{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# R demo for the 2018 BHI & BSN Data Challenge\n", "\n", "This notebook provides a simple introduction to analysing the MIMIC-III database. It was created as a demonstrator for the [2018 BHI & BSN Data Challenge](https://mimic.physionet.org/events/bhibsn-challenge/), which explores the following question:\n", "\n", "> Are patients admitted to the intensive care unit (ICU) on a weekend more likely to die in the hospital than those admitted on a weekday?\n", "\n", "We have provided an example slide template for final presentations (`slide-template.pptx`) at: https://github.com/MIT-LCP/bhi-bsn-challenge. There is no obligation to use it!" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Background on MIMIC-III\n", "\n", "MIMIC-III (‘Medical Information Mart for Intensive Care’) is a large, single-center database comprising information relating to patients admitted to critical care units at a large tertiary care hospital. \n", "\n", "Data includes vital signs, medications, laboratory measurements, observations and notes charted by care providers, fluid balance, procedure codes, diagnostic codes, imaging reports, hospital length of stay, survival data, and more. \n", "\n", "For details, see: https://mimic.physionet.org/. The data is downloaded as 26 CSV files, which can then be loaded into a database system. Scripts for loading the data into Postgres are provided in the [MIMIC Code Repository](https://mimic.physionet.org/gettingstarted/dbsetup/). A demo dataset is also available: https://mimic.physionet.org/gettingstarted/demo/\n", "\n", "Points to note:\n", "\n", "- A patient-level shift has been applied to dates. Day of week is retained. \n", "- Patients aged >89 years on first admission have been reassigned an age of ~300 years.\n", "- Patients may have multiple hospital admissions. Each hospital admission may comprise multiple ICU stays (e.g. a patient may visit the ICU, leave for surgery, and then return to the ICU for recovery, all within a single hospital admission).\n", "\n", "If you need help getting set up with access to MIMIC-III, please contact `data-challenge@physionet.org`." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Connect to the MIMIC-III database\n", "\n", "The toughest part is often getting the database running and the software connecting to it. This is the way I usually do it:\n", "\n", "1. run `echo 'password' > .pwfile_`\n", " in a command prompt, where `password` is replaced with your user's postgres password\n", "2. Run this script in the same directory as this file\n", "\n", "The resulting chunk may install some R packages on your computer." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\n", "Attaching package: ‘dplyr’\n", "\n", "The following objects are masked from ‘package:stats’:\n", "\n", " filter, lag\n", "\n", "The following objects are masked from ‘package:base’:\n", "\n", " intersect, setdiff, setequal, union\n", "\n", "\n", "Attaching package: ‘dbplyr’\n", "\n", "The following objects are masked from ‘package:dplyr’:\n", "\n", " ident, sql\n", "\n", "Loading required package: DBI\n" ] }, { "data": { "text/plain": [ "" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "if(!(\"dplyr\" %in% installed.packages()) |!(\"dbplyr\" %in% installed.packages()) | !(\"RPostgreSQL\" %in% installed.packages())) { \n", " install.packages(c(\"dplyr\",\"dbplyr\",\"RPostgreSQL\"))\n", "}\n", "\n", " \n", "library(dplyr); library(dbplyr); library(RPostgreSQL); library(DBI)\n", "\n", "# Create a database connection\n", "user <- 'jraffa'\n", "host <- '127.0.0.1'\n", "dbname <- 'mimic'\n", "schema <- 'mimiciii,public'\n", "port <- 5455\n", "passwd <- readLines(\"~/.pwfile_\")\n", "\n", "pg_src <- src_postgres(dbname = dbname, \n", " host = host, port = port, \n", " user = user, password = passwd,\n", " options=paste0(\"-c search_path=\", schema))\n", "\n", "m <- dbDriver(\"PostgreSQL\")\n", "con <- dbConnect(m, user=user, password=passwd, dbname=dbname,host=host,port=port) # The trickiest parts\n", "dbSendStatement(con,paste0(\"SET search_path TO \",schema))" ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "collapsed": true }, "outputs": [], "source": [ "# Run query and assign the results to a DataFrame\n", "# Requires the icustay_detail view from:\n", "# https://github.com/MIT-LCP/mimic-code/tree/master/concepts/demographics\n", "\n", "\n", "# To get these views, the following may work (uncomment to run):\n", "## NOTE:\n", "## This will overwrite old views on your machine if you had already created them.\n", "## If you have them already, you likely do not need to run this code!.\n", "## This will take a while, possibly several hours, depending on the speed of your computer/network\n", "#\n", "#\n", "## if(!(\"MIMICutil\" %in% installed.packages())) { devtools::install_github(\"jraffa/MIMICutil\") }\n", "## \n", "## library(MIMICutil); library(RPostgreSQL)\n", "## v <- get_views(URLlist=\"https://raw.githubusercontent.com/jraffa/MIMICutil/master/mat_view_urls_working\",dplyrDB=pg_src,con=con)\n", "\n", "###\n", "\n", "query = \"\n", "WITH first_icu AS (\n", " SELECT i.subject_id, i.hadm_id, i.icustay_id, i.gender, i.admittime admittime_hospital, \n", " i.dischtime dischtime_hospital, i.los_hospital, i.age, i.admission_type, \n", " i.hospital_expire_flag, i.intime intime_icu, i.outtime outtime_icu, i.los_icu, \n", " s.first_careunit\n", " FROM icustay_detail i\n", " LEFT JOIN icustays s\n", " ON i.icustay_id = s.icustay_id\n", " WHERE i.hospstay_seq = 1\n", " AND i.icustay_seq = 1\n", " AND i.age >= 16\n", ")\n", "SELECT f.