Code
knitr::opts_chunk$set(comment = NA)
library(tableone)
library(janitor)
library(tidyverse)Building Table 1 from the bradley.csv data
This is a “live code” example that Dr. Love developed for Class 12 of 432, on 2023-02-23.
1 Load Packages
2 Ingest the bradley.csv data
Code
brad_raw <- read_csv("data/bradley.csv",
show_col_types = FALSE) |>
clean_names()
brad_raw# A tibble: 1,374 × 7
subject status age sex race_eth married location
<dbl> <chr> <dbl> <chr> <chr> <dbl> <chr>
1 1 Control 64 Male white 1 Bed
2 2 Case 70 Male white 1 ICU
3 3 Control 68 Male white 0 Bed
4 4 Control 76 Male white 1 Bed
5 5 Control 70 Male white 1 Bed
6 6 Case 89 Male white 1 Other
7 7 Control 75 Male white 0 Bed
8 8 Control 84 Male white 1 Bed
9 9 Control 80 Male white 1 ICU
10 10 Control 89 Male white 0 Bed
# … with 1,364 more rows3 Attempt 1
Code
vars <- c("age", "sex", "race_eth", "married", "location")
trt <- c("status")
attempt_1 <- CreateTableOne(data = brad_raw,
vars = vars,
strata = trt)
print(attempt_1) Stratified by status
Case Control p test
n 687 687
age (mean (SD)) 73.78 (10.24) 72.60 (10.50) 0.035
sex = Male (%) 677 (98.5) 666 (96.9) 0.069
race_eth = white (%) 546 (79.5) 527 (76.7) 0.240
married (mean (SD)) 0.52 (0.50) 0.45 (0.50) 0.013
location (%) <0.001
Bed 446 (64.9) 580 (84.4)
ICU 186 (27.1) 65 ( 9.5)
Other 55 ( 8.0) 42 ( 6.1) 4 What’s wrong here?
Code
brad_new <- brad_raw |>
mutate(marital = fct_recode(factor(married),
"yes" = "1",
"no" = "0")) |>
mutate(loc = fct_relevel(location,
"ICU", "Bed", "Other"))5 Second Attempt
Code
vars <- c("age", "sex", "race_eth", "marital", "loc")
factorvars <- c("sex", "race_eth", "marital", "loc")
trt <- c("status")
attempt_2 <- CreateTableOne(data = brad_new,
vars = vars,
factorVars = factorvars,
strata = trt)
print(attempt_2) Stratified by status
Case Control p test
n 687 687
age (mean (SD)) 73.78 (10.24) 72.60 (10.50) 0.035
sex = Male (%) 677 (98.5) 666 (96.9) 0.069
race_eth = white (%) 546 (79.5) 527 (76.7) 0.240
marital = yes (%) 356 (51.8) 310 (45.1) 0.015
loc (%) <0.001
ICU 186 (27.1) 65 ( 9.5)
Bed 446 (64.9) 580 (84.4)
Other 55 ( 8.0) 42 ( 6.1) 6 Show alternative summaries?
Code
print(attempt_2,
nonnormal = c("age"),
exact = c("sex", "race_eth", "marital")) Stratified by status
Case Control p test
n 687 687
age (median [IQR]) 74.00 [67.00, 81.00] 73.00 [65.00, 80.00] 0.035 nonnorm
sex = Male (%) 677 (98.5) 666 (96.9) 0.068 exact
race_eth = white (%) 546 (79.5) 527 (76.7) 0.240 exact
marital = yes (%) 356 (51.8) 310 (45.1) 0.015 exact
loc (%) <0.001
ICU 186 (27.1) 65 ( 9.5)
Bed 446 (64.9) 580 (84.4)
Other 55 ( 8.0) 42 ( 6.1) 7 A more detailed summary?
Code
summary(attempt_2)
### Summary of continuous variables ###
status: Case
n miss p.miss mean sd median p25 p75 min max skew kurt
age 687 0 0 74 10 74 67 81 38 106 0.02 0.2
------------------------------------------------------------
status: Control
n miss p.miss mean sd median p25 p75 min max skew kurt
age 687 0 0 73 11 73 65 80 34 103 0.08 -0.1
p-values
pNormal pNonNormal
age 0.03527177 0.03548289
Standardize mean differences
1 vs 2
age 0.1137012
=======================================================================================
### Summary of categorical variables ###
status: Case
var n miss p.miss level freq percent cum.percent
sex 687 0 0.0 Female 10 1.5 1.5
Male 677 98.5 100.0
race_eth 687 0 0.0 non-white 141 20.5 20.5
white 546 79.5 100.0
marital 687 0 0.0 no 331 48.2 48.2
yes 356 51.8 100.0
loc 687 0 0.0 ICU 186 27.1 27.1
Bed 446 64.9 92.0
Other 55 8.0 100.0
------------------------------------------------------------
status: Control
var n miss p.miss level freq percent cum.percent
sex 687 0 0.0 Female 21 3.1 3.1
Male 666 96.9 100.0
race_eth 687 0 0.0 non-white 160 23.3 23.3
white 527 76.7 100.0
marital 687 0 0.0 no 377 54.9 54.9
yes 310 45.1 100.0
loc 687 0 0.0 ICU 65 9.5 9.5
Bed 580 84.4 93.9
Other 42 6.1 100.0
p-values
pApprox pExact
sex 6.926864e-02 6.760400e-02
race_eth 2.403791e-01 2.403489e-01
marital 1.513478e-02 1.510454e-02
loc 1.429017e-17 3.999616e-18
Standardize mean differences
1 vs 2
sex 0.10797814
race_eth 0.06690265
marital 0.13427981
loc 0.489231028 Send original attempt 2 to Excel for polish?
One option is to save the Table 1 to a .csv file, which you can then open directly in Excel. This is the approach I generally use. Note the addition of some quote, noSpaces and printToggle selections here.
Code
brad_t1 <- print(attempt_2,
quote = FALSE, noSpaces = TRUE, printToggle = FALSE)
write.csv(brad_t1,
file = "data/bradley_table1_result.csv")You can then open the bradley_table1_result.csv file in Excel and edit further.