turn on suggestions

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

Showing results for

Find a Community

- Home
- /
- SAS Programming
- /
- General Programming
- /
- Kaplan-Meier plot for up to two recurrent time-to-...

Topic Options

- Subscribe to 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
- Subscribe to RSS Feed
- Highlight
- Email to a Friend
- Report Inappropriate Content

07-18-2014 05:50 PM

Dear fellow SAS users,

I have a dataset on time to occlusions, where half of the patients had bilateral treatments with possibly bilateral occlusions that are correlated. How can I make the Kaplan-Meier plot for time to occlusions, taking into account those bilateral occlusions within patients?

I've found some literature about survival analysis on recurrent events, the problem here is that not all patients have recurrent events, and the number of recurrent events is up to two.

Any suggestions to SAS code or literature will be greatly appreciated!

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

07-18-2014 06:28 PM

Maybe totally off base here, especially since I'm having a hard time following the med term on a late Friday afternoon.

SAS FAQ: How to model repeated events survival analysis using proc phreg?

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

07-22-2014 10:39 AM

Thanks for your reply Reeza! Allow me to rephrase here:

I have a dataset on time to events, where half of the samples could develop up to two events and therefore have two time points that are correlated. How can I make the Kaplan-Meier plot for time to events, taking into account those two events within samples?

I've found some literature about survival analysis on recurrent events, the problem here is that not all samples have recurrent events, and the number of recurrent events is up to two.

Any suggestions to SAS code or literature will be greatly appreciated!

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

07-22-2014 10:46 AM

I don't know that a KM plot can account for that.

In terms of analysis, the reason you use survival analysis is because not all observations may have an event so that shouldn't be an issue. From your description a counting process model does seem like it would fit the analysis requirements.

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

07-22-2014 10:59 AM

Appreciate your response! Now the data is highly unbalanced (only half of the samples could develop two events), and the number of events is only up to two. Do you think a counting process would still work in this scenario?

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

07-22-2014 11:15 AM

What's the research question?

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

07-22-2014 12:55 PM

The question is: How to plot Kaplan-Meier with the type of data I mentioned in previous posts, accounting for correlation between two events within the same samples.

Thanks for your help!