*, o.icustay_expire_flag, o.oasis, o.oasis_prob\n", "FROM first_icu f\n", "LEFT JOIN oasis o\n", "ON f.icustay_id = o.icustay_id;\n", "\"\n", "\n", "rs <- dbSendQuery(con,query)\n", "dat <- dbFetch(rs)\n", "\n", "# or using dplyr:\n", "# uncomment below\n", "#\n", "## icustay_detail_tbl <- tbl(pg_src,\"icustay_detail\")\n", "## icustays_tbl <- tbl(pg_src,\"icustays\")\n", "## oasis_tbl <- tbl(pg_src,\"oasis\")\n", "##\n", "## first_icu_tbl <- icustay_detail_tbl %>% select(subject_id,hadm_id, icustay_id, gender,admittime_hospital=admittime, \n", "## dischtime_hospital=dischtime, los_hospital, admission_type, \n", "## hospital_expire_flag,intime_icu=intime, outtime_icu=outtime, los_icu,age,hospstay_seq,icustay_seq) %>% \n", "## left_join(icustays_tbl %>% select(icustay_id, first_careunit) ,by=\"icustay_id\") %>% \n", "## filter(hospstay_seq==1,icustay_seq ==1,age>=16) %>% select(-hospstay_seq,-icustay_seq)\n", "\n", "## dat <- first_icu_tbl %>% \n", "## left_join(oasis_tbl %>% select(icustay_id,icustay_expire_flag, oasis, oasis_prob),by=\"icustay_id\") %>% \n", "## collect(n=Inf)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Check Data Extracted\n", "\n", "Always a good idea to inspect the data after you have extracted it. We will look at the first six patients (rows), and then check the number of rows, and get some summary statistics of the dataset." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\n", "
subject_idhadm_idicustay_idgenderadmittime_hospitaldischtime_hospitallos_hospitalageadmission_typehospital_expire_flagintime_icuouttime_iculos_icufirst_careuniticustay_expire_flagoasisoasis_prob
3 145834 211552 M 2101-10-20 19:08:002101-10-31 13:58:0010.7847 76.5268 EMERGENCY 0 2101-10-20 19:10:112101-10-26 20:43:096.0646 MICU 0 57 0.748927417
6 107064 228232 F 2175-05-30 07:15:002175-06-15 16:00:0016.3646 65.9407 ELECTIVE 0 2175-05-30 21:30:542175-06-03 13:39:543.6729 SICU 0 11 0.008391576
9 150750 220597 M 2149-11-09 13:06:002149-11-14 10:15:00 4.8813 41.7902 EMERGENCY 1 2149-11-09 13:07:022149-11-14 20:52:145.3231 MICU 1 35 0.152891534
11 194540 229441 F 2178-04-16 06:18:002178-05-11 19:00:0025.5292 50.1483 EMERGENCY 0 2178-04-16 06:19:322178-04-17 20:21:051.5844 SICU 0 29 0.077479229
13 143045 263738 F 2167-01-08 18:43:002167-01-15 15:15:00 6.8556 39.8661 EMERGENCY 0 2167-01-08 18:44:252167-01-12 10:43:313.6660 CCU 0 34 0.137098607
17 194023 277042 F 2134-12-27 07:15:002134-12-31 16:05:00 4.3681 47.4543 ELECTIVE 0 2134-12-27 16:21:482134-12-29 18:04:032.0710 CSRU 0 38 0.209225538
\n" ], "text/latex": [ "\\begin{tabular}{r|lllllllllllllllll}\n", " subject\\_id & hadm\\_id & icustay\\_id & gender & admittime\\_hospital & dischtime\\_hospital & los\\_hospital & age & admission\\_type & hospital\\_expire\\_flag & intime\\_icu & outtime\\_icu & los\\_icu & first\\_careunit & icustay\\_expire\\_flag & oasis & oasis\\_prob\\\\\n", "\\hline\n", "\t 3 & 145834 & 211552 & M & 2101-10-20 19:08:00 & 2101-10-31 13:58:00 & 10.7847 & 76.5268 & EMERGENCY & 0 & 2101-10-20 19:10:11 & 2101-10-26 20:43:09 & 6.0646 & MICU & 0 & 57 & 0.748927417 \\\\\n", "\t 6 & 107064 & 228232 & F & 2175-05-30 07:15:00 & 2175-06-15 16:00:00 & 16.3646 & 65.9407 & ELECTIVE & 0 & 2175-05-30 21:30:54 & 2175-06-03 13:39:54 & 3.6729 & SICU & 0 & 11 & 0.008391576 \\\\\n", "\t 9 & 150750 & 220597 & M & 2149-11-09 13:06:00 & 2149-11-14 10:15:00 & 4.8813 & 41.7902 & EMERGENCY & 1 & 2149-11-09 13:07:02 & 2149-11-14 20:52:14 & 5.3231 & MICU & 1 & 35 & 0.152891534 \\\\\n", "\t 11 & 194540 & 229441 & F & 2178-04-16 06:18:00 & 2178-05-11 19:00:00 & 25.5292 & 50.1483 & EMERGENCY & 0 & 2178-04-16 06:19:32 & 2178-04-17 20:21:05 & 1.5844 & SICU & 0 & 29 & 0.077479229 \\\\\n", "\t 13 & 143045 & 263738 & F & 2167-01-08 18:43:00 & 2167-01-15 15:15:00 & 6.8556 & 39.8661 & EMERGENCY & 0 & 2167-01-08 18:44:25 & 2167-01-12 10:43:31 & 3.6660 & CCU & 0 & 34 & 0.137098607 \\\\\n", "\t 17 & 194023 & 277042 & F & 2134-12-27 07:15:00 & 2134-12-31 16:05:00 & 4.3681 & 47.4543 & ELECTIVE & 0 & 2134-12-27 16:21:48 & 2134-12-29 18:04:03 & 2.0710 & CSRU & 0 & 38 & 0.209225538 \\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "subject_id | hadm_id | icustay_id | gender | admittime_hospital | dischtime_hospital | los_hospital | age | admission_type | hospital_expire_flag | intime_icu | outtime_icu | los_icu | first_careunit | icustay_expire_flag | oasis | oasis_prob | \n", "|---|---|---|---|---|---|\n", "| 3 | 145834 | 211552 | M | 2101-10-20 19:08:00 | 2101-10-31 13:58:00 | 10.7847 | 76.5268 | EMERGENCY | 0 | 2101-10-20 19:10:11 | 2101-10-26 20:43:09 | 6.0646 | MICU | 0 | 57 | 0.748927417 | \n", "| 6 | 107064 | 228232 | F | 2175-05-30 07:15:00 | 2175-06-15 16:00:00 | 16.3646 | 65.9407 | ELECTIVE | 0 | 2175-05-30 21:30:54 | 2175-06-03 13:39:54 | 3.6729 | SICU | 0 | 11 | 0.008391576 | \n", "| 9 | 150750 | 220597 | M | 2149-11-09 13:06:00 | 2149-11-14 10:15:00 | 4.8813 | 41.7902 | EMERGENCY | 1 | 2149-11-09 13:07:02 | 2149-11-14 20:52:14 | 5.3231 | MICU | 1 | 35 | 0.152891534 | \n", "| 11 | 194540 | 229441 | F | 2178-04-16 06:18:00 | 2178-05-11 19:00:00 | 25.5292 | 50.1483 | EMERGENCY | 0 | 2178-04-16 06:19:32 | 2178-04-17 20:21:05 | 1.5844 | SICU | 0 | 29 | 0.077479229 | \n", "| 13 | 143045 | 263738 | F | 2167-01-08 18:43:00 | 2167-01-15 15:15:00 | 6.8556 | 39.8661 | EMERGENCY | 0 | 2167-01-08 18:44:25 | 2167-01-12 10:43:31 | 3.6660 | CCU | 0 | 34 | 0.137098607 | \n", "| 17 | 194023 | 277042 | F | 2134-12-27 07:15:00 | 2134-12-31 16:05:00 | 4.3681 | 47.4543 | ELECTIVE | 0 | 2134-12-27 16:21:48 | 2134-12-29 18:04:03 | 2.0710 | CSRU | 0 | 38 | 0.209225538 | \n", "\n", "\n" ], "text/plain": [ " subject_id hadm_id icustay_id gender admittime_hospital dischtime_hospital \n", "1 3 145834 211552 M 2101-10-20 19:08:00 2101-10-31 13:58:00\n", "2 6 107064 228232 F 2175-05-30 07:15:00 2175-06-15 16:00:00\n", "3 9 150750 220597 M 2149-11-09 13:06:00 2149-11-14 10:15:00\n", "4 11 194540 229441 F 2178-04-16 06:18:00 2178-05-11 19:00:00\n", "5 13 143045 263738 F 2167-01-08 18:43:00 2167-01-15 15:15:00\n", "6 17 194023 277042 F 2134-12-27 07:15:00 2134-12-31 16:05:00\n", " los_hospital age admission_type hospital_expire_flag intime_icu \n", "1 10.7847 76.5268 EMERGENCY 0 2101-10-20 19:10:11\n", "2 16.3646 65.9407 ELECTIVE 0 2175-05-30 21:30:54\n", "3 4.8813 41.7902 EMERGENCY 1 2149-11-09 13:07:02\n", "4 25.5292 50.1483 EMERGENCY 0 2178-04-16 06:19:32\n", "5 6.8556 39.8661 EMERGENCY 0 2167-01-08 18:44:25\n", "6 4.3681 47.4543 ELECTIVE 0 2134-12-27 16:21:48\n", " outtime_icu los_icu first_careunit icustay_expire_flag oasis\n", "1 2101-10-26 20:43:09 6.0646 MICU 0 57 \n", "2 2175-06-03 