From: "Martin, Tomas" <tmj@cptcnt2.cvm.ncsu.edu>
Subject: Prediction intervals
Date: Fri, 19 Nov 1999 17:58:10 -0500
Dear nm users:
I would appreciate your help in the following matter. Is there an easy way to compute prediction intervals for the predicted concentrations obtained after fitting a model with NONMEM?.
Thanks.
Tomas Martin
Prediction intervals
Date: Sun, 21 Nov 1999 08:12:51 -0800
From: Lewis Sheiner <lewis@c255.ucsf.edu>
Subject: Re: Prediction intervals
Yes.
Simulate the desired prediction under the desired design for say 500 simulated patients and use the percentiles of the simulated distribution to estimate the desired interval.
This prediction interval is CONDITIONAL on the model; it does not involve model uncertainty. To obtain an interal with model uncertainty is more difficult.
LBS.
--
Lewis B Sheiner Professor, Lab. Med., Biopharmaceut. Sci, Med.
Box 0626 - UCSF 415-476-1965 (voice)
San Francisco, CA 415-476-2796 (fax)
94143 lewis@c255.ucsf.edu