Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down on April 23, and its replacement, is already up and running.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Projected Survival Rates after -stcox-

From   Steve Samuels <>
Subject   Re: st: Projected Survival Rates after -stcox-
Date   Thu, 7 Mar 2013 12:55:23 -0500

Maarten, "survival rate" is customary terminology for a survival
probability. See, e.g.:

Also, when I wrote to the list yesterday, I had forgotten that -stcurve- 
can also plot smoothed estimates of the hazard function. 



Date: Thu, 28 Feb 2013 15:21:45 +0100
From: Maarten Buis <>
Subject: Re: st: Projected Survival Rates after -stcox-

On Thu, Feb 28, 2013 at 2:43 PM, Yuval Arbel wrote:
> I would like to predict the survival rates after -stcox- at the sample
> mean of each 114 periods in my sample.
> The predicted value should be bounded between 0 and 1.

That is not true, a rate is larger than or equal to 0 but not
necesarily less than 1. Whether that is true pretty much depends on
the unit you used to measure time.

> How could I calculate these predictions directly?

If you want this kind of thing than that is a sure sign that you do
not want to estimate a Cox model. The strength and weakness of a Cox
model is that it does not estimate the baseline hazard function. This
means that you cannot make a mistake in the baseline hazard functiton
(the strength) but it also means that it is only designed to estimate
hazard ratios (the weakness). You'll probably want to look at a model
like -stpm2- (type in Stata -findit stpm2-).

Hope this helps,

- ---------------------------------
Maarten L. Buis
Reichpietschufer 50
10785 Berlin
*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index