13:39:54 3.6729 SICU 0 11 \n", "3 2149-11-14 20:52:14 5.3231 MICU 1 35 \n", "4 2178-04-17 20:21:05 1.5844 SICU 0 29 \n", "5 2167-01-12 10:43:31 3.6660 CCU 0 34 \n", "6 2134-12-29 18:04:03 2.0710 CSRU 0 38 \n", " oasis_prob \n", "1 0.748927417\n", "2 0.008391576\n", "3 0.152891534\n", "4 0.077479229\n", "5 0.137098607\n", "6 0.209225538" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "38557" ], "text/latex": [ "38557" ], "text/markdown": [ "38557" ], "text/plain": [ "[1] 38557" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [ " subject_id hadm_id icustay_id gender \n", " Min. : 3 Min. :100001 Min. :200003 Length:38557 \n", " 1st Qu.:13980 1st Qu.:124930 1st Qu.:225240 Class :character \n", " Median :27912 Median :150093 Median :250275 Mode :character \n", " Mean :38266 Mean :150038 Mean :250222 \n", " 3rd Qu.:62965 3rd Qu.:175223 3rd Qu.:275288 \n", " Max. :99999 Max. :199999 Max. :299999 \n", " \n", " admittime_hospital dischtime_hospital los_hospital \n", " Min. :2100-06-07 19:59:00 Min. :2100-06-09 17:09:00 Min. : -0.9451 \n", " 1st Qu.:2125-09-28 09:42:00 1st Qu.:2125-10-08 17:35:00 1st Qu.: 4.0479 \n", " Median :2150-12-27 04:35:00 Median :2151-01-04 15:50:00 Median : 6.8938 \n", " Mean :2150-12-13 00:38:20 Mean :2150-12-22 22:10:28 Mean : 9.8973 \n", " 3rd Qu.:2176-05-16 07:15:00 3rd Qu.:2176-05-23 14:40:00 3rd Qu.: 11.8833 \n", " Max. :2208-08-19 07:15:00 Max. :2208-08-25 14:59:00 Max. :294.6604 \n", " \n", " age admission_type hospital_expire_flag\n", " Min. : 16.02 Length:38557 Min. :0.0000 \n", " 1st Qu.: 52.32 Class :character 1st Qu.:0.0000 \n", " Median : 65.67 Mode :character Median :0.0000 \n", " Mean : 74.57 Mean :0.1147 \n", " 3rd Qu.: 77.90 3rd Qu.:0.0000 \n", " Max. :310.28 Max. :1.0000 \n", " \n", " intime_icu outtime_icu los_icu \n", " Min. :2100-06-07 20:00:22 Min. :2100-06-08 14:59:31 Min. : 0.0001 \n", " 1st Qu.:2125-09-28 09:46:09 1st Qu.:2125-10-02 14:26:16 1st Qu.: 1.1863 \n", " Median :2150-12-27 11:21:41 Median :2150-12-30 01:03:44 Median : 2.0952 \n", " Mean :2150-12-14 01:20:00 Mean :2150-12-18 11:17:37 Mean : 4.0587 \n", " 3rd Qu.:2176-05-17 09:06:35 3rd Qu.:2176-05-20 12:42:46 3rd Qu.: 4.0921 \n", " Max. :2208-08-19 13:03:37 Max. :2208-08-21 05:38:09 Max. :153.9280 \n", " NA's :2 NA's :2 \n", " first_careunit icustay_expire_flag oasis oasis_prob \n", " Length:38557 Min. :0.00000 Min. : 3.00 Min. :0.003042 \n", " Class :character 1st Qu.:0.00000 1st Qu.:25.00 1st Qu.:0.048012 \n", " Mode :character Median :0.00000 Median :31.00 Median :0.097783 \n", " Mean :0.08284 Mean :31.16 Mean :0.143443 \n", " 3rd Qu.:0.00000 3rd Qu.:37.00 3rd Qu.:0.188911 \n", " Max. :1.00000 Max. :70.00 Max. :0.939936 \n", " " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Have a look at the dataset:\n", "head(dat)\n", "\n", "nrow(dat) #38557\n", "\n", "summary(dat)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Add day of week to DataFrame\n", "\n", "If we are going to examine the weekend effect, we need to pull this out of the dataset, as you can see, all we have above are dates. We will define a weekend, as anytime between Saturday (00:00:00) until Sunday (23:59:59). The dates above are shifted, and that's why they look odd, but they are matched on the day of week, so this aspect is preserved." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "\n", "Attaching package: ‘lubridate’\n", "\n", "The following object is masked from ‘package:base’:\n", "\n", " date\n", "\n" ] }, { "data": { "text/plain": [ "\n", " 1 2 3 4 5 6 7 \n", "3960 6097 6141 5985 5876 6263 4235 " ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [ "\n", " Sunday Monday Tuesday Wednesday Thursday Friday Saturday \n", " 3960 6097 6141 5985 5876 6263 4235 " ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [ "\n", "FALSE TRUE \n", "30362 8195 " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "if(!(\"lubridate\" %in% installed.packages())) { install.packages(\"lubridate\")}\n", "\n", "library(lubridate)\n", "\n", "dat$dow <- as.factor(wday(dat$intime_icu ))\n", "table(dat$dow)\n", "# Convert to text levels\n", "levels(dat$dow) <- c(\"Sunday\",\"Monday\",\"Tuesday\",\"Wednesday\",\"Thursday\",\"Friday\",\"Saturday\",\"Sunday\")\n", "table(dat$dow)\n", "\n", "dat$weekend <- dat$dow %in% c(\"Sunday\",\"Saturday\")\n", "table(dat$weekend)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Produce some Summary Statistics by DOW and Weekday vs. Weekend\n", "\n", "Next, it's good to look at some basic summaries of the data. We will compute simple averages and percentages/counts for each of the variables we have extracted, and look at it by day of week (`dow`) and weekend (`weekend`).\n" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\n", "
variable_namelevelSundayMondayTuesdayWednesdayThursdayFridaySaturdayptest
n 3960 6097 6141 5985 5876 6263 4235
gender (%) F 1736 (43.8) 2559 (42.0) 2671 (43.5) 2636 (44.0) 2603 (44.3) 2662 (42.5) 1857 (43.8) 0.097
M 2224 (56.2) 3538 (58.0) 3470 (56.5) 3349 (56.0) 3273 (55.7) 3601 (57.5) 2378 (56.2)
los_hospital (mean (sd)) 9.82 (10.70) 9.71 (9.92) 9.86 (10.58) 9.77 (10.22) 9.88 (10.90) 10.23 (11.61) 10.00 (10.84) 0.141
age (mean (sd)) 75.26 (60.66) 73.13 (51.37) 75.48 (55.22) 74.16 (53.88) 75.51 (55.46) 74.56 (53.22) 73.92 (58.58) 0.167
admission_type (%) ELECTIVE 101 ( 2.6) 1265 (20.7) 1292 (21.0) 1243 (20.8) 999 (17.0) 1016 (16.2) 162 ( 3.8) <0.001
EMERGENCY 3681 (93.0) 4687 (76.9) 4704 (76.6) 4600 (76.9) 4746 (80.8) 5118 (81.7) 3852 (91.0)
URGENT 178 ( 4.5) 145 ( 2.4) 145 ( 2.4) 142 ( 2.4) 131 ( 2.2) 129 ( 2.1) 221 ( 5.2)
hospital_expire_flag (%) 0 3388 (85.6) 5468 (89.7) 5491 (89.4) 5350 (89.4) 5202 (88.5) 5576 (89.0) 3658 (86.4) <0.001
1 572 (14.4) 629 (10.3) 650 (10.6) 635 (10.6) 674 (11.5) 687 (11.0) 577 (13.6)
los_icu (mean (sd)) 4.44 (6.29) 3.83 (5.52) 3.80 (5.65) 4.04 (5.94) 3.96 (6.12) 4.15 (6.19) 4.42 (6.58) <0.001
icustay_expire_flag (%) 0 3548 (89.6) 5650 (92.7) 5673 (92.4) 5514 (92.1) 5399 (91.9) 5768 (92.1) 3811 (90.0) <0.001
1 412 (10.4) 447 ( 7.3) 468 ( 7.6) 471 ( 7.9) 477 ( 8.1) 495 ( 7.9) 424 (10.0)
