One interesting functionality that might also be helpful to dig deeper into survival analysis is adjusting survival curves for CoxPH models (this is far less commonly employed in standard-clinical analysis, but offers some adavantages nonetheless). Often (unless in RCTs) clinical cohorts are not perfectly similar in terms of baseline clinical characteristics (as is even the case in large datasets). To account for differences in baseline clinical characteristics one can employ a CoxPH model to adjust the KM survival curves accordingly taking those dissimilarities at least in parts into account.
This could be helpful if ehrapy was to be used to stratify patients by a given feature but stratification by this feature yielded somehow dissimilar patient groups (i.e. gender, or age-deviated).
One implementation of this CoxPh-adjusted survival analysis can be found in the survminer package in R (https://search.r-project.org/CRAN/refmans/survminer/html/ggadjustedcurves.html).
That´s mostly it from the basic clinical side - lots more to explore though!
Description of feature
Hi all, hope you made it until this one...
One interesting functionality that might also be helpful to dig deeper into survival analysis is adjusting survival curves for CoxPH models (this is far less commonly employed in standard-clinical analysis, but offers some adavantages nonetheless). Often (unless in RCTs) clinical cohorts are not perfectly similar in terms of baseline clinical characteristics (as is even the case in large datasets). To account for differences in baseline clinical characteristics one can employ a CoxPH model to adjust the KM survival curves accordingly taking those dissimilarities at least in parts into account. This could be helpful if ehrapy was to be used to stratify patients by a given feature but stratification by this feature yielded somehow dissimilar patient groups (i.e. gender, or age-deviated). One implementation of this CoxPh-adjusted survival analysis can be found in the survminer package in R (https://search.r-project.org/CRAN/refmans/survminer/html/ggadjustedcurves.html). That´s mostly it from the basic clinical side - lots more to explore though!