The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.

KMunicate-Style Kaplan–Meier Plots

Codecov test coverage CRAN status CRAN_Logs_Badge CRAN_Logs_Badge_Total R-CMD-check

The goal of {KMunicate} is to produce Kaplan–Meier plots in the style recommended following the KMunicate study (TP Morris et al. Proposals on Kaplan–Meier plots in medical research and a survey of stakeholder views: KMunicate. BMJ Open, 2019, 9:e030215).

Installation

You can install {KMunicate} from CRAN by typing the following in your R console:

install.packages("KMunicate")

Alternatively, you can install the dev version of {KMunicate} from GitHub with:

# install.packages("devtools")
devtools::install_github("ellessenne/KMunicate-package")

Example

library(survival)
library(KMunicate)

The {KMunicate} package comes with a couple of bundled dataset, cancer and brcancer. The main function is named KMunicate:

KM <- survfit(Surv(rectime, censrec) ~ hormon, data = brcancer)
time_scale <- seq(0, max(brcancer$rectime), by = 365)
KMunicate(fit = KM, time_scale = time_scale)

KM <- survfit(Surv(studytime, died) ~ drug, data = cancer2)
time_scale <- seq(0, max(cancer2$studytime), by = 7)
KMunicate(fit = KM, time_scale = time_scale)

You also might wonder, does this work with a single arm? Yes, yes it does:

KM <- survfit(Surv(studytime, died) ~ 1, data = cancer2)
time_scale <- seq(0, max(cancer2$studytime), by = 7)
KMunicate(fit = KM, time_scale = time_scale)

Finally, you can also plot 1 - survival by using the argument .reverse = TRUE:

KM <- survfit(Surv(rectime, censrec) ~ hormon, data = brcancer)
time_scale <- seq(0, max(brcancer$rectime), by = 365)
KMunicate(fit = KM, time_scale = time_scale, .reverse = TRUE)

Customise Risk Table

By default, KMunicate() will build a risk table conform to the KMunicate style, e.g., with cumulative number of events and censored (the column-wise sum is equal to the total number of individuals at risk per arm):

KM <- survfit(Surv(rectime, censrec) ~ hormon, data = brcancer)
time_scale <- seq(0, max(brcancer$rectime), by = 365)
KMunicate(fit = KM, time_scale = time_scale)

Alternatively, it is possible to customise the risk table via the .risk_table argument. For instance, if one wants to have interval-wise number of events and censored, just pass the survfit value to the .risk_table argument:

KMunicate(fit = KM, time_scale = time_scale, .risk_table = "survfit")

This is the default output of the summary.survfit() function.

Finally, it is also possible to fully omit the risk table by setting .risk_table = NULL:

KMunicate(fit = KM, time_scale = time_scale, .risk_table = NULL)

Custom Fonts

Assuming you have set up your computer to use custom fonts with ggplot2, customising your KMunicate-style plot is trivial. All you have to do is pass the font name as the .ff argument:

KM <- survfit(Surv(studytime, died) ~ 1, data = cancer2)
time_scale <- seq(0, max(cancer2$studytime), by = 7)
KMunicate(fit = KM, time_scale = time_scale, .ff = "Times New Roman")

Further Customisation

Several options to further customise each plot are provided, see e.g. the introductory vignette for more details.

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.