oasis (mean (sd)) 32.07 (9.02) 31.21 (8.77) 30.90 (8.77) 30.58 (9.11) 31.10 (8.82) 31.17 (9.02) 31.49 (9.27) <0.001
oasis_prob (mean (sd)) 0.16 (0.15) 0.14 (0.14) 0.14 (0.14) 0.14 (0.14) 0.14 (0.14) 0.14 (0.14) 0.15 (0.15) <0.001
first_careunit (%) CCU 621 (15.7) 918 (15.1) 919 (15.0) 851 (14.2) 850 (14.5) 838 (13.4) 695 (16.4) <0.001
CSRU 194 ( 4.9) 1632 (26.8) 1575 (25.6) 1268 (21.2) 1282 (21.8) 1416 (22.6) 237 ( 5.6)
MICU 1706 (43.1) 1940 (31.8) 2019 (32.9) 2020 (33.8) 2020 (34.4) 2139 (34.2) 1765 (41.7)
SICU 743 (18.8) 865 (14.2) 933 (15.2) 1038 (17.3) 996 (17.0) 1044 (16.7) 743 (17.5)
TSICU 696 (17.6) 742 (12.2) 695 (11.3) 808 (13.5) 728 (12.4) 826 (13.2) 795 (18.8)
\n" ], "text/latex": [ "\\begin{tabular}{r|lllllllllll}\n", " variable\\_name & level & Sunday & Monday & Tuesday & Wednesday & Thursday & Friday & Saturday & p & test\\\\\n", "\\hline\n", "\t n & & 3960 & 6097 & 6141 & 5985 & 5876 & 6263 & 4235 & & \\\\\n", "\t gender (\\%) & F & 1736 (43.8) & 2559 (42.0) & 2671 (43.5) & 2636 (44.0) & 2603 (44.3) & 2662 (42.5) & 1857 (43.8) & 0.097 & \\\\\n", "\t & M & 2224 (56.2) & 3538 (58.0) & 3470 (56.5) & 3349 (56.0) & 3273 (55.7) & 3601 (57.5) & 2378 (56.2) & & \\\\\n", "\t los\\_hospital (mean (sd)) & & 9.82 (10.70) & 9.71 (9.92) & 9.86 (10.58) & 9.77 (10.22) & 9.88 (10.90) & 10.23 (11.61) & 10.00 (10.84) & 0.141 & \\\\\n", "\t age (mean (sd)) & & 75.26 (60.66) & 73.13 (51.37) & 75.48 (55.22) & 74.16 (53.88) & 75.51 (55.46) & 74.56 (53.22) & 73.92 (58.58) & 0.167 & \\\\\n", "\t admission\\_type (\\%) & ELECTIVE & 101 ( 2.6) & 1265 (20.7) & 1292 (21.0) & 1243 (20.8) & 999 (17.0) & 1016 (16.2) & 162 ( 3.8) & <0.001 & \\\\\n", "\t & EMERGENCY & 3681 (93.0) & 4687 (76.9) & 4704 (76.6) & 4600 (76.9) & 4746 (80.8) & 5118 (81.7) & 3852 (91.0) & & \\\\\n", "\t & URGENT & 178 ( 4.5) & 145 ( 2.4) & 145 ( 2.4) & 142 ( 2.4) & 131 ( 2.2) & 129 ( 2.1) & 221 ( 5.2) & & \\\\\n", "\t hospital\\_expire\\_flag (\\%) & 0 & 3388 (85.6) & 5468 (89.7) & 5491 (89.4) & 5350 (89.4) & 5202 (88.5) & 5576 (89.0) & 3658 (86.4) & <0.001 & \\\\\n", "\t & 1 & 572 (14.4) & 629 (10.3) & 650 (10.6) & 635 (10.6) & 674 (11.5) & 687 (11.0) & 577 (13.6) & & \\\\\n", "\t los\\_icu (mean (sd)) & & 4.44 (6.29) & 3.83 (5.52) & 3.80 (5.65) & 4.04 (5.94) & 3.96 (6.12) & 4.15 (6.19) & 4.42 (6.58) & <0.001 & \\\\\n", "\t icustay\\_expire\\_flag (\\%) & 0 & 3548 (89.6) & 5650 (92.7) & 5673 (92.4) & 5514 (92.1) & 5399 (91.9) & 5768 (92.1) & 3811 (90.0) & <0.001 & \\\\\n", "\t & 1 & 412 (10.4) & 447 ( 7.3) & 468 ( 7.6) & 471 ( 7.9) & 477 ( 8.1) & 495 ( 7.9) & 424 (10.0) & & \\\\\n", "\t oasis (mean (sd)) & & 32.07 (9.02) & 31.21 (8.77) & 30.90 (8.77) & 30.58 (9.11) & 31.10 (8.82) & 31.17 (9.02) & 31.49 (9.27) & <0.001 & \\\\\n", "\t oasis\\_prob (mean (sd)) & & 0.16 (0.15) & 0.14 (0.14) & 0.14 (0.14) & 0.14 (0.14) & 0.14 (0.14) & 0.14 (0.14) & 0.15 (0.15) & <0.001 & \\\\\n", "\t first\\_careunit (\\%) & CCU & 621 (15.7) & 918 (15.1) & 919 (15.0) & 851 (14.2) & 850 (14.5) & 838 (13.4) & 695 (16.4) & <0.001 & \\\\\n", "\t & CSRU & 194 ( 4.9) & 1632 (26.8) & 1575 (25.6) & 1268 (21.2) & 1282 (21.8) & 1416 (22.6) & 237 ( 5.6) & & \\\\\n", "\t & MICU & 1706 (43.1) & 1940 (31.8) & 2019 (32.9) & 2020 (33.8) & 2020 (34.4) & 2139 (34.2) & 1765 (41.7) & & \\\\\n", "\t & SICU & 743 (18.8) & 865 (14.2) & 933 (15.2) & 1038 (17.3) & 996 (17.0) & 1044 (16.7) & 743 (17.5) & & \\\\\n", "\t & TSICU & 696 (17.6) & 742 (12.2) & 695 (11.3) & 808 (13.5) & 728 (12.4) & 826 (13.2) & 795 (18.8) & & \\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "variable_name | level | Sunday | Monday | Tuesday | Wednesday | Thursday | Friday | Saturday | p | test | \n", "|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n", "| n | | 3960 | 6097 | 6141 | 5985 | 5876 | 6263 | 4235 | | | \n", "| gender (%) | F | 1736 (43.8) | 2559 (42.0) | 2671 (43.5) | 2636 (44.0) | 2603 (44.3) | 2662 (42.5) | 1857 (43.8) | 0.097 | | \n", "| | M | 2224 (56.2) | 3538 (58.0) | 3470 (56.5) | 3349 (56.0) | 3273 (55.7) | 3601 (57.5) | 2378 (56.2) | | | \n", "| los_hospital (mean (sd)) | | 9.82 (10.70) | 9.71 (9.92) | 9.86 (10.58) | 9.77 (10.22) | 9.88 (10.90) | 10.23 (11.61) | 10.00 (10.84) | 0.141 | | \n", "| age (mean (sd)) | | 75.26 (60.66) | 73.13 (51.37) | 75.48 (55.22) | 74.16 (53.88) | 75.51 (55.46) | 74.56 (53.22) | 73.92 (58.58) | 0.167 | | \n", "| admission_type (%) | ELECTIVE | 101 ( 2.6) | 1265 (20.7) | 1292 (21.0) | 1243 (20.8) | 999 (17.0) | 1016 (16.2) | 162 ( 3.8) | <0.001 | | \n", "| | EMERGENCY | 3681 (93.0) | 4687 (76.9) | 4704 (76.6) | 4600 (76.9) | 4746 (80.8) | 5118 (81.7) | 3852 (91.0) | | | \n", "| | URGENT | 178 ( 4.5) | 145 ( 2.4) | 145 ( 2.4) | 142 ( 2.4) | 131 ( 2.2) | 129 ( 2.1) | 221 ( 5.2) | | | \n", "| hospital_expire_flag (%) | 0 | 3388 (85.6) | 5468 (89.7) | 5491 (89.4) | 5350 (89.4) | 5202 (88.5) | 5576 (89.0) | 3658 (86.4) | <0.001 | | \n", "| | 1 | 572 (14.4) | 629 (10.3) | 650 (10.6) | 635 (10.6) | 674 (11.5) | 687 (11.0) | 577 (13.6) | | | \n", "| los_icu (mean (sd)) | | 4.44 (6.29) | 3.83 (5.52) | 3.80 (5.65) | 4.04 (5.94) | 3.96 (6.12) | 4.15 (6.19) | 4.42 (6.58) | <0.001 | | \n", "| icustay_expire_flag (%) | 0 | 3548 (89.6) | 5650 (92.7) | 5673 (92.4) | 5514 (92.1) | 5399 (91.9) | 5768 (92.1) | 3811 (90.0) | <0.001 | | \n", "| | 1 | 412 (10.4) | 447 ( 7.3) | 468 ( 7.6) | 471 ( 7.9) | 477 ( 8.1) | 495 ( 7.9) | 424 (10.0) | | | \n", "| oasis (mean (sd)) | | 32.07 (9.02) | 31.21 (8.77) | 30.90 (8.77) | 30.58 (9.11) | 31.10 (8.82) | 31.17 (9.02) | 31.49 (9.27) | <0.001 | | \n", "| oasis_prob (mean (sd)) | | 0.16 (0.15) | 0.14 (0.14) | 0.14 (0.14) | 0.14 (0.14) | 0.14 (0.14) | 0.14 (0.14) | 0.15 (0.15) | <0.001 | | \n", "| first_careunit (%) | CCU | 621 (15.7) | 918 (15.1) | 919 (15.0) | 851 (14.2) | 850 (14.5) | 838 (13.4) | 695 (16.4) | <0.001 | | \n", "| | CSRU | 194 ( 4.9) | 1632 (26.8) | 1575 (25.6) | 1268 (21.2) | 1282 (21.8) | 1416 (22.6) | 237 ( 5.6) | | | \n", "| | MICU | 1706 (43.1) | 1940 (31.8) | 2019 (32.9) | 2020 (33.8) | 2020 (34.4) | 2139 (34.2) | 1765 (41.7) | | | \n", "| | SICU | 743 (18.8) | 865 (14.2) | 933 (15.2) | 1038 (17.3) | 996 (17.0) | 1044 (16.7) | 743 (17.5) | | | \n", "| | TSICU | 696 (17.6) | 