FW: Re: Trough concentrations and ADDL

From: Alison Boeckmann [mailto:[email protected]] Sent: Wednesday, February 29, 2012 4:10 PM To: Bachman, William; Ludden, Thomas; Bauer, Robert Subject: Fwd: Re: [NMusers] Trough concentrations and ADDL Bill, I have tried sending this message to nmusers several times. It never got through. I can't believe it is too long. Can you try sending it for me? ----- Original message ----- From: "Alison Boeckmann" <[email protected]> To: "Elisabet Størset" <[email protected]>, [email protected] Date: Mon, 27 Feb 2012 11:21:25 -0800 Subject: Re: [NMusers] Trough concentrations and ADDL This is an answer to an email from Jan 11. Dear Elisabet, Here are some remarks on your data set, and also a suggestion for handling differing amounts (e.g. TDM) with ADDL/II (1) With PREDPP, a data item called AMT is needed, e.g., $INPUT ...DOSE=AMT (2) You cannot use comma (,) as decimal point. E.g., 6,1 should be 6.1 (3) DATE in 8th record is clearly a typo: 11-01-2001 should be 11-01-2008 (4) With ADDL, there is no need to interrupt the dosing pattern for an observation event. Specify as many ADDL doses as needed till the next change in the pattern; PREDPP will continue the dosing despite intervening observation events. (5) With ADDL and II, if the dosing interval II remains constant and only the amount changes, then a dose amount of 1 can be specified, and F1 can be used to change the effective dose amounts. Fragment of a control file: $INPUT ID DAT1=DROP TIME WT EVID C=DV DOSE AMT II ADDL .... $PK IF (DOSE> 0) SAVEDOSE=DOSE ; a technique for saving TDM amount F1=SAVEDOSE Fragmanent of the data file: ID DAT1=DROP TIME WT EVID C=DV DOSE AMT II ADDL 1 01.01.2008 20 10 1 0 4 1 12 99999 1 03.01.2008 8 . 0 6.1 0 0 0 0 1 04.01.2008 8 . 0 7.6 0 0 0 0 1 04.01.2008 20 . 2 0 3 0 0 0 1 06.01.2008 8 . 0 5.4 0 0 0 0 (6) It is still possible to model bio-availablity, e.g. F1=SAVEDOSE*THETA(2) (7) Note that the use of SAVEDOSE is similar to the well-known technique ` for saving a covariate value. E.g., IF (WT > 0) SAVEWT=WT K=THETA(1)*SAVEWT*EXP(ETA(1)) On Wed, Jan 11, 2012, at 03:47 PM, Elisabet Størset wrote: I am a new NONMEM user and have a simple question regarding the input file. I am analyzing trough concentrations in a multiple oral dose study. 20 trough concentrations are available per patient over a period of 10-12 weeks. The dose is frequently changing due to TDM. I wonder how it is recommended to code the dosing history in the inputfile effectively in this situation. I'm considering using ADDL, but am unfamiliar with it and unsure if I am using it correctly. This is a selected part of the inputfile: ID DAT1=DROP TIME EVID C=DV DOSE II ADDL 1 01.01.2008 20 1 0 4 12 4 1 03.01.2008 8 0 6,1 0 0 0 1 04.01.2008 8 0 7,6 0 0 0 1 04.01.2008 8 1 0 4 0 0 1 04.01.2008 20 1 0 3 12 12 1 06.01.2008 8 0 5,4 0 0 0 1 08.01.2008 8 0 5,6 0 0 0 1 11.01.2001 8 0 4,4 0 0 0 1 11.01.2008 8 1 0 3 0 0 1 11.01.2008 20 1 0 3,5 12 20 1 13.01.2008 8 0 5,0 0 0 0 Is this similar to the way other nmusers use the ADDL item? Will NONMEM analyze the above data file with the same results, compared to this more simple, but less effective code (only first records): ID DAT1=DROP TIME EVID C=DV DOSE 1 01.01.2008 20 1 0 4 1 02.01.2008 8 1 0 4 1 02.01.2008 20 1 0 4 1 03.01.2008 8 0 6,1 0 1 03.01.2008 8 1 0 4 1 03.01.2008 20 1 0 4 1 04.01.2008 8 0 7,6 0 1 04.01.2008 8 1 0 4 1 04.01.2008 20 1 0 3 1 05.01.2008 8 1 0 3 1 05.01.2008 20 1 0 3 When doing the top approach, I have to calculate the number of doses to fill the ADDL item, which can be complicated if the time before the dose is changed is several weeks. This could give errors. Is there a way NONMEM could understand that "the patient will use this dose at 12 hour intervall until a new dose record is stated"? Thank you in advance for helping! Kindly regards Elisabet Størset, new nonmemuser. -- Alison Boeckmann [email protected] -- Alison Boeckmann [email protected]