Add at risk table and 95% confidence intervals to adjusted survival curves using survminer package in r

Question

I want to add at-risk table and 95% confidence intervals to adjusted survival curves. Something like survival curves below (Link). I saw some relevant code here but it doesn't mention what I need.

Another question, would it be correct if I used the obtained weight variable (from iptw package) in the adjustment (as I have a column named weight in my actual dataset.

Here is my code:

library(survival);library(survminer)
data(lung);names(lung)
#fit <- coxph( Surv(time, status==2) ~ ph.karno + strata(sex), data = lung )
lung$sex <- ifelse(lung$sex == 1,  "Male", "Female")
fit <- coxph(Surv(time, status) ~  ph.ecog + age +strata(sex),  data = lung)

ggadjustedcurves(fit,
                 variable = "sex", 
                 data = lung,
                 method = "average",
                 palette = c("#E69F00", "#56B4E9"),
                 size = 1.3,
                 legend = "right",
                 legend.title = expression(bold("Legend title")),
                 xlab = "Time",font.legend = 12) + theme(legend.text.align = 0.5)

Answer 1

Short Answer: Currently there is no R-Package that allows you to directly plot confounder-adjusted survival curves with confidence intervals. However there is a function called ate in the riskRegression R-Package which can be used to calculate adjusted survival probability estimates at some points in time with 95% confidence intervals. If you do some coding you can definitly use that one to get what you want.

Alternatively you can wait a little longer. I am currently working on an R-Package that implements various methods of confounder-adjusted survival curves (and cumulative incidence functions when there are competing risks) complete with confidence intervals, hypothesis tests, risk tables and so on.

UPDATE (09.05.2022):

The adjustedCurves package can now be used to obtain adjusted survival curves with confidence intervals. For your example:

library(survival)
library(devtools)

# install adjustedCurves from github, load it
devtools::install_github("/RobinDenz1/adjustedCurves")
library(adjustedCurves)

# fit required model
lung$sex <- as.factor(ifelse(lung$sex == 1,  "Male", "Female"))
lung$status <- lung$status - 1

fit <- coxph(Surv(time, status) ~  ph.ecog + age + strata(sex), data=lung,
             x=TRUE)

# calculate and plot curves
adj <- adjustedsurv(data=lung, variable="sex", ev_time="time",
                    event="status", method="direct",
                    outcome_model=fit, conf_int=TRUE)
plot(adj, conf_int=TRUE)