742 (12.2) | 695 (11.3) | 808 (13.5) | 728 (12.4) | 826 (13.2) | 795 (18.8) | | | \n", "\n", "\n" ], "text/plain": [ " variable_name level Sunday Monday Tuesday \n", "1 n 3960 6097 6141 \n", "2 gender (%) F 1736 (43.8) 2559 (42.0) 2671 (43.5) \n", "3 M 2224 (56.2) 3538 (58.0) 3470 (56.5) \n", "4 los_hospital (mean (sd)) 9.82 (10.70) 9.71 (9.92) 9.86 (10.58)\n", "5 age (mean (sd)) 75.26 (60.66) 73.13 (51.37) 75.48 (55.22)\n", "6 admission_type (%) ELECTIVE 101 ( 2.6) 1265 (20.7) 1292 (21.0) \n", "7 EMERGENCY 3681 (93.0) 4687 (76.9) 4704 (76.6) \n", "8 URGENT 178 ( 4.5) 145 ( 2.4) 145 ( 2.4) \n", "9 hospital_expire_flag (%) 0 3388 (85.6) 5468 (89.7) 5491 (89.4) \n", "10 1 572 (14.4) 629 (10.3) 650 (10.6) \n", "11 los_icu (mean (sd)) 4.44 (6.29) 3.83 (5.52) 3.80 (5.65) \n", "12 icustay_expire_flag (%) 0 3548 (89.6) 5650 (92.7) 5673 (92.4) \n", "13 1 412 (10.4) 447 ( 7.3) 468 ( 7.6) \n", "14 oasis (mean (sd)) 32.07 (9.02) 31.21 (8.77) 30.90 (8.77) \n", "15 oasis_prob (mean (sd)) 0.16 (0.15) 0.14 (0.14) 0.14 (0.14) \n", "16 first_careunit (%) CCU 621 (15.7) 918 (15.1) 919 (15.0) \n", "17 CSRU 194 ( 4.9) 1632 (26.8) 1575 (25.6) \n", "18 MICU 1706 (43.1) 1940 (31.8) 2019 (32.9) \n", "19 SICU 743 (18.8) 865 (14.2) 933 (15.2) \n", "20 TSICU 696 (17.6) 742 (12.2) 695 (11.3) \n", " Wednesday Thursday Friday Saturday p test\n", "1 5985 5876 6263 4235 \n", "2 2636 (44.0) 2603 (44.3) 2662 (42.5) 1857 (43.8) 0.097 \n", "3 3349 (56.0) 3273 (55.7) 3601 (57.5) 2378 (56.2) \n", "4 9.77 (10.22) 9.88 (10.90) 10.23 (11.61) 10.00 (10.84) 0.141 \n", "5 74.16 (53.88) 75.51 (55.46) 74.56 (53.22) 73.92 (58.58) 0.167 \n", "6 1243 (20.8) 999 (17.0) 1016 (16.2) 162 ( 3.8) <0.001 \n", "7 4600 (76.9) 4746 (80.8) 5118 (81.7) 3852 (91.0) \n", "8 142 ( 2.4) 131 ( 2.2) 129 ( 2.1) 221 ( 5.2) \n", "9 5350 (89.4) 5202 (88.5) 5576 (89.0) 3658 (86.4) <0.001 \n", "10 635 (10.6) 674 (11.5) 687 (11.0) 577 (13.6) \n", "11 4.04 (5.94) 3.96 (6.12) 4.15 (6.19) 4.42 (6.58) <0.001 \n", "12 5514 (92.1) 5399 (91.9) 5768 (92.1) 3811 (90.0) <0.001 \n", "13 471 ( 7.9) 477 ( 8.1) 495 ( 7.9) 424 (10.0) \n", "14 30.58 (9.11) 31.10 (8.82) 31.17 (9.02) 31.49 (9.27) <0.001 \n", "15 0.14 (0.14) 0.14 (0.14) 0.14 (0.14) 0.15 (0.15) <0.001 \n", "16 851 (14.2) 850 (14.5) 838 (13.4) 695 (16.4) <0.001 \n", "17 1268 (21.2) 1282 (21.8) 1416 (22.6) 237 ( 5.6) \n", "18 2020 (33.8) 2020 (34.4) 2139 (34.2) 1765 (41.7) \n", "19 1038 (17.3) 996 (17.0) 1044 (16.7) 743 (17.5) \n", "20 808 (13.5) 728 (12.4) 826 (13.2) 795 (18.8) " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "#If this fails, you may have to install pkg-config and libnlopt-dev on your system\n", "# e.g., in Ubuntu/Debian:\n", "# sudo apt-get install pkg-config libnlopt-dev\n", "\n", "if(!(\"tableone\" %in% installed.packages())) { install.packages(\"tableone\")}\n", "\n", "library(tableone)\n", "\n", "vars <- c('gender', 'los_hospital', 'age', 'admission_type', 'hospital_expire_flag', \n", " 'los_icu','icustay_expire_flag', 'oasis', 'oasis_prob', 'first_careunit')\n", "\n", "group_var <- \"dow\"\n", "\n", "CreateTableOne(dat,vars=vars,strata=group_var,factorVars=c(\"hospital_expire_flag\",\"icustay_expire_flag\")) %>% \n", " print(\n", " printToggle = FALSE,\n", " showAllLevels = TRUE,\n", " cramVars = \"kon\"\n", ") %>% \n", "{data.frame(\n", " variable_name = gsub(\" \", \" \", rownames(.), fixed = TRUE), ., \n", " row.names = NULL, \n", " check.names = FALSE, \n", " stringsAsFactors = FALSE)} \n" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\t\n", "\n", "
variable_namelevelFALSETRUEptest
n 30362 8195
gender (%) F 13131 (43.2) 3593 (43.8) 0.341
M 17231 (56.8) 4602 (56.2)
los_hospital (mean (sd)) 9.89 (10.67) 9.91 (10.77) 0.879
age (mean (sd)) 74.56 (53.84) 74.57 (59.59) 0.997
admission_type (%) ELECTIVE 5815 (19.2) 263 ( 3.2) <0.001
EMERGENCY 23855 (78.6) 7533 (91.9)
URGENT 692 ( 2.3) 399 ( 4.9)
hospital_expire_flag (%) 0 27087 (89.2) 7046 (86.0) <0.001
1 3275 (10.8) 1149 (14.0)
los_icu (mean (sd)) 3.96 (5.89) 4.43 (6.44) <0.001
icustay_expire_flag (%) 0 28004 (92.2) 7359 (89.8) <0.001
1 2358 ( 7.8) 836 (10.2)
oasis (mean (sd)) 31.00 (8.90) 31.77 (9.15) <0.001
oasis_prob (mean (sd)) 0.14 (0.14) 0.15 (0.15) <0.001
first_careunit (%) CCU 4376 (14.4) 1316 (16.1) <0.001
CSRU 7173 (23.6) 431 ( 5.3)
MICU 10138 (33.4) 3471 (42.4)
SICU 4876 (16.1) 1486 (18.1)
TSICU 3799 (12.5) 1491 (18.2)
\n" ], "text/latex": [ "\\begin{tabular}{r|llllll}\n", " variable\\_name & level & FALSE & TRUE & p & test\\\\\n", "\\hline\n", "\t n & & 30362 & 8195 & & \\\\\n", "\t gender (\\%) & F & 13131 (43.2) & 3593 (43.8) & 0.341 & \\\\\n", "\t & M & 17231 (56.8) & 4602 (56.2) & & \\\\\n", "\t los\\_hospital (mean (sd)) & & 9.89 (10.67) & 9.91 (10.77) & 0.879 & \\\\\n", "\t age (mean (sd)) & & 74.56 (53.84) & 74.57 (59.59) & 0.997 & \\\\\n", "\t admission\\_type (\\%) & ELECTIVE & 5815 (19.2) & 263 ( 3.2) & <0.001 & \\\\\n", "\t & EMERGENCY & 23855 (78.6) & 7533 (91.9) & & \\\\\n", "\t & URGENT & 692 ( 2.3) & 399 ( 4.9) & & \\\\\n", "\t hospital\\_expire\\_flag (\\%) & 0 & 27087 (89.2) & 7046 (86.0) & <0.001 & \\\\\n", "\t & 1 & 3275 (10.8) & 1149 (14.0) & & \\\\\n", "\t los\\_icu (mean (sd)) & & 3.96 (5.89) & 4.43 (6.44) & <0.001 & \\\\\n", "\t icustay\\_expire\\_flag (\\%) & 0 & 28004 (92.2) & 7359 (89.8) & <0.001 & \\\\\n", "\t & 1 & 2358 ( 7.8) & 836 (10.2) & & \\\\\n", "\t oasis (mean (sd)) & & 31.00 (8.90) & 31.77 (9.15) & <0.001 & \\\\\n", "\t oasis\\_prob (mean (sd)) & & 0.14 (0.14) & 0.15 (0.15) & <0.001 & \\\\\n", "\t first\\_careunit (\\%) & CCU & 4376 (14.4) & 1316 (16.1) & <0.001 & \\\\\n", "\t & CSRU & 7173 (23.6) & 431 ( 5.3) & & \\\\\n", "\t & MICU & 10138 (33.4) & 3471 (42.4) & & \\\\\n", "\t & SICU & 4876 (16.1) & 1486 (18.1) & & \\\\\n", "\t & TSICU & 3799 (12.5) & 1491 (18.2) & & \\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "variable_name | level | FALSE | TRUE | p | test | \n", "|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|\n", "| n | | 30362 | 8195 | | | \n", "| gender (%) | F | 13131 (43.2) | 3593 (43.8) | 0.341 | | \n", "| | M | 17231 (56.8) | 4602 (56.2) | | | \n", "| los_hospital (mean (sd)) | | 9.89 (10.67) | 9.91 (10.77) | 0.879 | | \n", "| age (mean (sd)) | | 74.56 (53.84) | 74.57 (59.59) | 0.997 | | \n", "| admission_type (%) | ELECTIVE | 5815 (19.2) | 263 ( 3.2) | <0.001 | | \n", "| | EMERGENCY | 23855 (78.6) | 7533 (91.9) | | | \n", "| | URGENT | 692 ( 2.3) | 399 ( 4.9) | | | \n", "| hospital_expire_flag (%) | 0 | 27087 (89.2) | 7046 (86.0) | <0.001 | | \n", "| | 1 | 3275 (10.8) | 1149 (14.0) | | | \n", "| los_icu (mean (sd)) | | 3.96 (5.89) | 4.43 (6.44) | <0.001 | | \n", "| icustay_expire_flag (%) | 0 | 28004 (92.2) | 7359 (89.8) | <0.001 | | \n", "| | 1 | 2358 ( 7.8) | 836 (10.2) | | | \n", "| oasis (mean (sd)) | | 31.00 (8.90) | 31.77 (9.15) | <0.001 | | \n", "| oasis_prob (mean (sd)) | | 0.14 (0.14) | 0.15 (0.15) | <0.001 | | \n", "| first_careunit (%) | CCU | 4376 (14.4) | 1316 (16.1) | <0.001 | | \n", "| | CSRU | 7173 (23.6) | 431 ( 5.3) | | | \n", "| | MICU | 10138 (33.4) | 3471 (42.4) | | | \n", "| | SICU | 4876 (16.1) | 1486 (18.1) | | | \n", "| | TSICU | 3799 (12.5) | 1491 (18.2) | | | \n", "\n", "\n" ], "text/plain": [ " variable_name level FALSE TRUE p test\n", "1 n 30362 8195 \n", "2 gender (%) F 13131 (43.2) 3593 (43.8) 0.341 \n", "3 M 17231 (56.8) 4602 (56.2) \n", "4 los_hospital (mean (sd)) 9.89 (10.67) 9.91 (10.77) 0.879 \n", "5 age (mean (sd)) 74.56 (53.84) 74.57 (59.59) 0.997 \n", "6 admission_type (%) ELECTIVE 5815 (19.2) 263 ( 3.2) <0.001 \n", "7 EMERGENCY 23855 (78.6) 7533 (91.9) \n", "8 URGENT 692 ( 2.3) 399 ( 4.9) \n", "9 hospital_expire_flag (%) 0 27087 (89.2) 7046 (86.0) <0.001 \n", "10 1 3275 (10.8) 1149 (14.0) \n", "11 los_icu (mean (sd)) 3.96 (5.89) 4.43 (6.44) <0.001 \n", "12 icustay_expire_flag (%) 0 28004 (92.2) 7359 (89.8) <0.001 \n", "13 1 2358 ( 7.8) 836 (10.2) \n", "14 oasis (mean (sd)) 31.00 (8.90) 31.77 (9.15) <0.001 \n", "15 oasis_prob (mean (sd)) 0.14 (0.14) 0.15 (0.15) <0.001 \n", "16 first_careunit (%) CCU 4376 (14.4) 1316 (16.1) <0.001 \n", "17 CSRU 7173 (23.6) 431 ( 5.3) \n", "18 MICU 10138 (33.4) 3471 (42.4) \n", "19 SICU 4876 (16.1) 1486 (18.1) \n", "20 TSICU 3799 (12.5) 1491 (18.2) " ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "vars <- c('gender', 'los_hospital', 'age', 'admission_type', 'hospital_expire_flag', \n", " 'los_icu','icustay_expire_flag', 'oasis', 'oasis_prob', 'first_careunit')\n", "\n", "group_var <- \"weekend\"\n", "\n", "CreateTableOne(dat,vars=vars,strata=group_var,factorVars=c(\"hospital_expire_flag\",\"icustay_expire_flag\")) %>% \n", " print(\n", " printToggle = FALSE,\n", " showAllLevels = TRUE,\n", " cramVars = \"kon\"\n", ") %>% \n", "{data.frame(\n", " variable_name = gsub(\" \", \" \", rownames(.), fixed = TRUE), ., \n", " row.names = NULL, \n", " check.names = FALSE, \n", " stringsAsFactors = FALSE)} " ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It looks like there's a higher rate of hospital mortality (14.0% vs 10.8%) and ICU mortality (10.2% vs 7.8%) on weekends when compared to weekdays. There are also statistically significant differences between several other important variables, including: admission type, disease severity (OASIS), and the patient's first care unit, suggesting that these groups may be fundamentally different in some way. Let's explore this a little further.\n" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": {}, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94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\nEOwAAAAALALBDgAAAMAi/j/M9r7zVO0kWAAAAABJRU5ErkJggg==", "text/plain": [ "plot without title" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "library(ggplot2)\n", "\n", "dat %>% group_by(admission_type,dow) %>% \n", " summarise(n=n(),mortRate=mean(hospital_expire_flag==1),\n", " LL95 = mortRate - qnorm(0.975)*sqrt(mortRate*(1-mortRate)/n),\n", " UL95 = mortRate + qnorm(0.975)*sqrt(mortRate*(1-mortRate)/n)) %>%\n", " ggplot(aes(dow,mortRate,group=admission_type,col=admission_type)) + \n", " geom_line() + \n", " geom_ribbon(aes(ymax=UL95,ymin=LL95,fill=admission_type,col=NULL),alpha=0.1) +\n", " xlab(\"Day of Week\") + ylab(\"Hospital Mortality Rate\")" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [ { "data": {}, "metadata": {}, "output_type": "display_data" }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94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\nTkSuQmCC", "text/plain": [ "plot without title" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "dat$weekend <- as.factor(dat$weekend)\n", "dat$admission_type <- as.factor(dat$admission_type)\n", "dat %>% group_by(admission_type,weekend) %>% \n", " summarise(n=n(),mortRate=mean(hospital_expire_flag==1),\n", " LL95 = mortRate - qnorm(0.975)*sqrt(mortRate*(1-mortRate)/n),\n", " UL95 = mortRate + qnorm(0.975)*sqrt(mortRate*(1-mortRate)/n)) %>%\n", " ggplot(aes(weekend,mortRate,group=admission_type,col=admission_type)) + \n", " geom_line() + \n", " geom_ribbon(aes(ymax=UL95,ymin=LL95,fill=admission_type,col=NULL),alpha=0.1) +\n", " xlab(\"Weekend\") + ylab(\"Hospital Mortality Rate\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# Model building\n", "\n", "Let's try to incorporate what we saw above into a very simple model. We will use logistic regression with hospital mortality as our outcome. First an unadjusted estimate, and then we will try to adjust for admission type.\n", "\n", "The unadjusted analysis should mirror pretty closely what we saw in the one of the tables above. The odds ratio corresponding with 14.0% and 10.8% mortality in the the weekend and weekday groups, respectively, is about 1.35. Performing logistic regression on the same data:\n" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "\n", "Call:\n", "glm(formula = hospital_expire_flag ~ weekend, family = \"binomial\", \n", " data = dat)\n", "\n", "Deviance Residuals: \n", " Min 1Q Median 3Q Max \n", "-0.5497 -0.4778 -0.4778 -0.4778 2.1104 \n", "\n", "Coefficients:\n", " Estimate Std. Error z value Pr(>|z|) \n", "(Intercept) -2.1127 0.0185 -114.204 < 2e-16 ***\n", "weekendTRUE 0.2992 0.0368 8.129 4.33e-16 ***\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: 27477 on 38556 degrees of freedom\n", "Residual deviance: 27413 on 38555 degrees of freedom\n", "AIC: 27417\n", "\n", "Number of Fisher Scoring iterations: 4\n" ] }, "metadata": {}, "output_type": "display_data" }, { "name": "stderr", "output_type": "stream", "text": [ "Waiting for profiling to be done...