Survival analysis with python - how to obtain adjusted KM estimates of cumulative risk

I am new to survival analysis. I have a database with mortality data for people with or without a certain condition. I would like to plot the cumulative risk of mortality, and also calculate the hazard ratio with confidence interval and p-value for the log-rank test. Model assumptions for COX are fine.

I research the literature, and most articles with a similar objective obtain the following: I) Kaplan–Meier estimates of the cumulative risk (adjusted by age and sex); 2) Hazard ratios and corresponding 95% confidence intervals are obtained from stratified Cox proportional-hazards models.

My questions are:

I) How to plot age and sex-adjusted Kaplan Meier estimates of cumulative risk in python? I checked the documentation for the lifelines package but was not able to find instructions. In R, the survminer package provides adjusted survival curves (ggcoxadjustedcurves), but I couldn't find a way to plot adjusted cumulative risk KM estimates.

II) I did find a way to plot adjusted cumulative hazard curves using the lifelines package. How are those different from the KM-estimates of cumulative risk? How to interpret each? If they are similar, why the literature prefers the KM estimate?

https://stackoverflow.com/questions/66937929/survival-analysis-with-python-how-to-obtain-adjusted-km-estimates-of-cumulativ April 04, 2021 at 12:06PM