Hi, I have some tissue concentration data I am trying to use the sparse sampling function for- and keep getting negative estimates for the AUCs. I have concentrations at 2 PK visits (first dose and steady state) in 3 tissue types. At each of 6 time points, I have concentrations from 2 different subjects (each subject contributed one concentration per PK visit), and at one time point at first dose PK, there are data from 3 subjects. I’m not sure what the issue is- I am mapping the time to the nominal time column, and it seems like it should be pretty straightforward following the instruction manual. Or can I only do this if I have multiple samples per subject and not just one? Any explanations or suggestions? Thanks, Julie
Julie. For NCA you will need to split the data to each visit, since they are different dosing statuses Plus you’ll be splitting by tissue type, with the amount of data you describe - it won’t hurt to use the sparse sampling option but it probably won’t add much to your analysis. t o understand your ‘issue’ I think an example project plus error message would help Simon.
I split them up, and that took care of the negative AUC issue, but I’m not sure what it’s doing, as I’m getting negative AUMC/MRT values, and for the PK2 (several days of dosing) the Cmin/Cmax/Tmax values are totally off (and the same for each tissue). It almost looks like I have the time/concentration columns switched in places, but they are mapped correctly. It is estimating concentrations for time values out to 10^4 hours. My workbook is attached. [file name=Dumond_NCA_example.phxproj size=1086536]/extranet/media/kunena/attachments/legacy/files/Dumond_NCA_example.phxproj[/file]Dumond_NCA_example-20120323.phxproj (1.04 MB)
Ok, let me try this again…my earlier message looks blank. Breaking up the data resolved the negative AUC issue…but I still have negative AUMC values (and thus, negative MRT values). It looks to me like it’s extrapolating out to times on the order of 10^4, and in the PK2, the Tmax, Cmax, Tmin, Cmin values are not reflective of the data set. Sample project attachedDumond_NCA_example-20120323.phxproj (1.04 MB)
Julie - I think there are a couple of things to look at here, firstly if Day 1 is single dose then Tau should be left blank; 
Secondly for the Sparse sampling tool you should use Nominal time but there seems to be significant delays when we compare the data with the nominal time in the first graph compared to actual time after 
I’ll try and look at your project a little more tomorrow as despite the above I think there is something strange going on in the NCA object itself. Simon.
Hi Julie I worked out the problem, we discovered a critical issue QC10833: “NCA Sparse Sampling: Columns mapped as Subject ID cannot contain spaces in their values” last year. We have now fixed this in 6.3 but the workaround for this project is to remove the spaces in the ID column. I used a the column transformation; replace(subject_id,5,1,“”) which means it Looks at variable SUbject_id, finds the fifth character (which is where the space is), and for 1 character, replaces it with “nothing” I then simply re-mapped your NCA and used the newsubID I created instead and it worked fine. 
