Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

Find a Community

- Home
- /
- Analytics
- /
- Stat Procs
- /
- How do I generate a Restricted Mean Survival Time ...

- RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

02-15-2018 03:28 PM

I am working on a longitudinal project with an enormous dataset (over 1.5 million participants). I am looking at their survival based on several factors including some measure of disease and age. I have been able to plot KM curves through proc lifetest, and adjusted KM curves using the baseline statement in proc phreg. The investigator I am working for, however, does not feel that Kaplan Meier curves or hazard rates are intuitive from a clinician point of view. He said that clinicians expect to see a "Restricted Mean Survival Time Curve." RMST is defined as the area under the survival curve up to a point t*.

I have attached an example of the type of result I am looking for to this post. I have been unable to find anything regarding how to generate such a curve. If you notice, instead of survival probability on the Y-axis, it is Mean Survival Time, and on the X-axis it is age rather than time. For each of my participants, I have starting age, degree of disease (a categorical variable with 5 levels), and date of death/age at death. I have seen many items regarding calculating a RMST at a point in time, but I have not seen anything about generating this type of curve in SAS. Any guidance would be much appreciated.

Thank you

- Mark as New
- Bookmark
- Subscribe
- RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Posted in reply to Nusseey

02-15-2018 03:51 PM

Nusseey wrote:

. The investigator I am working for, however, does not feel that Kaplan Meier curves or hazard rates are intuitive from a clinician point of view. He said that clinicians expect to see a "Restricted Mean Survival Time Curve." RMST is defined as the area under the survival curve up to a point t*.

I haven't been doing survival analysis for the last 5 years, but in the 4 years where that was all I did everyday, with clinicians side by side, I never once generated the plot shown and we always used survival curves and hazard rates.

Search the literature for your field and see what's actually being done. But I'm the person that pushes back on reviewers as well when they ask for stupid things...which did seem to surprise several of my colleagues. The reviewers did accept the article without the changes as well