
    "\n",
    "# used to calculate AUROC/accuracy\n",
    "from sklearn import metrics\n",
    "\n",
    "# default colours for prettier plots\n",
    "col = [[0.9047, 0.1918, 0.1988],\n",
    "    [0.2941, 0.5447, 0.7494],\n",
    "    [0.3718, 0.7176, 0.3612],\n",
    "    [1.0000, 0.5482, 0.1000],\n",
    "    [0.4550, 0.4946, 0.4722],\n",
    "    [0.6859, 0.4035, 0.2412],\n",
    "    [0.9718, 0.5553, 0.7741],\n",
    "    [0.5313, 0.3359, 0.6523]];\n",
    "marker = ['v','o','d','^','s','o','+']\n",
    "ls = ['-','-','-','-','-','s','--','--']\n",
    "\n",
    "import colorsys\n",
    "def gg_color_hue(n):\n",
    "    hues = np.linspace(15, 375, n)\n",
    "    hsv_tuples = [(x*1.0/360.0, 0.5, 0.8) for x in hues]\n",
    "    rgb_tuples = map(lambda x: colorsys.hsv_to_rgb(*x), hsv_tuples)\n",
    "    return rgb_tuples\n",
    "\n",
    "# path used for data - can be a relative path to the current folder\n",
    "data_path = 'data'\n",
    "\n",
    "%matplotlib inline\n",
    "plt.style.use('ggplot')\n",
    "\n",
    "font = {'family' : 'DejaVu Sans',\n",
    "        'size'   : 20}\n",
    "\n",
    "matplotlib.rc('font', **font)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(os.path.join(data_path,'sepsis3-df.csv'))\n",
    "\n",
    "# add the composite outcome\n",
    "df['composite_outcome'] = ( (df['hospital_expire_flag']==1) | (df['icu_los']>=3) ).astype(int)\n",
    "\n",
    "labels = OrderedDict([['suspicion_poe', 'BC + ABX (Prescribed)']])\n",
    "\n",
    "# add some other useful variables\n",
    "df['blood culture'] = (~df['blood_culture_time'].isnull())\n",
    "df['suspicion_poe'] = (~df['suspected_infection_time_poe_days'].isnull())\n",
    "\n",
    "df['abx_poe'] = (~df['antibiotic_time_poe'].isnull())\n",
    "\n",
    "df['sepsis-3'] = ((df['suspicion_poe']==1) & (df['sofa']>=2)).astype(int)\n",
    "df['sofa>=2'] = (df['sofa']>=2).astype(int)\n",
    "\n",
    "\n",
    "for c in ['intime','outtime',\n",
    "          'suspected_infection_time_mv','suspected_infection_time',\n",
    "          'suspected_infection_time_poe', 'blood_culture_time']:\n",
    "    if c in df.columns:\n",
    "        if df[c].dtype == 'object':\n",
    "            df[c] = pd.to_datetime(df[c])\n",
    "\n",
    "# list of the sepsis definitions\n",
    "sepsis_list = ['sepsis_angus','sepsis_martin', 'sepsis_explicit',\n",
    "               'sepsis_cdc','sepsis_nqf',\n",
    "               'sepsis-3']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Results for Abstract"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11791 patients.\n",
      " 7061 (60%) suspected of infection.\n",
      " 8868 (75.21%) have SOFA >= 2.\n",
      " 5784 (49.05%) have Sepsis-3 criteria (intersection of above two).\n",
      " 1277 (10.83%) have suspicion and SOFA < 2.\n",
      " 3368 (28.56%) - sepsis_angus\n",
      " 1734 (14.71%) - sepsis_martin\n",
      " 1062 (9.01%) - sepsis_explicit\n",
      " 3761 (31.90%) - sepsis_cdc\n",
      " 1302 (11.04%) - sepsis_nqf\n",
      " 5784 (49.05%) - sepsis-3\n",
      " 2917 (24.74%) have Sepsis-3 criteria but not Angus.\n",
      "\n",
      " === Cronbach Alpha ===\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>angus</th>\n",
       "      <th>martin</th>\n",
       "      <th>explicit</th>\n",
       "      <th>cdc</th>\n",
       "      <th>nqf</th>\n",
       "      <th>sepsis-3</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>angus</th>\n",
       "      <td></td>\n",
       "      <td>0.69 [0.68-0.70]</td>\n",
       "      <td>0.62 [0.61-0.63]</td>\n",
       "      <td>0.62 [0.60-0.63]</td>\n",
       "      <td>0.63 [0.62-0.65]</td>\n",
       "      <td>0.62 [0.61-0.64]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>martin</th>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0.85 [0.84-0.85]</td>\n",
       "      <td>0.54 [0.52-0.55]</td>\n",
       "      <td>0.91 [0.90-0.92]</td>\n",
       "      <td>0.49 [0.48-0.50]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>explicit</th>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0.49 [0.48-0.51]</td>\n",
       "      <td>0.89 [0.89-0.90]</td>\n",
       "      <td>0.40 [0.39-0.41]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cdc</th>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0.53 [0.52-0.55]</td>\n",
       "      <td>0.76 [0.75-0.76]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>nqf</th>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td></td>\n",
       "      <td>0.45 [0.43-0.46]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# see get-data for the exclusions\n",
