Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

Re: st: survival curve for a stratified analysis using matched data


From   Steve Samuels <[email protected]>
To   [email protected]
Subject   Re: st: survival curve for a stratified analysis using matched data
Date   Tue, 26 Nov 2013 16:53:57 -0500

You are very welcome., Katie. My use of the word "unmatched" was
incorrect; one can speak of a "matched" or "unmatched" analysis, but the
data, in fact, are well-matched. That's the point of the Ho article,
which I urge you to read. 
Steve


On Nov 26, 2013, at 9:59 AM, <[email protected]> <[email protected]> wrote:

Thanks for reference and link.  I suspected there wasn't a way to create one overall survival curve from stratified data, but wanted to make sure I wasn't mistaken.  I will present the unadjusted, unmatched survival curves since my strata are too small to present individually.

Thanks!
Katie
________________________________________
From: [email protected] [[email protected]] On Behalf Of Steve Samuels [[email protected]]
Sent: Monday, November 25, 2013 6:59 PM
To: [email protected]
Subject: Re: st: survival curve for a stratified analysis using matched data

I should have said "Do the unmatched analysis and present the curves from that".

S.

The standard theory after -stcox- will estimate a survival curve for
every stratum. Is that what the reviewer is asking for? If so, then

. predict surv, basesurv
. sort group _t
. twoway connect surv _t, by(group)

If your groups are too small for this (they should be if the matching is
well-done), then present curves for the unmatched data. See Ho et al.
2007.

Reference:

Ho, Daniel, Kosuke Imai, Gary King, and Elizabeth Stuart. 2007. Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference. Political Analysis 15: 199–236.

Link at: http://gking.harvard.edu/category/research-interests/methods/causal-inference


Steve

Steven J. Samuels
18 Cantine's Island
Saugerties NY 12477 USA


On Nov 25, 2013, at 4:46 PM, <[email protected]> <[email protected]> wrote:

Hi,

I am working on a survival analysis using a 1:2 variably matched dataset.  A reviewer is asking for a survival curve to accompany the proportional hazards model I've used and Stata is not allowing me to use the "stcurve, survival" command on a stratified model.  I need to keep the model stratified on matched group to account for the matched pairs.

My code:
xi: stcox i.rx_tzd, str(group) vce(robust)
stcurve, survival at1(_Irx_tzd_1=1)  at2(_Irx_tzd_1=0)
** error message from Stata:   "stcurve not allowed after stcox, strata()"

Is there any way to create a survival curve from a stratified analysis in Stata?
I am using Stata/SE version 12.1.

Thanks for your help,

Katie Lynch

This message is intended for the use of the person(s) to whom it may be addressed. It may contain information that is privileged, confidential, or otherwise protected from disclosure under applicable law. If you are not the intended recipient, any dissemination, distribution, copying, or use of this information is prohibited. If you have received this message in error, please permanently delete it and immediately notify the sender. Thank you.

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index