Skip to contents

Publication Ready Kaplan-Meier Estimator

Source: R/surv.plot.R
surv.plot.Rd

Provide an open-source, user-friendly tool designed to enhance the creation and customization of Kaplan-Meier plots and incorporating various statistics and layout customization options using surv.plot(fit, ...).

Arguments

fit

An object of class survival::survfit containing survival data.

reference.arm

A character string specifying the reference arm for comparison.

time.unit

A character string specifying the time unit which was used to create the fit object. Note: time.unit will not convert the time of the fit object.

Option include: "day", "week", "month","year".

y.unit

A character string specifying the unit of the y-axis.

Option include: "probability", "percent".

censoring.mark

A logical parameter indicating whether censoring events should be marked on the survival curves. Default is TRUE.

censoring.cex

A numeric value specifying the size of the marks for censored patients. Default is 1.3.

conf.int

A numeric value controlling the confidence interval on survival curves. Default is 0.95, corresponding to a 95% confidence interval. Values between 0 and 1 represent the desired confidence interval. If set to 0, no confidence intervals are displayed.

conf.band

A logical parameter indicating whether to display the confidence band on the survival curves. Default is TRUE.

conf.band.col

A colour which is used for the confidence band. Can accept a single colour value or a vector of colours.

conf.band.transparent

A numeric value between 0 and 1 controlling the transparency of the confidence band. Default is 0.25.

conf.line.lty

A strings specifying the line type of the confidence lines.

Options include: "blank", "solid", "dashed", "dotted", "dotdash", "longdash", "twodash". Default is "blank".

conf.line.lwd

A numeric value specifying the width of the confidence lines.

conf.type

Transformation type for the confidence interval.

Options include: "log", "log-log", "plain", "logit", "arcsin". Default is log-log.

grid

A logical parameter specifying whether to draw a grid. Default is FALSE.

col

A colour which is used for the survival curves. Can accept a single colour value or a vector of colours.

main

Title of the plot.

sub

Subtitle of the plot. Note: A subtitle is only displayed if no risk table is shown.

xlab

X-axis label.

ylab

Y-axis label.

xticks

A numeric vector specifying the ticks of the x-axis.

Can be specified as seq(from = , to = , by = ).

  • from: starting value

  • to: end value

  • by: number; increment of the sequence

yticks

A numeric vector specifying the ticks of the y-axis.

Can be specified as seq(from = , to = , by = ).

  • from: starting value

  • to: end value

  • by: number; increment of the sequence

Note: It should always be specified as probability.

xlab.pos

Defines the margin line where the X-axis label (xlab) is displayed, starting at 0 and counting outwards. Default is 1.5.

ylab.pos

Defines the margin line the Y-axis label (ylab) is displayed, starting at 0 counting outwards. Default is 3.

xlab.cex

A numeric value specifying the size of the X-axis label.

ylab.cex

A numeric value specifying the size of the Y-axis label.

cex

A numeric value specifying the size of all all text elements (labels, annotations, etc.).

axis.cex

A numeric value specifying the size of the axis elements.

bty

Determines the style of the box drawn around the plot.

Options include: "n" ,"o","c","u". Default is "n".

lty

A string specifying the line type of of the curve(s).

Options include: "blank", "solid", "dashed", "dotted", "dotdash", "longdash", "twodash".

lwd

A numeric value specifying the width of the line.

legend

A logical parameter specifying whether to display legend. By default, the legend is displayed if there is more than one arm.

legend.position

Position of the legend.

Options include: "c(x,y)", "bottomright", "bottom", "bottomleft", "left"`,

legend.name

A vector of character string specifying of the name(s) of the arm(s).

legend.cex

A numeric value specifying the size of the legend text.

legend.text.font

Font style of the legend text. Possible values:

  • 1 normal

  • 2 bold

  • 3 italic

  • 4 bold and italic

legend.title

Title of the legend.

legend.title.cex

A numeric value specifying the size of the legend title.

segment.type

A numeric value specifying the layout of the segment. Possible values:

  • 1 full width

  • 2 half width

  • 3 vertical and horizontal segment (default)

segment.timepoint

A single value or a vector of fixed time points to be drawn as segment(s).

segment.quantile

A single value or a vector of fixed quantile to be drawn as segment(s). Example: 0.5 corresponds to median.

segment.main

Title of the segment text.

segment.confint

A logical parameter specifying whether to display the confidence interval for the segment.

NOTE: Only possible to set segment.confint = FALSE if there are two arms. Default is TRUE.

segment.annotation

Position of the segment annotation.

Options include: c(x,y),"bottomleft", "left", "right", "top", "none".

segment.annotation.two.lines

A logical parameter to force that the annotation is displayed on two lines even if there is only one arm. This parameter only has an effect if there is only one arm. Default: FALSE

segment.annotation.col

A colour which is used for the segment annotation. Can accept a single colour value or a vector of colours.

segment.annotation.space

Spacing between the text in units of x-coordinates.

segment.col

A colour which is used for the segment. Can accept a single colour value.

segment.lty

A strings specifying the line type of the segment(s).