    "print('{:5g} patients.'.format(df.shape[0]))\n",
    "\n",
    "print('{:5g} ({:2.0f}%) suspected of infection.'.format(\n",
    "        df['suspicion_poe'].sum(), df['suspicion_poe'].sum()*100.0/df.shape[0]))\n",
    "\n",
    "print('{:5g} ({:2.2f}%) have SOFA >= 2.'.format(\n",
    "    df['sofa>=2'].sum(),100.0*df['sofa>=2'].mean()))\n",
    "\n",
    "print('{:5g} ({:2.2f}%) have Sepsis-3 criteria (intersection of above two).'.format(\n",
    "    df['sepsis-3'].sum(),100.0*df['sepsis-3'].mean()))\n",
    "\n",
    "print('{:5g} ({:2.2f}%) have suspicion and SOFA < 2.'.format(\n",
    "        ((df['sofa>=2']==0) & (df['suspicion_poe']==1)).sum(),\n",
    "        ((df['sofa>=2']==0) & (df['suspicion_poe']==1)).sum()*100.0/df.shape[0]))\n",
    "\n",
    "for c in sepsis_list:\n",
    "    print('{:5g} ({:3.2f}%) - {}'.format(\n",
    "        df[c].sum(), df[c].sum()*100.0/df.shape[0], c))\n",
    "\n",
    "print('{:5g} ({:2.2f}%) have Sepsis-3 criteria but not Angus.'.format(\n",
    "        ((df['sepsis_angus']==0) & (df['sepsis-3']==1)).sum(),\n",
    "        ((df['sepsis_angus']==0) & (df['sepsis-3']==1)).sum()*100.0/df.shape[0]))\n",
    "\n",
    "\n",
    "np.random.seed(21381)\n",
    "# cronbach alpha for construct validity\n",
    "calpha = su.cronbach_alpha_table(df, sepsis_list)\n",
    "\n",
    "# remove \"sepsis_\" prefix from columns/indices\n",
    "calpha.columns = [x.replace('sepsis_','') for x in calpha.columns]\n",
    "calpha.index = [x.replace('sepsis_','') for x in calpha.index]\n",
    "\n",
    "print('\\n === Cronbach Alpha ===')\n",
    "display(HTML(calpha.fillna('').to_html()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Results section\n",
    "\n",
    "We now print out the results in the same order as they are in the paper."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Demographics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "N                   \t11791\n",
      "age                 \t64.5 [51.1, 78.5]\n",
      "gender              \t6478 (54.9%)\n",
      "bmi                 \t28.7 +- 8.4\n",
      "Race                \t\n",
      "White               \t8497 (72.1%)\n",
      "Black               \t1110 (9.4%)\n",
      "Hispanic            \t 457 (9.4%)\n",
      "elixhauser_hospital \t1.0 [-1.0, 6.0]\n",
      "qsofa               \t2.0 [1.0, 2.0]\n",
      "sirs                \t3.0 [2.0, 3.0]\n",
      "sofa                \t3.0 [2.0, 5.0]\n",
      "mlods               \n",
      "lactate_max         \n",
      "vent                \t4151 (35.2%)\n",
      "icu_los             \t1.9 [1.1, 3.5]\n",
      "hosp_los            \t5.6 [3.1, 9.6]\n",
      "thirtyday_expire_flag\t1619 (13.7%)\n",
      "hospital_expire_flag\t1277 (10.8%)\n",
      "\n",
      "Alive vs. dead\n",
      "N                   \t10514     \t1277     \t     \n",
      "age                 \t63.3 [50.0, 77.5]\t74.9 [61.8, 83.7]\t< 0.001\n",
      "gender              \t5795 (55.1%)\t 683 (53.5%)\t0.281\n",
      "bmi                 \t28.7 +- 8.2\t28.1 +- 10.1\t0.167\n",
      "ethnicity           \t          \t          \t< 0.001\n",
      "  black             \t1036 (9.9%)\t  74 (5.8%)\t     \n",
      "  hispanic          \t 424 (4.0%)\t  33 (2.6%)\t     \n",
      "  other             \t1424 (13.5%)\t 303 (23.7%)\t     \n",
      "  white             \t7630 (72.6%)\t 867 (67.9%)\t     \n",
      "elixhauser_hospital \t0.0 [-1.0, 6.0]\t5.0 [0.0, 10.0]\t< 0.001\n",
      "qsofa               \t2.0 [1.0, 2.0]\t2.0 [2.0, 2.0]\t< 0.001\n",
      "sirs                \t3.0 [2.0, 3.0]\t3.0 [3.0, 4.0]\t< 0.001\n",
      "sofa                \t3.0 [1.0, 5.0]\t6.0 [4.0, 10.0]\t< 0.001\n",
      "mlods               \n",
      "lactate_max         \n",
      "vent                \t3275 (31.1%)\t 876 (68.6%)\t< 0.001\n",
      "icu_los             \t1.9 [1.1, 3.2]\t2.4 [1.1, 5.6]\t< 0.001\n",
      "hosp_los            \t5.8 [3.4, 9.8]\t3.2 [1.1, 7.9]\t< 0.001\n",
      "thirtyday_expire_flag\t 375 (3.6%)\t1244 (97.4%)\t< 0.001\n",
      "hospital_expire_flag\t   0 (0.0%)\t1277 (100.0%)\t< 0.001\n",
      "\n",
      "10189 have SIRS >= 2 (86.41%) on admission.\n",
      " 7497 have qSOFA >= 2 (63.58%) on admission.\n",
      " 8868 have SOFA >= 2 (75.21%).\n",
      " 9285 have LODS >= 2 (78.75%).\n"
     ]
    }
   ],
   "source": [
    "# Call the print_demographics subfunction, which prints out a reasonably formatted table\n",
    "su.print_demographics(df)\n",
    "\n",
    "print('\\nAlive vs. dead')\n",
