Hello Andrzej
Yes, it should be cenind(cstat), not censtat(cstat). Sorry for my mistake.
And, by -hazard-, I mean a variable that is supposed to predict hazard rate, such as the predicted hazard ratio, also known as the relative hazard. This variable might be calculated after a Cox regression by using the -predict- command with the -hr- option.
Note that, if you want Harrell's c to have confidence limits that can be taken seriously, then the hazard ratio should be derived (using out-of-sample prediction) from a Cox model fitted to a different dataset from the dataset in which you estimate Harrell's c. For instance, if you have a large sample of subjects, then you can split that sample into a training set and a test set, fit the Cox regression to the training set, bring the test set into the memory, calculate the hazard ratio in the test set, and then calculate confidence limits for Harrell's c (or Somers' D) in the test set. And, of course, you can calculate multiple hazard ratios from multiple Cox regressions, and compare their predictive power using confidence intervals for the differences between Harrell's c indices.
I hope this helps.
Best wishes
Roger
Roger B Newson BSc MSc DPhil
Lecturer in Medical Statistics
Respiratory Epidemiology and Public Health Group
National Heart and Lung Institute
Imperial College London
Royal Brompton Campus
Room 33, Emmanuel Kaye Building
1B Manresa Road
London SW3 6LR
UNITED KINGDOM
Tel: +44 (0)20 7352 8121 ext 3381
Fax: +44 (0)20 7351 8322
Email: [email protected]
Web page: http://www.imperial.ac.uk/nhli/r.newson/
Departmental Web page:
http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/popgenetics/reph/
Opinions expressed are those of the author, not of the institution.
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Andrzej Niemierko
Sent: 30 September 2009 23:17
To: Stata
Subject: Re: st: RE: stcrreg and Harrell's c index
Roger,
I am getting "option censtat() not allowed" message. Should it be
-cenind()-, maybe? Also, by -hazard- you mean subhazard ratio or cumulative
subhazard function?
Andrzej
On 9/11/09 5:40 PM, "Newson, Roger B" <[email protected]> wrote:
> I don't know why Harrell's c is not available afert parametric survival model.
> However, it is definitely available after the -somersd- command, downloadable
> from SSC. If in Stata you type
>
> somersd lifetime hazard, censtat(cstat) transf(c) tdist
>
> and -lifetime- is a lifetime variable, -hazard- is a predictor of hazard rate,
> and -cstat- is a censorship status variable (positive for right-censored
> lifetimes, negative for left-censored lifetimes and zero for uncensored
> lifetimes), then you will get an estimate for Harrell's c, which is stored in
> e(b).
>
> I hope this helps. Let me know if you have any further queries.
>
> Best wishes
>
> Roger
>
>
> Roger B Newson BSc MSc DPhil
> Lecturer in Medical Statistics
> Respiratory Epidemiology and Public Health Group
> National Heart and Lung Institute
> Imperial College London
> Royal Brompton Campus
> Room 33, Emmanuel Kaye Building
> 1B Manresa Road
> London SW3 6LR
> UNITED KINGDOM
> Tel: +44 (0)20 7352 8121 ext 3381
> Fax: +44 (0)20 7351 8322
> Email: [email protected]
> Web page: http://www.imperial.ac.uk/nhli/r.newson/
> Departmental Web page:
> http://www1.imperial.ac.uk/medicine/about/divisions/nhli/respiration/popgeneti
> cs/reph/
>
> Opinions expressed are those of the author, not of the institution.
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of Andrzej Niemierko
> Sent: 11 September 2009 14:46
> To: Stata
> Subject: st: stcrreg and Harrell's c index
>
> Harrell's c index is not available as a postestimation statistic after
> stcrreg (or any parametric survival model). Is there a theoretical reason
> for this? If not, how can I compute it after stcrreg?
> Andrzej
>
>
>
>
>
>
> The information in this e-mail is intended only for the person to whom it is
> addressed. If you believe this e-mail was sent to you in error and the e-mail
> contains patient information, please contact the Partners Compliance HelpLine
> at
> http://www.partners.org/complianceline . If the e-mail was sent to you in
> error
> but does not contain patient information, please contact the sender and
> properly
> dispose of the e-mail.
>
> *
> * 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/
*
* 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/
