Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

R: st: re Kaplan-Meier survival analysis model


From   "Carlo Lazzaro" <[email protected]>
To   <[email protected]>
Subject   R: st: re Kaplan-Meier survival analysis model
Date   Wed, 23 Jan 2008 12:54:28 +0100

Dear Ziad, provided that I have understood well your question, I suppose you
may model it via Kaplan-Meier estimates (including censoring). 

Please take a look at what follows below:
-------------------------------begin
example-------------------------------------------------------------
set obs 20
g id=_n
replace dummy_res=0 in 1/10
replace dummy_res=1 in 11/20
g In=0
g Out = 1
replace Out=0 in 1
replace Out=0 in 20
g Conv_HIV=1 in 6/16
replace Conv_HIV=0 if Conv_HIV==.
g risk_time= Out- In
stset risk_time, id(id) failure(Conv_HIV==1)
sts list, by(dummy_res) failure
stcox dummy_res
--------------------------------end
example------------------------------------------------------------ I would
suggest you to test whether the prerequisites for the Cox proportional risk
model hold before relying on the Hazard ratio (and related p-value) of the
Cox regression. Hope this helps and Kind Regards,

Carlo
-----Messaggio originale-----
Da: [email protected]
[mailto:[email protected]] Per conto di Ziad El-Khatib
Inviato: mercoled´┐Ż 23 gennaio 2008 12.03
A: [email protected]
Oggetto: Re: st: re Kaplan-Meier survival analysis model

Thank you!
well my intention is to see how the progress is over time.
to be clearer and more specific, to compare time took group of HIV
patients with resistance at baseline versus group of HIV patients
without resistance at baseline to show treatment failure.
 option 1) treatment failure due to baseline resistance
 option 2) treatment failure due to resistance development during treatment
 option 3) no treatment failure at all
 option 4) patient dropped out (maybe this can not be added here?)


Thank you and best regards
ziad

On Jan 23, 2008 2:55 AM, Maarten buis <[email protected]> wrote:
> --- Ziad El-Khatib <[email protected]> wrote:
> > My understanding to KM survival model is to have the variable related
> > to the 'survival' event to be coded as (0: as nothing happened or 1:
> > as the event did happen) For example if we are monitoring death: 1)
> > died, 0) did not die.
> >
> > I am assessing variable for sickness, to see if there is difference
> > in time to show between
> >   option 1) patients who were sick before treatment and became sicker
> > after treatment
> >
> >   option 2) patients who were not sick before treatment but became
> > sicker after treatment
> >
> >  option3) And there are patients who did not show any sickness at
> > all.
> >
> >
> > would i code them as follow:
> > option1 : code 1
> > option2: code 2
> > option3: code 0
>
> Survival models model the duration till an effent, so that does not
> seem to apply in your case. I would suppose that you are interested in
> whether or not someone became sicker after treatment, and you want to
> control for whether or not someone was already sick. I would create two
> dummies:
> sicker: 1) got sicker, 0) not sicker
> sick: 1) was sick, 0) was not sick.
>
> and than type:
> logit sicker sick
>
> Hope this helps,
> maarten
>
> -----------------------------------------
> Maarten L. Buis
> Department of Social Research Methodology
> Vrije Universiteit Amsterdam
> Boelelaan 1081
> 1081 HV Amsterdam
> The Netherlands
>
> visiting address:
> Buitenveldertselaan 3 (Metropolitan), room Z434
>
> +31 20 5986715
>
> http://home.fsw.vu.nl/m.buis/
> -----------------------------------------
>
>
>       __________________________________________________________
> Sent from Yahoo! Mail - a smarter inbox http://uk.mail.yahoo.com
>
>
> *
> *   For searches and help try:
> *   http://www.stata.com/support/faqs/res/findit.html
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>



-- 
Ziad El-Khatib
AIDS Unit
National Institute for Communicable Diseases (NICD)
1 Modderfontein Road
Private Bag X4, Sandringham, 2131
Johannesburg, South Africa
Mobile: +27 (0) 72-52 39 716
Phone: +27 (0) 11-386 6433
Fax (NICD): +27 (0)11 386 6453
Fax (alternative): +27 (0)866 18 2871
http://www.nicd.ac.za/

Division of International Health (IHCAR)
Karolinska Institutet
www.phs.ki.se/ihcar
*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index