    "su.print_demographics(df, idx=(df.hospital_expire_flag.values==1))\n",
    "\n",
    "print('')\n",
    "\n",
    "print('{:5g} have SIRS >= 2 ({:2.2f}%) on admission.'.format(\n",
    "    (df.sirs.values >= 2).sum(),100.0*(df.sirs.values >= 2).mean()))\n",
    "\n",
    "print('{:5g} have qSOFA >= 2 ({:2.2f}%) on admission.'.format(\n",
    "    (df.qsofa.values >= 2).sum(),100.0*(df.qsofa.values >= 2).mean()))\n",
    "\n",
    "print('{:5g} have SOFA >= 2 ({:2.2f}%).'.format(\n",
    "    (df.sofa.values >= 2).sum(),100.0*(df.sofa.values >= 2).mean()))\n",
    "\n",
    "print('{:5g} have LODS >= 2 ({:2.2f}%).'.format(\n",
    "    (df.lods.values >= 2).sum(),100.0*(df.lods.values >= 2).mean()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Frequency of primary/secondary outcomes for each score\n",
    "\n",
    "First print a table, then plot the figure."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "=== hospital_expire_flag ===\n",
      "\n",
      "Criteria       \tN       \tp(death|c)\tp(death|~c)\n",
      "suspicion_poe  \t7061, 59.9%\t12.5%\t\t8.3%\n",
      "sofa>=2        \t8868, 75.2%\t13.2%\t\t3.6%\n",
      "sepsis-3       \t5784, 49.1%\t14.5%\t\t7.3%\n",
      "sepsis_angus   \t3368, 28.6%\t17.9%\t\t8.0%\n",
      "sepsis_martin  \t1734, 14.7%\t22.7%\t\t8.8%\n",
      "sepsis_explicit\t1062, 9.0%\t30.1%\t\t8.9%\n",
      "sepsis_cdc     \t3761, 31.9%\t18.6%\t\t7.2%\n",
      "sepsis_nqf     \t1302, 11.0%\t27.2%\t\t8.8%\n",
      "\n",
      "=== composite_outcome ===\n",
      "\n",
      "Criteria       \tN       \tp(death|c)\tp(death|~c)\n",
      "suspicion_poe  \t7061, 59.9%\t46.3%\t\t19.9%\n",
      "sofa>=2        \t8868, 75.2%\t41.2%\t\t19.2%\n",
      "sepsis-3       \t5784, 49.1%\t50.0%\t\t21.9%\n",
      "sepsis_angus   \t3368, 28.6%\t61.2%\t\t25.5%\n",
      "sepsis_martin  \t1734, 14.7%\t60.1%\t\t31.5%\n",
      "sepsis_explicit\t1062, 9.0%\t70.7%\t\t32.2%\n",
      "sepsis_cdc     \t3761, 31.9%\t61.1%\t\t23.8%\n",
      "sepsis_nqf     \t1302, 11.0%\t64.7%\t\t32.1%\n"
     ]
    }
   ],
   "source": [
    "# list probability of outcome for each score\n",
    "scores = ['suspicion_poe','sofa>=2',\n",
    "          'sepsis-3',\n",
    "          'sepsis_angus','sepsis_martin','sepsis_explicit',\n",
    "          'sepsis_cdc','sepsis_nqf']\n",
    "\n",
    "scores_dict = {\n",
    "    'suspicion_poe': 'Suspected infection',\n",
    "    'sofa>=2': 'SOFA >= 2',\n",
    "    'sepsis-3': 'Sepsis-3',\n",
    "    'sepsis_angus': 'Angus et al. criteria',\n",
    "    'sepsis_martin': 'Martin et al. criteria',\n",
    "    'sepsis_explicit': 'Explicit',\n",
    "    'sepsis_cdc': 'CDC',\n",
    "    'sepsis_nqf': 'CMS'\n",
    "}\n",
    "target_header = \"hospital_expire_flag\"\n",
    "idx = df[target_header]==1\n",
    "\n",
    "print()\n",
    "print('=== {} ==='.format(target_header))\n",
    "print()\n",
    "print('{:15s}\\t{:8s}\\t{:5s}\\t{:5s}'.format(\n",
    "    'Criteria','N','p(death|c)', 'p(death|~c)'))\n",
    "for c in scores:\n",
    "    print('{:15s}\\t{:4d}, {:2.1f}%\\t{:2.1f}%\\t\\t{:2.1f}%'.format(\n",
    "            c,\n",
    "            np.sum( df[c]==1 ),\n",
    "            np.sum( df[c]==1 )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&idx )*100.0 / np.sum( df[c]==1 ),\n",
    "            np.sum( (df[c]!=1)&idx )*100.0 / np.sum( df[c]!=1 )\n",
    "        ))\n",
    "    \n",
    "target_header = \"composite_outcome\"\n",
    "idx = df[target_header]==1\n",
    "\n",
    "print()\n",
    "print('=== {} ==='.format(target_header))\n",
    "print()\n",
    "\n",
    "print('{:15s}\\t{:8s}\\t{:5s}\\t{:5s}'.format(\n",
    "    'Criteria','N','p(death|c)', 'p(death|~c)'))\n",
    "for c in scores:\n",
    "    print('{:15s}\\t{:4d}, {:2.1f}%\\t{:2.1f}%\\t\\t{:2.1f}%'.format(\n",
    "            c,\n",
    "            np.sum( df[c]==1 ),\n",
    "            np.sum( df[c]==1 )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&idx )*100.0 / np.sum( df[c]==1 ),\n",
    "            np.sum( (df[c]!=1)&idx )*100.0 / np.sum( df[c]!=1 )\n",
    "        ))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAj4AAAEhCAYAAAB/Wqp/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f3bbffebb90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Criteria       \tN       \tp(death|c)\tp(death|~c)\tp(comp|c)\tp(comp|~c)\n",
      "sofa>=2        \t8868, 75.2%\t13.2%\t\t3.6%\t\t41.2%\t\t19.2%\n",
      "suspicion_poe  \t7061, 59.9%\t12.5%\t\t8.3%\t\t46.3%\t\t19.9%\n",
      "sepsis-3       \t5784, 49.1%\t14.5%\t\t7.3%\t\t50.0%\t\t21.9%\n",