\n", "Warning message:\n", "“Removed 1 rows containing missing values (geom_hline).”" ] }, { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAAA0gAAANICAIAAAByhViMAAAACXBIWXMAABJ0AAASdAHeZh94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\nYQcAEISwAwAI4v8DilzOVUtSHRsAAAAASUVORK5CYII=", "text/plain": [ "plot without title" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "simple.glm <- glm(hospital_expire_flag ~ weekend,data=dat,family=\"binomial\")\n", "summary(simple.glm)\n", "\n", "# Uncomment for pretty graph\n", "## if(!(\"sjPlot\" %in% installed.packages())) { install.packages(\"sjPlot\")}\n", "## library(sjPlot)\n", "## sjp.glm(simple.glm)\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "yields the same results. The above coefficient for weekend is on the log scale, so when we exponentiate it, we get the odds-ratio: `exp(0.2992) = 1.35`. So, looking at these crude rates and odds ratios, we can see that patients admitted on a weekend have about a 35% increase in the odds of dying in the hospital when compared to those on a weekday. This effect is statistically significant (p<0.001).\n", "\n", "Are we done?\n", "\n", "I hope not. We saw from the tables and figures above, there is likely some confounding and maybe even effect modification happening. Next let's look at admission type and weekend ICU admission in the same model. There are two such models we could consider. The first adjusts for admission type, but assumes that the effect of weekend admission is the same regardless if the patient is of any of the admission types. The second one adjusts for admission type, but then allows the effect of weekend ICU admission to vary across the different levels of admission type. The first type of model would be able to account for confounding (when a nuisance variable is associated with both the outcome and the exposure/variable of interest), while the second permits what is called effect modification or a statistical interaction. Interactions are sometimes difficult to understand, but if ignored, can lead to incorrect conclusions about the effect of one or more of the variables.\n", "\n", "In this example, we fit both models, output estimates of the log-odds ratios, and perform a hypothesis test which evaluates the statistical significance of dropping one of the variables. Below is the resulting output:" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "\n", "Call:\n", "glm(formula = hospital_expire_flag ~ weekend + admission_type, \n", " family = \"binomial\", data = dat)\n", "\n", "Deviance Residuals: \n", " Min 1Q Median 3Q Max \n", "-0.5587 -0.5224 -0.5224 -0.2302 2.6995 \n", "\n", "Coefficients:\n", " Estimate Std. Error z value Pr(>|z|) \n", "(Intercept) -3.61725 0.08014 -45.137 < 2e-16 ***\n", "weekendTRUE 0.14442 0.03715 3.887 0.000101 ***\n", "admission_typeEMERGENCY 1.69439 0.08221 20.610 < 2e-16 ***\n", "admission_typeURGENT 1.58810 0.12308 12.903 < 2e-16 ***\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: 27477 on 38556 degrees of freedom\n", "Residual deviance: 26721 on 38553 degrees of freedom\n", "AIC: 26729\n", "\n", "Number of Fisher Scoring iterations: 6\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\t\n", "\n", "
DfDevianceAICLRTPr(>Chi)
<none>NA 26721.01 26729.01 NA NA
weekend 1 26735.87 26741.87 14.85949 1.158241e-04
admission_type 2 27412.85 27416.85 691.83845 5.877446e-151
\n" ], "text/latex": [ "\\begin{tabular}{r|lllll}\n", " & Df & Deviance & AIC & LRT & Pr(>Chi)\\\\\n", "\\hline\n", "\t & NA & 26721.01 & 26729.01 & NA & NA\\\\\n", "\tweekend & 1 & 26735.87 & 26741.87 & 14.85949 & 1.158241e-04\\\\\n", "\tadmission\\_type & 2 & 27412.85 & 27416.85 & 691.83845 & 5.877446e-151\\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "| | Df | Deviance | AIC | LRT | Pr(>Chi) | \n", "|---|---|---|\n", "| | NA | 26721.01 | 26729.01 | NA | NA | \n", "| weekend | 1 | 26735.87 | 26741.87 | 14.85949 | 1.158241e-04 | \n", "| admission_type | 2 | 27412.85 | 27416.85 | 691.83845 | 5.877446e-151 | \n", "\n", "\n" ], "text/plain": [ " Df Deviance AIC LRT Pr(>Chi) \n", " NA 26721.01 26729.01 NA NA\n", "weekend 1 26735.87 26741.87 14.85949 1.158241e-04\n", "admission_type 2 27412.85 27416.85 691.83845 5.877446e-151" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/plain": [ "\n", "Call:\n", "glm(formula = hospital_expire_flag ~ weekend * admission_type, \n", " family = \"binomial\", data = dat)\n", "\n", "Deviance Residuals: \n", " Min 1Q Median 3Q Max \n", "-0.5605 -0.5242 -0.5242 -0.2192 2.7353 \n", "\n", "Coefficients:\n", " Estimate Std. Error z value Pr(>|z|) \n", "(Intercept) -3.71692 0.08615 -43.143 < 2e-16 ***\n", "weekendTRUE 1.32317 0.23880 5.541 3.01e-08 ***\n", "admission_typeEMERGENCY 1.80138 0.08830 20.401 < 2e-16 ***\n", "admission_typeURGENT 1.60954 0.14959 10.760 < 2e-16 ***\n", "weekendTRUE:admission_typeEMERGENCY -1.20715 0.24185 -4.991 6.00e-07 ***\n", "weekendTRUE:admission_typeURGENT -0.98723 0.30357 -3.252 0.00115 ** \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: 27477 on 38556 degrees of freedom\n", "Residual deviance: 26700 on 38551 degrees of freedom\n", "AIC: 26712\n", "\n", "Number of Fisher Scoring iterations: 6\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\n", "
DfDevianceAICLRTPr(>Chi)
<none>NA 26700.44 26712.44 NA NA
weekend:admission_type 2 26721.01 26729.01 20.57385 3.407566e-05
