st: RE: RE: statistical test and sensitivity analysis for matched pairs with censoring

["Ariel Linden, DrPH" <ariel.linden@gmail.com>]

Sho, 

I suggest you read chapter 4 in Rosenbaum PR.Observational Studies (2nd
edition). Springer, New York, 2002.

In particular, on page 134, there is a section that describes censored data.

I hope this helps

Ariel


Date: Tue, 28 Dec 2010 10:29:40 +0900
From: "Shoryoku Hino" <shoryok@ninus.ocn.ne.jp>
Subject: st: RE: statistical test and sensitivity analysis for matched pairs
with censoring

Dear Joseph,

Thank you for your reply and kind advice. I should have made my question
clearer.
All I want to know is the test for survival data in case of paired sample.

For example, if it were not survival data, we would use signed rank test for
continuous variable in pared sample instead of unpaired t-test or rank sum
test. I would use McNemar test for proportion instead of chi square test.
Rosenbaum's sensitivity test could be applicable in these cases.

In my case, survival data, I think there might be a better, more powerful
test for paired sample than log-rank test or Wilcoxon test.
I would like you to give me any suggestion.

Sho
  

- -----Original Message-----
From: Joseph Coveney [mailto:jcoveney@bigplanet.com] 
Sent: Sunday, December 26, 2010 8:08 PM
To: statalist@hsphsun2.harvard.edu
Cc: Shoryoku Hino
Subject: Re: statistical test and sensitivity analysis for matched pairs
with censoring


Shoryoku Hino wrote:

I am using Stata11 and working on the observational data.
I produced the one-to-one matched pairs using PSMATCH2 to compare the
treatment A with B in terms of maintenance effect.
The primary outcome is the time to recurrence of the disease.
There are two questions I would like to ask you.
1) What is the appropriate statistical test in this situation? RF.Woolson
seems to recommend censored-data version of signed rank test. But I have no
idea how we can do it on Stata. 
2) How to perform sensitivity analysis to determine the magnitude of hidden
bias in this case?

I wonder if you could help me to know the way to operate Stata for these
purpose.

-
----------------------------------------------------------------------------
- ----

1) Help files for Stata's official survival time commands can be seen by
typing

    help st

at the command line.  Stat has a selection of tests for censored survival
time data.  The help file for these tests can be seen by typing

    help sts_test

at the command line.  I don't know whether the particular one you mention is
among the ones that Stata offers.  

Stata also has modeling commands for this type of data.  You can click on
the hyperlinks in the viewer window that displays after typing "help st" to
see more about these modeling commands.

There are also user-written commands for survival time or censored data.
You can see more about these by typing

    findit survival

and

    findit censored

at the command line, and then clicking on the hyperlinks that appear in the
viewer window that pops up.

2) I don't understand your question and so can't offer a specific answer to
it, but in general you can use Monte Carlo simulation in order to explore
the properties of a method.  To see Stata's help file for Monte Carlo
simulation, type

    help simulate

at the command line.

Joseph Coveney

*
*   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/