      "sepsis_cdc     \t3761, 31.9%\t18.6%\t\t7.2%\t\t61.1%\t\t23.8%\n",
      "sepsis_angus   \t3368, 28.6%\t17.9%\t\t8.0%\t\t61.2%\t\t25.5%\n",
      "sepsis_martin  \t1734, 14.7%\t22.7%\t\t8.8%\t\t60.1%\t\t31.5%\n",
      "sepsis_nqf     \t1302, 11.0%\t27.2%\t\t8.8%\t\t64.7%\t\t32.1%\n",
      "sepsis_explicit\t1062, 9.0%\t30.1%\t\t8.9%\t\t70.7%\t\t32.2%\n"
     ]
    }
   ],
   "source": [
    "# create a figure of the above frequencies\n",
    "scores = ['suspicion_poe','sofa>=2',\n",
    "          'sepsis-3',\n",
    "          'sepsis_angus','sepsis_martin','sepsis_explicit',\n",
    "          'sepsis_cdc','sepsis_nqf']\n",
    "\n",
    "scores_dict = {\n",
    "    'suspicion_poe': 'Suspected infection',\n",
    "    'sofa>=2': 'SOFA >= 2',\n",
    "    'sepsis-3': 'Sepsis-3',\n",
    "    'sepsis_angus': 'Angus et al. criteria',\n",
    "    'sepsis_martin': 'Martin et al. criteria',\n",
    "    'sepsis_explicit': 'Explicit',\n",
    "    'sepsis_cdc': 'CDC',\n",
    "    'sepsis_nqf': 'CMS'\n",
    "}\n",
    "\n",
    "idx = df['hospital_expire_flag']==1\n",
    "idxComp = df['composite_outcome']==1\n",
    "\n",
    "score_plot = np.zeros( [len(scores), 3] )\n",
    "for i, c in enumerate(scores):\n",
    "    # proportion of patients\n",
    "    score_plot[i, 0] = np.sum( df[c]==1 )*100.0/df.shape[0]\n",
    "    \n",
    "    # with mort\n",
    "    score_plot[i, 1] = np.sum( (df[c]==1)&idx )*100.0 / np.sum( df[c]==1 )\n",
    "    # with comp\n",
    "    score_plot[i, 2] = np.sum( (df[c]==1)&idxComp )*100.0 / np.sum( df[c]==1 )\n",
    "    \n",
    "S = len(scores)\n",
    "\n",
    "idxSort = np.argsort(score_plot[:,0])\n",
    "plt.figure()\n",
    "plt.barh( range(S), score_plot[idxSort,0], color=col[1], align='center')\n",
    "plt.barh( range(S), score_plot[idxSort,0]*score_plot[idxSort,2]/100.0, color=col[3], align='center', height=0.6)\n",
    "plt.barh( range(S), score_plot[idxSort,0]*score_plot[idxSort,1]/100.0, color=col[0], align='center', height=0.4)\n",
    "\n",
    "plt.yticks(range(S), [scores_dict[scores[x]] for x in idxSort])\n",
    "plt.xlabel('Percentage of patients')\n",
    "plt.xlim([0,100])\n",
    "plt.show()\n",
    "\n",
    "print('')\n",
    "print('{:15s}\\t{:8s}\\t{:5s}\\t{:5s}\\t{:5s}\\t{:5s}'.format(\n",
    "    'Criteria','N','p(death|c)', 'p(death|~c)', 'p(comp|c)', 'p(comp|~c)'))\n",
    "for i in idxSort[-1::-1]:\n",
    "    c=scores[i]\n",
    "    print('{:15s}\\t{:4d}, {:2.1f}%\\t{:2.1f}%\\t\\t{:2.1f}%\\t\\t{:2.1f}%\\t\\t{:2.1f}%'.format(\n",
    "            c,\n",
    "            np.sum( df[c]==1 ),\n",
    "            np.sum( df[c]==1 )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&idx )*100.0 / np.sum( df[c]==1 ),\n",
    "            np.sum( (df[c]!=1)&idx )*100.0 / np.sum( df[c]!=1 ),\n",
    "            np.sum( (df[c]==1)&idxComp )*100.0 / np.sum( df[c]==1 ),\n",
    "            np.sum( (df[c]!=1)&idxComp )*100.0 / np.sum( df[c]!=1 )\n",
    "        ))\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAj4AAAEhCAYAAAB/Wqp/AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nHetkAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f3b8ecdbb10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# print the frequencies\n",
    "# list probability of outcome for each score\n",
    "scores = ['suspicion_poe','sofa>=2',\n",
    "          'sepsis-3',\n",
    "          'sepsis_angus','sepsis_martin','sepsis_explicit',\n",
    "          'sepsis_cdc','sepsis_nqf']\n",
    "\n",
    "scores_dict = {\n",
    "    'suspicion_poe': 'Suspected infection',\n",
    "    'sofa>=2': 'SOFA >= 2',\n",
    "    'sepsis-3': 'Sepsis-3',\n",
    "    'sepsis_angus': 'Angus et al. criteria',\n",
    "    'sepsis_martin': 'Martin et al. criteria',\n",
    "    'sepsis_explicit': 'Explicit',\n",
    "    'sepsis_cdc': 'CDC',\n",
    "    'sepsis_nqf': 'CMS'\n",
    "}\n",
    "\n",
    "idx = df['hospital_expire_flag']==1\n",
    "idxComp = df['composite_outcome']==1\n",
    "\n",
    "score_plot = np.zeros( [len(scores), 3] )\n",
    "for i, c in enumerate(scores):\n",
    "    # proportion of patients\n",
    "    score_plot[i, 0] = np.sum( df[c]==1 )*100.0/df.shape[0]\n",
    "    \n",
    "    # with mort\n",
    "    score_plot[i, 1] = np.sum((df[c]==1)&idx)*100.0 / np.sum( df[c]==1 )\n",
    "    # with comp\n",
    "    score_plot[i, 2] = np.sum( (df[c]==1)&idxComp )*100.0 / np.sum( df[c]==1 )\n",
    "    \n",
    "S = len(scores)\n",
    "\n",
    "idxSort = np.argsort(score_plot[:,0])\n",
    "\n",
    "plt.figure()\n",
    "plt.barh( range(S), score_plot[idxSort,0], color=col[1], align='center', label='Patients')\n",
