Dear Moleps,
to my real shame in my previous reply I have forgotten to add to the
references the following Sven Juul's, really helpful textbook An
Introduction to Stata for Health Researchers. College Station, TX: Stata
Press, 2006. The 2nd edition is now available (please, see www.stata.com).
Among other interesting topics, I have found it particularly straight to the
point for learning how to - stset - tha data set before performing survival
analysis.
Kind Regards,
Carlo
-----------------------------------------------------------
Dear Moleps:
-----Original message-----
Da: [email protected]
[mailto:[email protected]] Per conto di moleps islon
Inviato: marted� 23 settembre 2008 21.51
A: [email protected]
Oggetto: st: survival analysis
Hello,
I've got several questions where I would greatly appreciate feedback
1) Reading the medical literature most retrospective survival analysis
is done with Kaplan Meier curves. Is it not more appropriate to use
cox graphs where other covariates have been adjusted for?
1_answer)The main concern regarding semiparametric Cox proportional hazards
model (and the related regression) is the holding of the proportional
hazards statement throughout the follow-up. This statement is violated
whenever the survival curves are not parallel (please, see - help stphplot -
and -help stcoxkm - in Stata 9.2; I do not know the analogous commands in
Stata 10.1).
Cleves MA, Gould WW, Gutierrez RG. An Introduction to survival analysis
using Stata.2nd ed. College Station: Stata Press, 2006 offers a thorough
covering of this and related topics.
Moreover, two other relevant contributors of the Statalist - Maarten Buis
(http://home.fsw.vu.nl/m.buis/) and Stephen Jenkins
(http://www.iser.essex.ac.uk/teaching/degree/stephenj/ec968/index.php.)
published really interesting papers as well as teaching-notes on the topics
you are interested in.
2) Some report both univariate and then multivariate data on their
variables. What is the correct thing to do?
2_answer)The wisest (but not that supporting) answer would probably be: it
depends on the data you can rely on and the aim(s) of your research project.
3)I'm unable to make my individual graphs in stata v10.1 appear with
diamonds, triangles and squares instead of the multitude of colors. I
can change the line pattern, but I cant find a way to report the lines
with diamons etc. Surely this must be possible-But despite several
days of working on this I cant manage..
3-answer) When compared to other Stata routine, which are really quick and
easy to learn, I find the graphic side of the matter particularly
user-demanding; however, please see - [G] graph intro - in Stata manual.
4) is there an option to make a table that I can put beneath the graph
delineating the number of patients in the different groups at
different timepoints?
4_answer) In Stata 9.2 this option is provided by the following:
SJ-4-4 st0058_2 . . . . . . . . . . . . . . . Software update for
stsatrisk
. . . . . . . . . . . . J. M. Linhart, J. S. Pitblado, and J.
Hassell
(help stsatrisk if installed)
Q4/04 SJ 4(4):491
bug fix for stsatrisk.ado
If I recall it correctly, Stata 10.1 has an ad hoc in-built command for this
option.
HTH and Kind Regards,
Carlo
Regards
Moleps
*
* 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/
