reportRmd

The goal of reportRmd is to automate the reporting of clinical data in Rmarkdown environments. Functions include table one-style summary statistics, compilation of multiple univariate models, tidy output of multivariable models and side by side comparisons of univariate and multivariable models. Plotting functions include customisable survival curves, forest plots, and automated bivariate plots.

Installation

Installing from CRAN:

install.packages('reportRmd')

You can install the development version of reportRmd from GitHub with:

# install.packages("devtools")
devtools::install_github("biostatsPMH/reportRmd", ref="development")

New Features

• Survival curves have been improved and now return ggplots
• Variable labels will now be automatically output in tables

Documentation

Online Documentation

Examples

Summary statistics by Sex

library(reportRmd)
data("pembrolizumab")
rm_covsum(data=pembrolizumab, maincov = 'sex',
covs=c('age','pdl1','change_ctdna_group'),
show.tests=TRUE)

Using Variable Labels

var_names <- data.frame(var=c("age","pdl1","change_ctdna_group"),
                          label=c('Age at study entry',
                                  'PD L1 percent',
                                  'ctDNA change from baseline to cycle 3'))

pembrolizumab <- set_labels(pembrolizumab,var_names)

rm_covsum(data=pembrolizumab, maincov = 'sex',
covs=c('age','pdl1','change_ctdna_group'))

Multiple Univariate Regression Analyses

rm_uvsum(data=pembrolizumab, response='orr',
covs=c('age','pdl1','change_ctdna_group'))
#> Waiting for profiling to be done...
#> Waiting for profiling to be done...
#> Waiting for profiling to be done...

Tidy multivariable analysis

glm_fit <- glm(orr~change_ctdna_group+pdl1+cohort,
               family='binomial',
               data = pembrolizumab)
rm_mvsum(glm_fit,showN=T)

Combining univariate and multivariable models

uvsumTable <- rm_uvsum(data=pembrolizumab, response='orr',
covs=c('age','sex','pdl1','change_ctdna_group'),tableOnly = TRUE)
#> Waiting for profiling to be done...
#> Waiting for profiling to be done...
#> Waiting for profiling to be done...
#> Waiting for profiling to be done...

glm_fit <- glm(orr~change_ctdna_group+pdl1,
               family='binomial',
               data = pembrolizumab)
mvsumTable <- rm_mvsum(glm_fit,tableOnly = TRUE)

rm_uv_mv(uvsumTable,mvsumTable)

Simple survival summary table

Shows events, median survival, survival rates at different times and the log rank test. Does not allow for covariates or strata, just simple tests between groups

 rm_survsum(data=pembrolizumab,time='os_time',status='os_status',
 group="cohort",survtimes=c(12,24),
# group="cohort",survtimes=seq(12,36,12),
# survtimesLbls=seq(1,3,1),
 survtimesLbls=c(1,2),
 survtimeunit='yr')

Summarise Cumulative incidence

library(survival)
data(pbc)
rm_cifsum(data=pbc,time='time',status='status',group=c('trt','sex'),
eventtimes=c(1825,3650),eventtimeunit='day')
#> 106 observations with missing data were removed.

Plotting survival curves

ggkmcif2(response = c('os_time','os_status'),
cov='cohort',
data=pembrolizumab)

Plotting odds ratios

require(ggplot2)
#> Loading required package: ggplot2
forestplot2(glm_fit)
#> Warning: `forestplot2()` was deprecated in reportRmd 0.1.0.
#> ℹ Please use `forestplotUV()` instead.
#> This warning is displayed once every 8 hours.
#> Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
#> generated.
#> Warning: Vectorized input to `element_text()` is not officially supported.
#> ℹ Results may be unexpected or may change in future versions of ggplot2.

Plotting bivariate relationships

These plots are designed for quick inspection of many variables, not for publication.

require(ggplot2)
plotuv(data=pembrolizumab, response='orr',
covs=c('age','cohort','pdl1','change_ctdna_group'))
#> Boxplots not shown for categories with fewer than 20 observations.
#> Boxplots not shown for categories with fewer than 20 observations.