    "#plt.barh( range(S), score_plot[idxSort,2], color=col[3], align='center', height=0.6)\n",
    "plt.barh( range(S), score_plot[idxSort,1], color=col[0], align='center', height=0.4, label='Patients who died')\n",
    "#ax2.set_xlim(ax2.get_xlim()[::-1])\n",
    "\n",
    "#plt.plot( score_plot[idxSort,1], range(S), color=col[0], marker='o', markersize=10)\n",
    "#ax1.barh( range(S), score_plot[idxSort,0]*score_plot[idxSort,2]/100.0, color=col[3], align='center', height=0.6)\n",
    "#\n",
    "#ax2 = ax1.twiny()\n",
    "#ax2.barh( range(S), (score_plot[idxSort,0]*score_plot[idxSort,1]/100.0), color=col[0], align='center', height=0.4)\n",
    "#ax2.set_xlim(ax2.get_xlim()[::-1])\n",
    "#ax2.plot(range(S), score_plot[idxSort,0]*score_plot[idxSort,1]/100.0, color=col[0] )\n",
    "\n",
    "plt.yticks(range(S), [scores_dict[scores[x]] for x in idxSort])\n",
    "plt.xlabel('Percentage of patients')\n",
    "#plt.legend(loc='lower right')\n",
    "plt.xlim([0,100])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Criteria</th>\n",
       "      <th>N</th>\n",
       "      <th>p(death|c)</th>\n",
       "      <th>p(death|~c)</th>\n",
       "      <th>p(comp|c)</th>\n",
       "      <th>p(comp|~c)</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>SOFA &gt;= 2</th>\n",
       "      <td>8868.0</td>\n",
       "      <td>75.21</td>\n",
       "      <td>13.22</td>\n",
       "      <td>3.59</td>\n",
       "      <td>41.16</td>\n",
       "      <td>19.19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Suspected infection</th>\n",
       "      <td>7061.0</td>\n",
       "      <td>59.88</td>\n",
       "      <td>12.55</td>\n",
       "      <td>8.27</td>\n",
       "      <td>46.34</td>\n",
       "      <td>19.85</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sepsis-3</th>\n",
       "      <td>5784.0</td>\n",
       "      <td>49.05</td>\n",
       "      <td>14.45</td>\n",
       "      <td>7.34</td>\n",
       "      <td>50.03</td>\n",
       "      <td>21.92</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CDC</th>\n",
       "      <td>3761.0</td>\n",
       "      <td>31.90</td>\n",
       "      <td>18.64</td>\n",
       "      <td>7.17</td>\n",
       "      <td>61.07</td>\n",
       "      <td>23.84</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Angus et al. criteria</th>\n",
       "      <td>3368.0</td>\n",
       "      <td>28.56</td>\n",
       "      <td>17.90</td>\n",
       "      <td>8.00</td>\n",
       "      <td>61.19</td>\n",
       "      <td>25.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Martin et al. criteria</th>\n",
       "      <td>1734.0</td>\n",
       "      <td>14.71</td>\n",
       "      <td>22.72</td>\n",
       "      <td>8.78</td>\n",
       "      <td>60.09</td>\n",
       "      <td>31.51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CMS</th>\n",
       "      <td>1302.0</td>\n",
       "      <td>11.04</td>\n",
       "      <td>27.19</td>\n",
       "      <td>8.80</td>\n",
       "      <td>64.75</td>\n",
       "      <td>32.11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Explicit</th>\n",
       "      <td>1062.0</td>\n",
       "      <td>9.01</td>\n",
       "      <td>30.13</td>\n",
       "      <td>8.92</td>\n",
       "      <td>70.72</td>\n",
       "      <td>32.25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create a dataframe for these probabilities\n",
    "df_outcome = pd.DataFrame(columns=['Criteria','N','p(death|c)', 'p(death|~c)', 'p(comp|c)', 'p(comp|~c)'], dtype=float)\n",
    "# same print but to csv\n",
    "for i in idxSort[-1::-1]:\n",
    "    c=scores[i]\n",
    "    df_outcome.loc[scores_dict[c], :] = [np.sum( df[c]==1 ),\n",
    "            np.sum( df[c]==1 )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&idx )*100.0 / np.sum( df[c]==1 ),\n",
    "            np.sum( (df[c]!=1)&idx )*100.0 / np.sum( df[c]!=1 ),\n",
    "            np.sum( (df[c]==1)&idxComp )*100.0 / np.sum( df[c]==1 ),\n",
    "            np.sum( (df[c]!=1)&idxComp )*100.0 / np.sum( df[c]!=1 )]\n",
    "    #print('{},{:4d} {:2.1f}%,{:2.1f}%,{:2.1f}%,{:2.1f}%,{:2.1f}%'.format(\n",
    "    #    ))\n",
    "    \n",
    "# display .. maximum of 2 decimal places\n",
    "display(HTML(df_outcome.round(decimals=2).to_html()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Comparison to past literature performance.\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>hospital_expire_flag</th>\n",
       "      <th>composite_outcome</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Seymour et al., 2016</th>\n",
