Jump to content


Photo

high shrinkage solutions


  • Please log in to reply
3 replies to this topic

#1 amashehri

amashehri

    Newbie

  • Members
  • Pip
  • 8 posts

Posted 22 October 2021 - 11:30 PM

Hi, I am trying to build a model to fit a plasma conc for a drug after a single dose. I created a base model with two compartments and Tlag. I got a reasonable fit, but I still get very high shrinkage in one of the Pk parameters (Vd)
Would you please take a look to see if there is any way to decrease the shrinkage value?
I have attached the model below.
Thanks  

Attached Files



#2 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,316 posts

Posted 23 October 2021 - 07:03 AM

Hi Amashehri, 

   I'm not sure if there is anything you can to decrease the shrinkage; it is a diagnostic that tells you the data does not have enough information to estimate random effects on this parameter.  I think with the second compartment parameters identifiability can be an issue, since even when you remove random effects for  v2, shrinkage is still a little high for CL2 

 

    I don't think considering covariates would help, but I'd be interested to hear others experiences.

Attached Thumbnails

  • cl2.jpg
  • ran_eff.jpg


#3 amashehri

amashehri

    Newbie

  • Members
  • Pip
  • 8 posts

Posted 05 November 2021 - 02:49 PM

Thanks Simon for your reply 

 

 



#4 XQLI

XQLI

    Member

  • Members
  • PipPip
  • 15 posts

Posted 27 February 2022 - 02:58 AM

amashehri, 在 05 十一月 2021 - 10:49 下午, 说:

谢谢西蒙的回复

How did you end up dealing with high shrinkage values? The shrinkage value of one of my parameters was not high before the addition of the covariable model, but after the addition of the covariable model, the shrinkage value of the parameter was close to 0.98. Do you have any suggestions?






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users