\n" ], "text/latex": [ "\\begin{tabular}{r|lllll}\n", " & Df & Deviance & AIC & LRT & Pr(>Chi)\\\\\n", "\\hline\n", "\t & NA & 26700.44 & 26712.44 & NA & NA\\\\\n", "\tweekend:admission\\_type & 2 & 26721.01 & 26729.01 & 20.57385 & 3.407566e-05\\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "| | Df | Deviance | AIC | LRT | Pr(>Chi) | \n", "|---|---|\n", "| | NA | 26700.44 | 26712.44 | NA | NA | \n", "| weekend:admission_type | 2 | 26721.01 | 26729.01 | 20.57385 | 3.407566e-05 | \n", "\n", "\n" ], "text/plain": [ " Df Deviance AIC LRT Pr(>Chi) \n", " NA 26700.44 26712.44 NA NA\n", "weekend:admission_type 2 26721.01 26729.01 20.57385 3.407566e-05" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "# Without effect modification\n", "adj.glm <- glm(hospital_expire_flag ~ weekend + admission_type,data=dat,family=\"binomial\")\n", "summary(adj.glm)\n", "drop1(adj.glm,test=\"Chisq\")\n", "\n", "# With effect modification (interaction)\n", "adj.int.glm <- glm(hospital_expire_flag ~ weekend*admission_type,data=dat,family=\"binomial\")\n", "summary(adj.int.glm)\n", "drop1(adj.int.glm,test=\"Chisq\")" ] }, { "cell_type": "code", "execution_count": null, "metadata": { "collapsed": true }, "outputs": [], "source": [ "In the first model (no interaction), we see that although the effect of weekend is almost halved, it remains statistically significant, after adjusting for admission type (p<0.001).\n", "\n", "In the second model, we are primarily interested in the significance of the interaction. We can see when assessed with the `drop1` function, the interaction (`weekend:admission_type`) is statistically significant (p<0.001), suggesting that the effect of weekend may be different depending on which hospital admission type you are. How exactly to interpret this:\n", "\n", "One way of looking at this complexity is by computing the odds ratio in each of the levels of admission type. We can do this using the `predict` function, which by default outputs the log-odds of death. If for each hospital admission type, we calculate the log odds of death for each of the levels of weekend," ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\t\n", "\n", "
weekendadmission_typepred
TRUE ELECTIVE -2.393754
TRUE EMERGENCY-1.799525
TRUE URGENT -1.771439
\n" ], "text/latex": [ "\\begin{tabular}{r|lll}\n", " weekend & admission\\_type & pred\\\\\n", "\\hline\n", "\t TRUE & ELECTIVE & -2.393754\\\\\n", "\t TRUE & EMERGENCY & -1.799525\\\\\n", "\t TRUE & URGENT & -1.771439\\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "weekend | admission_type | pred | \n", "|---|---|---|\n", "| TRUE | ELECTIVE | -2.393754 | \n", "| TRUE | EMERGENCY | -1.799525 | \n", "| TRUE | URGENT | -1.771439 | \n", "\n", "\n" ], "text/plain": [ " weekend admission_type pred \n", "1 TRUE ELECTIVE -2.393754\n", "2 TRUE EMERGENCY -1.799525\n", "3 TRUE URGENT -1.771439" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\t\n", "\n", "
weekendadmission_typepred
FALSE ELECTIVE -3.716925
FALSE EMERGENCY-1.915548
FALSE URGENT -2.107381
\n" ], "text/latex": [ "\\begin{tabular}{r|lll}\n", " weekend & admission\\_type & pred\\\\\n", "\\hline\n", "\t FALSE & ELECTIVE & -3.716925\\\\\n", "\t FALSE & EMERGENCY & -1.915548\\\\\n", "\t FALSE & URGENT & -2.107381\\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "weekend | admission_type | pred | \n", "|---|---|---|\n", "| FALSE | ELECTIVE | -3.716925 | \n", "| FALSE | EMERGENCY | -1.915548 | \n", "| FALSE | URGENT | -2.107381 | \n", "\n", "\n" ], "text/plain": [ " weekend admission_type pred \n", "1 FALSE ELECTIVE -3.716925\n", "2 FALSE EMERGENCY -1.915548\n", "3 FALSE URGENT -2.107381" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "onWeekend <- expand.grid(weekend=\"TRUE\",admission_type=levels(dat$admission_type))\n", "onWeekday <- expand.grid(weekend=\"FALSE\",admission_type=levels(dat$admission_type))\n", "\n", "onWeekend$pred <- predict(adj.int.glm,newdata=onWeekend)\n", "onWeekend \n", "\n", "\n", "onWeekday$pred <- predict(adj.int.glm,newdata=onWeekday)\n", "onWeekday \n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ " We can now compute the log odds ratio ($log(OR) = logOdds_{weekend} - logOdds_{weekday}$), and exponentiate to get the odds ratio:\n" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "text/html": [ "\n", "\n", "\n", "\t\n", "\t\n", "\t\n", "\n", "
weekend.xadmission_typepred.xweekend.ypred.yOR
TRUE ELECTIVE -2.393754FALSE -3.7169253.755307
TRUE EMERGENCY-1.799525FALSE -1.9155481.123022
TRUE URGENT -1.771439FALSE -2.1073811.399257
\n" ], "text/latex": [ "\\begin{tabular}{r|llllll}\n", " weekend.x & admission\\_type & pred.x & weekend.y & pred.y & OR\\\\\n", "\\hline\n", "\t TRUE & ELECTIVE & -2.393754 & FALSE & -3.716925 & 3.755307 \\\\\n", "\t TRUE & EMERGENCY & -1.799525 & FALSE & -1.915548 & 1.123022 \\\\\n", "\t TRUE & URGENT & -1.771439 & FALSE & -2.107381 & 1.399257 \\\\\n", "\\end{tabular}\n" ], "text/markdown": [ "\n", "weekend.x | admission_type | pred.x | weekend.y | pred.y | OR | \n", "|---|---|---|\n", "| TRUE | ELECTIVE | -2.393754 | FALSE | -3.716925 | 3.755307 | \n", "| TRUE | EMERGENCY | -1.799525 | FALSE | -1.915548 | 1.123022 | \n", "| TRUE | URGENT | -1.771439 | FALSE | -2.107381 | 1.399257 | \n", "\n", "\n" ], "text/plain": [ " weekend.x admission_type pred.x weekend.y pred.y OR \n", "1 TRUE ELECTIVE -2.393754 FALSE -3.716925 3.755307\n", "2 TRUE EMERGENCY -1.799525 FALSE -1.915548 1.123022\n", "3 TRUE URGENT -1.771439 FALSE -2.107381 1.399257" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "onWeekend %>% inner_join(onWeekday,by=\"admission_type\") %>% mutate(OR=exp((pred.x-pred.y)))\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ " So, this mirrors what we saw above. While there may be differences between `EMERGENCY` and `URGENT` admission types, an `ELECTIVE` admission occurring on a weekend has an odds of mortality almost four times that of an `ELECTIVE` admission on a weekday. This seems particularly odd -- patients usually do not get admitted to a hospital electively on a weekend. \n", "\n", " What do you think? \n", "\n", " Do patients admitted on a weekend have a higher rate of mortality than those admitted during the week? \n", "\n", " Who is most affected, if at all?\n", "\n", " What factors can you rule out might be causing this effect? e.g., is it because the patients are simply sicker on a weekend? Are they more likely to have complications?\n", "\n", " Looking forward to see what you guys come up with!" ] } ], "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": "3.4.3" } }, "nbformat": 4, "nbformat_minor": 2 }