       "      <td>0.74 [0.73 - 0.76]</td>\n",
       "      <td>~0.74 (inferred from eFigure 6)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Raith et al., 2017</th>\n",
       "      <td>0.753 [0.750 - 0.757]</td>\n",
       "      <td>0.736 [0.733 - 0.739]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Our results</th>\n",
       "      <td>0.740 [0.722 - 0.759]</td>\n",
       "      <td>0.688 [0.676 - 0.701]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# note: we need to subselect to the cohort suspected of infection in order to compare AUROCs to previous literature\n",
    "idx = df['suspicion_poe'].values == 1\n",
    "pred_header = 'sofa'\n",
    "\n",
    "print('Comparison to past literature performance.')\n",
    "#print'{}'\n",
    "\n",
    "\n",
    "df_past_literature = pd.DataFrame(columns=['hospital_expire_flag','composite_outcome'])\n",
    "df_past_literature.loc['Seymour et al., 2016',:] = ['0.74 [0.73 - 0.76]', '~0.74 (inferred from eFigure 6)']\n",
    "df_past_literature.loc['Raith et al., 2017',:] = ['0.753 [0.750 - 0.757]', '0.736 [0.733 - 0.739]']\n",
    "np.random.seed(7891)\n",
    "# AUROC of sofa for in-hospital mortality\n",
    "target_header = 'hospital_expire_flag'\n",
    "auc_hosp, ci_hosp = ru.calc_auc(df.loc[idx, pred_header].values, df.loc[idx, target_header].values, with_ci=True)\n",
    "target_header = 'composite_outcome'\n",
    "auc_comp, ci_comp = ru.calc_auc(df.loc[idx, pred_header].values, df.loc[idx, target_header].values, with_ci=True)\n",
    "\n",
    "df_past_literature.loc['Our results', :] = ['{:0.3f} [{:0.3f} - {:0.3f}]'.format(auc_hosp, ci_hosp[0], ci_hosp[1]),\n",
    "                                           '{:0.3f} [{:0.3f} - {:0.3f}]'.format(auc_comp, ci_comp[0], ci_comp[1])]\n",
    "\n",
    "\n",
    "display(HTML(df_past_literature.to_html()))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In the above, it is worth noting that:\n",
    "\n",
    "* The results of Seymour et al. are for a model incorporating age, gender, race, and comorbid status\n",
    "* The results of Raith et al. are for the uncalibrated univariable score (\"crude\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Metric\n",
      "\n",
      "     \tSOFA>=2 \n",
      "TN   \t 2362\n",
      "FP   \t 5218\n",
      "FN   \t  561\n",
      "TP   \t 3650\n",
      "Sens \t87 [86, 88]\n",
      "Spec \t31 [30, 32]\n",
      "PPV  \t41 [40, 42]\n",
      "NPV  \t81 [79, 82]\n",
      "F1   \t 55.81   \n",
      "NTP  \t 30.96   \n",
      "NFP  \t 44.25   \n"
     ]
    }
   ],
   "source": [
    "# operating point statistics on sofa >= 2\n",
    "yhat_dict = OrderedDict([['SOFA>=2', df['sofa>=2']]\n",
    "                        ])\n",
    "stats_all = su.get_op_stats(yhat_dict, df[target_header].values)\n",
    "su.print_op_stats(stats_all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Sepsis\n",
    "\n",
    "Create the sepsis-3 criteria: SOFA >= 2 and suspicion of infection."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11791 (100.00%)  first ICU stay for adults.\n",
      " 7061 (59.88%)  suspected of infection\n",
      " 2803 (23.77%)  with a positive blood culture\n",
      "\n",
      " 7061 (59.88%) - suspicion_poe\n"
     ]
    }
   ],
   "source": [
    "print('{:5g} ({:3.2f}%)  first ICU stay for adults.'.format(\n",
    "        df['icustay_id'].count(), 100))\n",
    "print('{:5g} ({:3.2f}%)  suspected of infection'.format(\n",
    "        np.sum(df['suspicion_poe']),\n",
    "        np.sum(df['suspicion_poe'])*100.0/df.shape[0]))\n",
    "print('{:5g} ({:3.2f}%)  with a positive blood culture'.format(\n",
    "    df['blood_culture_positive'].sum(), df['blood_culture_positive'].sum()*100.0/df.shape[0]))\n",
    "\n",
    "print()\n",
    "for i, c in enumerate(labels):\n",
    "    print('{:5g} ({:3.2f}%) - {}'.format(\n",
    "        df[c].sum(), df[c].sum()*100.0/df.shape[0], c))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Venn diagrams of Sepsis-3 against other criteria"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeUAAAH+CAYAAACiIRY6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nGGNslOCkzBhjjI0S/wWLWSJ8IbhtvwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f3b8ecdbb10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1420 patients (12.0%) satisfied all criteria.\n",
      "5448 patients (46.2%) satisfied no criteria.\n",
      "12.9% (1522) - sepsis_angus & sepsis_martin\n",
      "13.3% (1574) - sepsis-3 & sepsis_martin\n",
