Statalist


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

Re: st: comparing -ltable- and -sts list-


From   Michael McCulloch <mm@pinest.org>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: comparing -ltable- and -sts list-
Date   Sun, 2 Nov 2008 21:25:30 -0800

Thank you Philip for the very helpful explanation.
Michael


Michael


I think that to perform a test *at a particular time*, as opposed to an overall (logrank style) test, using the K-M estimator you need to

(i) estimate the S(t) and the Greenwood variance for each group at the chosen time
(ii) perform a z test


Step (i) can be done by sts list:

. sts list , by(drug) at(0,2)

         failure _d:  died
   analysis time _t:  studytime

              Beg.                      Survivor      Std.
Time Total Fail Function Error [95% Conf. Int.]
-------------------------------------------------------------------------------
drug=1
       0         0        0              1.0000         .          .         .
       2        18        3              0.8500    0.0798     0.6038    0.9490
drug=2
       0         0        0              1.0000         .          .         .
       2         0        0              1.0000         .          .         .
drug=3
       0         0        0              1.0000         .          .         .
       2         0        0              1.0000         .          .         .
-------------------------------------------------------------------------------


Note the syntax of the -at()- option. Had I just put -at(2)- Stata would have given me its 2 chosen times, not necessarily t=2. (Try it and see!)

Then you can retrieve S(t=2) for each of the two groups of interest and the SEs and thus calculate the Greenwood estimate of the variance of the difference in S(t). The z test follows...

Step (ii):

    z = (S1(t=2) - S2(t=2)) /  [  sqrt(Var(S1(t=2) + Var(S2(t=2))]

where Var(S(t=2) is the square of the SE for the respective group at the desired time.

whence you can use the -normal()- function to get your P value.


Doesn't look like your particular data (at t=2) will support this test - we only get an estimate for drug=1. But the above is the general idea. Someone else may know of a more direct way.

Phil

At 11:33 AM 3/11/2008, you wrote:
Thanks Phil.

Is there a way to use -sts list- to statistically compare the 2-year survival rates between two groups? I notice that the -compare- option simply places them next to each other without a statistical test.

Michael

Michael

use the   -noadjust-   option  on  -ltable-

from -help ltable- :

noadjust suppresses the actuarial adjustment for deaths and censored observations. The default is to consider the adjusted number at risk at the start of the interval to be total at the start minus (the number dead or censored)/2. If noadjust is specified, the number at risk is simply the total at the start, corresponding to the standard Kaplan-Meier assumption. noadjust should be specified when using ltable to list results corresponding to those
        produced by sts list.


Phil


At 09:51 AM 3/11/2008, you wrote:
Dear Statalist members,
I had the impression that both -ltable- and -sts list- would list the survivor function per failure period. However, they give slightly different answers. In the example below, for drug3 -ltable- shows survival at 24 months as 77.14%, whereas -sts list- says it's 64.29%

sysuse cancer.dta, clear
sts list, by(drug)
ltable _t died, by(drug)

Could anyone help me to clarify this?

--

Best wishes,
Michael McCulloch



Pine Street Foundation
124 Pine St., San Anselmo, CA 94960-2674
Tel:    (415) 407-1357
Fax:    (415) 485-1065
mcculloch@pinestreetfoundation.org
www.pinestreetfoundation.org
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/

Philip Ryan
Professor,
Discipline of Public Health

Director, Data Management & Analysis Centre

Associate Dean (IT)
Faculty of Health Sciences

postal address:
Discipline of Public Health
Mail Drop DX650 511
University of Adelaide 5005
South Australia

location:
Level 6, Room 6-18
Bice Building
Royal Adelaide Hospital
North Terrace
Adelaide

tel +61 8 8303 3570
fax +61 8 8223 4075
http://www.public-health.adelaide.edu.au/
CRICOS Provider Number 00123M
-----------------------------------------------------------
This email message is intended only for the addressee(s)
and contains information that may be confidential and/or
copyright. If you are not the intended recipient please
notify the sender by reply email and immediately delete
this email. Use, disclosure or reproduction of this email
by anyone other than the intended recipient(s) is strictly
prohibited. No representation is made that this email or
any attachments are free of viruses. Virus scanning is
recommended and is the responsibility of the recipient.

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

Philip Ryan
Professor,
Discipline of Public Health

Director, Data Management & Analysis Centre

Associate Dean (IT)
Faculty of Health Sciences

postal address:
Discipline of Public Health
Mail Drop DX650 511
University of Adelaide 5005
South Australia

location:
Level 6, Room 6-18
Bice Building
Royal Adelaide Hospital
North Terrace
Adelaide

tel +61 8 8303 3570
fax +61 8 8223 4075
http://www.public-health.adelaide.edu.au/
CRICOS Provider Number 00123M
-----------------------------------------------------------
This email message is intended only for the addressee(s)
and contains information that may be confidential and/or
copyright. If you are not the intended recipient please
notify the sender by reply email and immediately delete
this email. Use, disclosure or reproduction of this email
by anyone other than the intended recipient(s) is strictly
prohibited. No representation is made that this email or
any attachments are free of viruses. Virus scanning is
recommended and is the responsibility of the recipient.

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



© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index