Options include: "blank", "solid", "dashed", "dotted", "dotdash", "longdash", "twodash".

segment.lwd

A numeric value specifying the width of the segment line.

segment.cex

A numeric value specifying the size of the segment text size.

segment.font

A numeric value specifying the font face. Possible values:

  • 1 plain

  • 2 bold

  • 3 italic

  • 4 bold-italic

segment.main.font

A numeric value specifying the font face for the segment text. Possible values:

  • 1 plain

  • 2 bold

  • 3 italic

  • 4 bold-italic

stat.fit

An object of class survival::survfit containing survival data. Used for calculation of statistics, allowing to add stratification factors.

Note: If not specified the fit object will be used for the stat.

stat

Statistics displayed in the plot.

Options:

  • "logrank" gives the p value of the conducted logrank test using survdiff{survival}. To tests if there is a difference between two or more survival curves.

  • "coxph" gives the hazard ratio (HR) and its CI (default: 95% CI)of the conducted Cox proportional hazards regression using coxph{survival}. Note: This option only works if there are two arms.

  • "coxph_logrank" combines the hazard ratio (HR), its CI (default: 95% CI) and the logrank test. Note: This option only works if there are two arms.

  • 'none' no statistic is displayed (default).

Note: Confidence interval can be adjusted with the argument stat.conf.int.

stat.position

Position where the stat should be displayed.

Options include: c(x,y),"bottomleft", "left", "right", "top", "topright","bottomright", "none".

stat.conf.int

A numeric value controlling the confidence interval on the stat (hazard ratio). Default is 0.95, corresponds to a 95% confidence interval. Values between 0 and 1 represent the desired confidence interval.

stat.col

A colour which is used for the statistics text. Can accept a single colour value or a vector of colours.

stat.cex

A numeric value specifying the size of the `statistics text size.

stat.font

The font face of the statistics Possible values:

  • 1 plain

  • 2 bold

  • 3 italic

  • 4 bold-italic

risktable

A logical parameter indicating whether to draw risk table. Default is TRUE.

risktable.censoring

A logical parameter indicating whether to display number of censored patients. Default is FALSE.

risktable.pos

Defines on which margin line of the xlab is displayed, starting at 0 counting outwards. Default is at line 3.

risktable.name

Names of the arms for the risk table.

risktable.cex

A numeric value specifying the size of the risk table text size.

risktable.col

A coulour which is used for the risk table. Can accept a single colour value or a vector of colours. Default is black.

Note: If risktable.col = TRUE then the colours of the curves are used.

risktable.title

Title of risk table.

risktable.title.font

Font style of the risk table. Possible values:

  • 1 normal

  • 2 bold

  • 3 italic

  • 4 bold and italic

risktable.title.col

A colour which is used for the risk table title. Can accept a single colour value.

risktable.title.position

A numeric value specifying the position of the title on the x-axis.

risktable.title.cex

A numeric value specifying the size of the risk table title size.

risktable.name.cex

A numeric value specifying the size of the risk table legend name size.

risktable.name.font

Font style of the risk table legend name(s). Possible values:

  • 1 normal

  • 2 bold

  • 3 italic

  • 4 bold and italic

risktable.name.col

A colour which is used for the risk table name. Can accept a single colour value.

risktable.name.position

A numeric value specifying the position of the legend name(s) on the x-axis.

margin.bottom

Specifies the bottom margin of the plotting area in line units. Default is 5.

margin.left

Specifies the left margin of the plotting area in line units. Default is 6 (with risktable) or 4 (without risktable).

margin.top

Specifies the top margin of the plotting area in line units. Default is 3.

margin.right

Specifies the right margin of the plotting area in line units. Default is 2.

theme

Built-in layout options. Options include: ("none", "SAKK", "Lancet", "JCO", "WCLC", "ESMO")

Value

Kaplan-Meier curves of the input fit, incorporating various statistics and layout option(s).

Details

The survSAKK R package provides the surv.plot() function, facilitating Kaplan-Meier survival analysis. Designed with user-friendliness and efficiency in mind. Offering robust tool for analysing survival data. It utilises the functionalities of survival::survfit().

For a comprehensive manual visit: https://sakk-statistics.github.io/survSAKK/articles/surv.plot.html

References

Therneau T (2024). A Package for Survival Analysis in R. R package version 3.5-8, https://CRAN.R-project.org/package=survival.

Terry M. Therneau, Patricia M. Grambsch (2000). Modeling Survival Data: Extending the Cox Model. Springer, New York. ISBN 0-387-98784-3.

See also

Author

Vithersan Somasundaram and Katrin Gysel

Examples

 require(survival)
 require(survSAKK)

# Create survival object
 fit <- survfit(Surv(lung$time/365.25*12, status) ~ sex, data = lung)

# Generate survival plots
 surv.plot(fit = fit,
   time.unit = "month",
   legend.name =  c("male", "female"))