0

What is the difference between survreg and survfit?

Survreg: Fit a parametric survival regression model. These are location-scale models for an arbitrary transform of the time variable; the most common cases use a log transformation, leading to accelerated failure time models.

Survfit: Computes an estimate of a survival curve for censored data using either the Kaplan-Meier or the Fleming-Harrington method or computes the predicted survivor function for a Cox proportional hazards model.

Ben
  • 3,443
  • 2
    Does this answer your question? In survival analysis, why do we use semi-parametric models (Cox proportional hazards) instead of fully parametric models?. It's just the difference between a parametric (survreg) and a non- or semi-parametric model (survfit). – EdM Apr 11 '22 at 16:11
  • yes, thanks! why is there so much more content available about survfit than about survreg on google? – Ben Apr 11 '22 at 18:14
  • 1
    The non-parametric Kaplan-Meier or semi-parametric Cox models handled by survfit don't require knowledge of the underlying baseline survival curve. Cox models do require a proportional hazards (PH) assumption, but that can be checked from the data and the model can often then be adapted to fit PH. In particular, a Weibull model is a PH model but not all PH models are Weibull models. So a Cox model with results displayed by survfit is very often the initial go-to choice, particularly in clinical survival work. The better distinction should be survreg versus coxph, the model fitters. – EdM Apr 11 '22 at 18:20

0 Answers0