      "24.3% (2867) - sepsis-3 & sepsis_angus\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeYAAAH+CAYAAABJFq05AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f3bbc8f0fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1062 patients (9.0%) satisfied all criteria.\n",
      "5715 patients (48.5%) satisfied no criteria.\n",
      "9.2% (1084) - sepsis_cdc & sepsis_nqf\n",
      "10.4% (1228) - sepsis-3 & sepsis_nqf\n",
      "29.9% (3521) - sepsis-3 & sepsis_cdc\n"
     ]
    }
   ],
   "source": [
    "# define labels here\n",
    "# first label = red (top left)\n",
    "# second label = green (top right)\n",
    "# third label = blue (bottom)\n",
    "\n",
    "venn_labels = OrderedDict([\n",
    "        ['sepsis_martin', 'Martin criteria'],\n",
    "        ['sepsis_angus', 'Angus criteria'],\n",
    "        ['sepsis-3', 'Sepsis-3 criteria']\n",
    "    ])\n",
    "su.create_venn_diagram(df, venn_labels)\n",
    "\n",
    "venn_labels = OrderedDict([\n",
    "        ['sepsis_nqf', 'CMS'],\n",
    "        ['sepsis_cdc', 'CDC'],\n",
    "        ['sepsis-3', 'Sepsis-3']\n",
    "    ])\n",
    "su.create_venn_diagram(df, venn_labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Mortality rates for each group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0               dead            alive          \n",
      "sepsis_angus     674 5.7%\t 7749 65.7%  8.0%\n",
      "sepsis_martin    883 7.5%\t 9174 77.8%  8.8%\n",
      "sepsis_explicit  957 8.1%\t 9772 82.9%  8.9%\n",
      "suspicion_poe    391 3.3%\t 4339 36.8%  8.3%\n",
      "sepsis-3         441 3.7%\t 5566 47.2%  7.3%\n",
      "sofa>=2          105 0.9%\t 2818 23.9%  3.6%\n",
      "\n",
      "1\n",
      "sepsis_angus     603 5.1%\t 2765 23.5%  17.9%\n",
      "sepsis_martin    394 3.3%\t 1340 11.4%  22.7%\n",
      "sepsis_explicit  320 2.7%\t  742 6.3%  30.1%\n",
      "suspicion_poe    886 7.5%\t 6175 52.4%  12.5%\n",
      "sepsis-3         836 7.1%\t 4948 42.0%  14.5%\n",
      "sofa>=2         1172 9.9%\t 7696 65.3%  13.2%\n"
     ]
    }
   ],
   "source": [
    "target_header = \"hospital_expire_flag\"\n",
    "idx = df[target_header]==1\n",
    "\n",
    "# make a confusion matrix with multiple scores in each square\n",
    "scores = ['sepsis_angus','sepsis_martin','sepsis_explicit','suspicion_poe','sepsis-3','sofa>=2']\n",
    "\n",
    "print('{:15s} {:15s} {:15s}'.format('0','dead','alive','outcome %'))\n",
    "\n",
    "for c in scores:\n",
    "    print('{:15s} {:4d} {:1.1f}%\\t{:5d} {:1.1f}%  {:1.1f}%'.format(\n",
    "            c,\n",
    "            np.sum( (df[c]!=1)&idx ),  np.sum( (df[c]!=1)&idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]!=1)&~idx ), np.sum( (df[c]!=1)&~idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]!=1)&idx )*100.0/np.sum(df[c]!=1)\n",
    "        ))\n",
    "print()\n",
    "print('1')\n",
    "for c in scores:\n",
    "    print('{:15s} {:4d} {:1.1f}%\\t{:5d} {:1.1f}%  {:1.1f}%'.format(\n",
    "            c,\n",
    "            np.sum( (df[c]==1)&idx ),  np.sum( (df[c]==1)&idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&~idx ), np.sum( (df[c]==1)&~idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&idx )*100.0/np.sum(df[c]==1)\n",
    "        ))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Composite outcome for each group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0               dead/hi-LOS  alive/lo-LOS   \n",
      "sepsis_angus    2150 18.2%    6273 53.2%  25.5%\n",
      "sepsis_martin   3169 26.9%    6888 58.4%  31.5%\n",
      "sepsis_explicit 3460 29.3%    7269 61.6%  32.2%\n",
      "suspicion_poe    939 8.0%    3791 32.2%  19.9%\n",
      "sepsis-3        1317 11.2%    4690 39.8%  21.9%\n",
      "sofa>=2          561 4.8%    2362 20.0%  19.2%\n",
      "\n",
      "1\n",
      "sepsis_angus    2061 17.5%    1307 11.1%  61.2%\n",
      "sepsis_martin   1042 8.8%     692 5.9%  60.1%\n",
      "sepsis_explicit  751 6.4%     311 2.6%  70.7%\n",
      "suspicion_poe   3272 27.7%    3789 32.1%  46.3%\n",
      "sepsis-3        2894 24.5%    2890 24.5%  50.0%\n",
      "sofa>=2         3650 31.0%    5218 44.3%  41.2%\n"
     ]
    }
   ],
   "source": [
    "target_header = \"composite_outcome\"\n",
    "idx = (df['hospital_expire_flag']==1) | (df['icu_los']>=3)\n",
    "\n",
    "# make a confusion matrix with multiple scores in each square\n",
    "scores = ['sepsis_angus','sepsis_martin','sepsis_explicit','suspicion_poe','sepsis-3','sofa>=2']\n",
    "\n",
    "\n",
    "print('{:15s} {:12s} {:15s}'.format('0','dead/hi-LOS','alive/lo-LOS','outcome %'))\n",
