Kaplan-Meier Plot with 'ggplot2'

The function 'jskm()' creates publication quality Kaplan-Meier plot with at risk tables below. 'svyjskm()' provides plot for weighted Kaplan-Meier estimator.

Kaplan-Meier Plot with 'ggplot2': 'survfit' and 'svykm' objects from 'survival' and 'survey' packages.

Install

install.packages("jskm")
install.packages("remotes")
remotes::install_github("jinseob2kim/jskm")
 
library(jskm)

Example

Survival probability

#Load dataset
library(survival)
data(colon)
fit <- survfit(Surv(time,status)~rx, data=colon)
 
#Plot the data
jskm(fit)

jskm(fit, table = T, pval = T, label.nrisk = "No. at risk", size.label.nrisk = 8, 
     xlabs = "Time(Day)", ylabs = "Survival", ystratalabs = c("Obs", "Lev", "Lev + 5FU"), ystrataname = "rx",
     marks = F, timeby = 365, xlims = c(0, 3000), ylims = c(0.25, 1))

Cumulative hazard: 1- Survival probability

jskm(fit, ci = T, cumhaz = T,  mark = F, ylab = "Cumulative hazard (%)", surv.scale = "percent", pval =T, pval.size = 6, pval.coord = c(300, 0.7))

Weighted Kaplan-Meier plot - svykm.object in survey package

library(survey)
data(pbc, package="survival")
pbc$randomized <- with(pbc, !is.na(trt) & trt>0)
biasmodel <- glm(randomized~age*edema,data=pbc)
pbc$randprob <- fitted(biasmodel)
 
dpbc<-svydesign(id=~1, prob=~randprob, strata=~edema, data=subset(pbc,randomized))
 
s1 <-svykm(Surv(time,status>0) ~ 1, design = dpbc)
s2 <-svykm(Surv(time,status>0) ~ sex, design = dpbc)
 
svyjskm(s1)

svyjskm(s2, pval = T,  table = T, design = dpbc)

svyjskm(s2, cumhaz = T, ylab = "Cumulative (%)", surv.scale = "percent", pval = T, design = dpbc, pval.coord = c(300, 0.7)) 

If you want to get confidence interval, you should apply se = T option to svykm object.

s3 <- svykm(Surv(time,status>0) ~ sex, design=dpbc, se = T)
svyjskm(s3)

svyjskm(s3, ci = F)