NONMEM Users Network Archive

Hosted by Cognigen

Re: Categorical covariates

From: Nick Holford <n.holford>
Date: Thu, 14 Aug 2008 22:52:03 +1200


What kind of dependent variable do you have in your model?

If you are doing a PK analysis then my guess its probably not worth the
effort to see if tumour type is a covariate unless you have lots and
lots of patients with each tumour type (e.g. 100/tumour). A more
informative covariate might be a performance score like ECOG which you
might have in most patients without having to consider a wide spectrum
of tumour types.

If you are doing something more interesting like looking at tumour size
and response to treatment then you may be able to pool the tumour types
into groups based on biological origin and perhaps learn something about
tumour growth. See Tham et al. 2008 for an example of a tumour growth

Best wishes,


Tham LS, Wang L, Soo RA, Lee SC, Lee HS, Yong WP, et al. A
pharmacodynamic model for the time course of tumor shrinkage by
gemcitabine + carboplatin in non-small cell lung cancer patients. Clin
Cancer Res. 2008;14(13):4213-8.

Hudachek,Susan wrote:
> Greetings All! Does anyone have any suggestions on how to add in cancer type (roughly 10 different tumor types) as a covariate? Thanks for any and all input!
> Susan
> Susan Hudachek, M.S., Ph.D.
> Animal Cancer Center
> Veterinary Teaching Hospital
> Colorado State University
> 300 West Drake Road
> Fort Collins, CO 80523-1620
> PHONE: (970) 219-7599
> FAX: (970) 297-1254
> EMAIL: Susan.Hudachek

Nick Holford, Dept Pharmacology & Clinical Pharmacology
University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand

Received on Thu Aug 14 2008 - 06:52:03 EDT

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to:

Once subscribed, you may contribute to the discussion by emailing: