Skip to contents

Create Kaplan-Meier Plot

Source: R/plotting_survival.R
create_km_plot.Rd

Kaplan-Meier plot using ggplot2 and survival.

Usage

create_km_plot(
  data,
  time_var = "AVAL",
  event_var = "CNSR",
  trt_var = "TRT01P",
  title = "Kaplan-Meier Plot",
  xlab = "Time",
  ylab = "Survival Probability",
  risk_table = FALSE,
  show_median = FALSE,
  show_ci = FALSE,
  show_censor = TRUE,
  landmarks = NULL,
  xlim = NULL,
  palette = NULL,
  conf_level = 0.95,
  base_size = 11,
  type = c("km", "loglog")
)

Arguments

data

ADaMData object or data frame containing survival data

time_var

Time variable name (default: "AVAL")

event_var

Event variable name. If "CNSR" (ADaM censoring flag), it will be automatically inverted (0=event becomes 1=event). Otherwise expects 1=event, 0=censor. Default: "CNSR"

trt_var

Treatment variable name (default: "TRT01P")

title

Plot title

xlab

X-axis label

ylab

Y-axis label

risk_table

Logical, include risk table below. Requires patchwork.

show_median

Logical, add horizontal/vertical lines at median survival time (default: FALSE)

show_ci

Logical, show confidence bands around survival curves (default: FALSE)

show_censor

Logical, show censoring marks as crosses (default: TRUE)

landmarks

Numeric vector of timepoints to highlight with vertical lines (e.g., c(12, 24) for 12 and 24 months). NULL for none.

xlim

Optional x-axis limits as c(min, max)

palette

Optional color palette for treatment groups. Can be a character vector of colors, or NULL to use getOption("pharmhand.palette"). Defaults to the CVD-friendly "Okabe-Ito (reordered)" palette (orange, sky blue for first two arms). Other built-in options: "Okabe-Ito", "R4", "Tableau 10", "Alphabet", etc. (see grDevices::palette.pals()).

conf_level

Confidence level for CI bands (default: 0.95)

base_size

Base font size for plot text elements (default: 11). Also used for risk table text.

type

Character. Plot type: "km" for Kaplan-Meier or "loglog" for log(-log(S(t))) vs log(t). For "loglog", median lines, CI bands, risk tables, and landmarks are intentionally omitted to keep the diagnostic scale uncluttered; censor marks are supported via show_censor.

Value

A ClinicalPlot object

Examples

if (FALSE) { # \dontrun{
# Basic KM plot
km <- create_km_plot(
  data = adtte,
  time_var = "AVAL",
  event_var = "CNSR",
  title = "Overall Survival"
)

# With median lines and CI bands
km <- create_km_plot(
  data = adtte,
  show_median = TRUE,
  show_ci = TRUE,
  risk_table = TRUE
)
} # }