    "\n",
    "for c in scores:\n",
    "    print('{:15s} {:4d} {:1.1f}%   {:5d} {:1.1f}%  {:1.1f}%'.format(\n",
    "            c,\n",
    "            np.sum( (df[c]!=1)&idx ),  np.sum( (df[c]!=1)&idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]!=1)&~idx ), np.sum( (df[c]!=1)&~idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]!=1)&idx )*100.0/np.sum(df[c]!=1)\n",
    "        ))\n",
    "print()\n",
    "print('1')\n",
    "for c in scores:\n",
    "    print('{:15s} {:4d} {:1.1f}%   {:5d} {:1.1f}%  {:1.1f}%'.format(\n",
    "            c,\n",
    "            np.sum( (df[c]==1)&idx ),  np.sum( (df[c]==1)&idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&~idx ), np.sum( (df[c]==1)&~idx )*100.0/df.shape[0],\n",
    "            np.sum( (df[c]==1)&idx )*100.0/np.sum(df[c]==1)\n",
    "        ))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Mortality: operating point statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Metric\n",
      "\n",
      "     \tSOFA    \tSIRS    \tqSOFA   \n",
      "TN   \t 2818\t\t 1536\t\t 4031\n",
      "FP   \t 7696\t\t 8978\t\t 6483\n",
      "FN   \t  105\t\t   66\t\t  263\n",
      "TP   \t 1172\t\t 1211\t\t 1014\n",
      "Sens \t92 [90, 93]\t95 [93, 96]\t79 [77, 82]\n",
      "Spec \t27 [26, 28]\t15 [14, 15]\t38 [37, 39]\n",
      "PPV  \t13 [13, 14]\t12 [11, 13]\t14 [13, 14]\n",
      "NPV  \t96 [96, 97]\t96 [95, 97]\t94 [93, 95]\n",
      "F1   \t 23.10   \t 21.12   \t 23.11   \n",
      "NTP  \t 9.94   \t 10.27   \t 8.60   \n",
      "NFP  \t 65.27   \t 76.14   \t 54.98   \n"
     ]
    }
   ],
   "source": [
    "target_header = \"hospital_expire_flag\"\n",
    "\n",
    "# sepsis3 defined as qSOFA >= 2 and SOFA >= 2\n",
    "yhat_dict = OrderedDict([['SOFA', df.sofa.values >= 2],\n",
    "                        ['SIRS', df.sirs.values >= 2],\n",
    "                        ['qSOFA', df.qsofa.values >= 2]])\n",
    "\n",
    "stats_all = su.get_op_stats(yhat_dict, df[target_header].values)\n",
    "\n",
    "su.print_op_stats(stats_all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Composite outcome: operating point statistics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Metric\n",
      "\n",
      "     \tSOFA    \tSIRS    \tqSOFA   \n",
      "TN   \t 2362\t\t 1302\t\t 3236\n",
      "FP   \t 5218\t\t 6278\t\t 4344\n",
      "FN   \t  561\t\t  300\t\t 1058\n",
      "TP   \t 3650\t\t 3911\t\t 3153\n",
      "Sens \t87 [86, 88]\t93 [92, 94]\t75 [74, 76]\n",
      "Spec \t31 [30, 32]\t17 [16, 18]\t43 [42, 44]\n",
      "PPV  \t41 [40, 42]\t38 [37, 39]\t42 [41, 43]\n",
      "NPV  \t81 [79, 82]\t81 [79, 83]\t75 [74, 77]\n",
      "F1   \t 55.81   \t 54.32   \t 53.86   \n",
      "NTP  \t 30.96   \t 33.17   \t 26.74   \n",
      "NFP  \t 44.25   \t 53.24   \t 36.84   \n"
     ]
    }
   ],
   "source": [
    "target_header = \"composite_outcome\"\n",
    "\n",
    "# sepsis3 defined as qSOFA >= 2 and SOFA >= 2\n",
    "yhat_dict = OrderedDict([['SOFA', df.sofa.values >= 2],\n",
    "                        ['SIRS', df.sirs.values >= 2],\n",
    "                        ['qSOFA', df.qsofa.values >= 2]])\n",
    "\n",
    "stats_all = su.get_op_stats(yhat_dict, df[target_header].values)\n",
    "\n",
    "su.print_op_stats(stats_all)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Severity of illness stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11791 (100.0%) first ICU stay for adults.\n",
      "10189 (86.4%)  with sirs >= 2\n",
      " 7497 (63.6%)  with qsofa >= 2\n",
      " 8868 (75.2%)  with sofa >= 2\n",
      " 5784 (49.1%)  with sepsis-3\n",
      " 3368 (28.6%)  with sepsis_angus\n",
      " 1734 (14.7%)  with sepsis_martin\n",
      " 1062 (9.0%)  with sepsis_explicit\n"
     ]
    }
   ],
   "source": [
    "print('{:5g} ({:3.1f}%) first ICU stay for adults.'.format(\n",
    "        df['icustay_id'].count(), 100))\n",
    "for c in ['sirs','qsofa','sofa','sepsis-3',\n",
    "          'sepsis_angus','sepsis_martin','sepsis_explicit']:\n",
    "    if df[c].max() == 1:\n",
    "        print('{:5g} ({:3.1f}%)  with {}'.format(\n",
    "                (df[c]==1).sum(),\n",
    "                (df[c]==1).sum()*100.0/df.shape[0], c))\n",
    "    else:\n",
    "        print('{:5g} ({:3.1f}%)  with {} >= 2'.format(\n",
    "                (df[c]>=2).sum(),\n",
    "                (df[c]>=2).sum()*100.0/df.shape[0], c))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "sepsis3-py3",
   "language": "python",
   "name": "sepsis3-py3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}