Statalist


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

st: R: Hazard rates with left truncation?


From   "Carlo Lazzaro" <carlo.lazzaro@tin.it>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: R: Hazard rates with left truncation?
Date   Fri, 11 Sep 2009 16:39:22 +0200

Dear Michael,
left truncation (delayed entry) in Stata (9.2 SE in my release) is managed
via  nonparametric, semiparametric (please, invoke -help stcox- from within
Stata) and parametric (please, invoke -help streg - from within
Stata)regression models for survival analysis. 
Assuming a Weibull regression is a wise choice, you may want to try the
following one:
------------------------begin example----------------------------
set obs 20
g id=_n
g In=0 in 1/19
replace In=2 if In==.
g Out=3 in 10/20
replace Out=2 in 5/9
replace Out=1 if Out==.
g death=1 in 1/5
replace death =1 in 15/19
replace death=0 if death==.
g drug=1 in 10/20
replace drug=0 if drug==.
g Risk_time=Out-In
stset Risk_time, failure(death==1)
streg drug, d(weibull)
stcurve, hazard outfile("C:\Michael_hz.dta" , replace)
----------------------end example---------------------------------

Probably more relevant, but for sure you have already check it yourself in
your textbook, the core of the task is choosing what kind of model fits
better with your dataset and your research aim (and why).

Just to top off what above, for more detail on yours and other topics abou
survival analysis, I would refer you to the following:
Cleves MA, Gould WW, Gutierrez R. An Introduction To Survival Analysis Using
Stata. Revised edition. College Station: StataPress, 2004.

Please note that an updated version of the same textbook has been published
in 2008.

HTH and Kind Regards,
Carlo
-----Messaggio originale-----
Da: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] Per conto di Schwartz, Michael
Inviato: venerdì 11 settembre 2009 12.53
A: statalist@hsphsun2.harvard.edu
Oggetto: st: Hazard rates with left truncation?

I want to analyze a dataset that includes some observations with delayed
entry (left truncation). Basically, I follow the textbooks and
everything works. I use time0(time to delayed entry). Stata produces the
output with sts list and I do get the values for survivor function or
failure function. However, what I need is the hazard rate, but I have no
idea how Stata produces this output (taking into account the left
truncation!). I have also tried an alternative way using the origin()
and enter() commands. This does not work either.

Thank you for your help!

Michael Schwartz
Halle Insitute for Urban Economic Research
Michael.Schwartz@iwh-halle.de




*
*   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