Incorporate creatinine clearance as a time-varying variable

Post technical questions about making Pmetrics carry out a particular task.
Post Reply
sofiedhaese
Posts: 8
Joined: Sun Jan 22, 2017 4:58 pm

Incorporate creatinine clearance as a time-varying variable

Post by sofiedhaese » Thu Dec 06, 2018 11:31 pm

Dear professor Neely


I am having some trouble adding kidney function as a time-varying (or time dependent) covariate to CL. My patients have an estimated creatine clearance at several timepoints. When I run the model (see below), the result looks good but if I use the command
cov.11, I can see that estimated CL does not change in my patients when creatinine clearance changes...So I was wondering if I did something wrong when adding creatinine clearance (CaG in the model file)?

Thank you

#Pri
V, 1, 30
Cl1, 1, 25
Cl2, 1, 25
Vmax, 10, 4000
Km, 10,750
KPC, 0.001, 5
KCP, 0.001, 5

#Cov
WEIGHT
CaG


#Sec
Clc=Cl*(CaG/112)
Vw=V*(WEIGHT/76)
Ke = Clc/Vw

#Dif
XP(1)=RATEIV(1)-(Vmax/((Km*V)+X(1))+(Clc/Vw)+KCP)*X(1)+(KPC*X(2))
XP(2)=(KCP*X(1))-(KPC*X(2))

#Out
Y(1) = X(1)/V

#Err

User avatar
mneely
Site Admin
Posts: 220
Joined: Tue Nov 08, 2011 11:08 am
Contact:

Re: Incorporate creatinine clearance as a time-varying variable

Post by mneely » Tue Dec 18, 2018 11:09 am

Hi Sofie,

A few of things on your model file.
  • In #COV, you don't need Ke = Clc/Vw. Since you are using differential equations, you can name your variables anything.
  • In #SEC, you set Clc=Cl*(CaG/112), but you haven't defined Cl anywhere.
  • In #PRI, you have Cl1 and Cl2, but they aren't used in any equation.

sofiedhaese
Posts: 8
Joined: Sun Jan 22, 2017 4:58 pm

Re: Incorporate creatinine clearance as a time-varying variable

Post by sofiedhaese » Wed Dec 19, 2018 5:55 am

Dear professor Neely

I've made corrections to the model file. However, it is still not clear to me how I can introduce a time-dependent covariate (mCRCL) in my modelfile?

Thank you,
Sofie

Corrected model file:

#Pri
V, 1, 30
Cl, 1, 25
Vmax, 10, 4000
Km, 10,750
KPC, 0.001, 5
KCP, 0.001, 5

#Cov
OCC
mCRCL
GENDER
AGE
HEIGHT
WEIGHT
CaG
MDRD
SOFA
SCREAT
ALB


#Sec
Clm=Cl*(mCRCL/100)

#Dif
XP(1)=RATEIV(1)-(Vmax/((Km*V)+X(1))+(Clm/V)+KCP)*X(1)+(KPC*X(2))
XP(2)=(KCP*X(1))-(KPC*X(2))

#Out
Y(1) = X(1)/V

#Err
G=1
1,0.1,0,0

User avatar
mneely
Site Admin
Posts: 220
Joined: Tue Nov 08, 2011 11:08 am
Contact:

Re: Incorporate creatinine clearance as a time-varying variable

Post by mneely » Wed Dec 19, 2018 7:34 am

Hi Sofie,

It seems to me that you already have. Since Clm=Cl*(mCRCL/100), as mCRCL changes, so will Clm. Your Cl term is in units of L/h/mCRCL. So it can be constant, but you still have a clearance (Clm) which is indirectly dependent on time through the time-dependence of mCRCL.

sofiedhaese
Posts: 8
Joined: Sun Jan 22, 2017 4:58 pm

Re: Incorporate creatinine clearance as a time-varying variable

Post by sofiedhaese » Thu Dec 20, 2018 4:33 am

Dear professor Neely,

Thank you very much for your answer.
Is it possible in Pmetrics to see the predictions for 'CLm' for every individual patient over a certain course of time? I'd like to see if the values of CLm change within one patient. If I use the cov.* command, I can only see CL but not CLm.

Kind regards,
Sofie

Post Reply