NONMEM Users Network Archive

Hosted by Cognigen

RE: Your suggestions/thoughts needed on allometric base or final model

From: Stephen Duffull <stephen.duffull>
Date: Sat, 12 Jul 2008 08:21:29 +1200

Atul

> It is always a good idea to estimate the allometric
> coefficient if you have adequate (weight ranges, PK sampling
> etc) data collected.

An interesting point. If you estimate the allometric coefficient then it
will almost always give a better fit to your data as your model has more
degrees of freedom. However if you do estimate it you will invariably get
it wrong. Many studies would be underpowered to detect this parameter and
estimating it will therefore tend to find more extreme values (either bigger
or smaller) since these are the only values that will be detected at your
predetermined alpha-level.

The question remains:
Do you go with biology and fix the coefficient at 0.75 for CL? or
Do you go with your data and estimate the coefficient?

In the latter setting you run the risk of describing your data too well at
the cost of generality to other data sets. In the former setting you may
inflate your remaining random BSV with noise from weight ...

I'm sure there is no absolute answer here.

Steve
PS Body weight is always important - we just can't always detect it.
Received on Fri Jul 11 2008 - 16:21:29 EDT

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to: nmusers-request@iconplc.com.

Once subscribed, you may contribute to the discussion by emailing: nmusers@globomaxnm.com.