Assigning occasions to grouped data

Dear all, I would like to obtain some feedback from the group about how to assign occasions for the estimation of interoccasion variability (IOV). I have three oral PK datasets: 1. concentration-time profiles after multiple doses (first dose at 0h; dosing interval of 24h) from 0 to 96h (56 subjects); 2. concentration-time profiles after multiple doses (first dose at 0h; dosing intervals of 6, 8 or 12h) from 0 to 72h (44 subjects); and 3. concentration-time profiles after a single dose (at 0h) from 0 to 24h (20 subjects). What I would like to do is to combine these three datasets and perform a popPK analysis. For this purpose, should I define an occasion based on: a. time (e.g. occasion1 for any samples taken during 0 to 24h; occasion2 for any samples taken during 24 to 48h; ...); b. ingested dose (e.g. occasion1 includes PK profiles from 0 to 24h for dataset #1, PK profiles from 0 to 6h, 0 to 8h or 0 to 12h for dataset #2, and PK profiles from 0 to 24h for dataset #3; occasion2 denotes PK profiles from 24 to 48h for dataset #1, and PK profiles from 6 to 12h, 8 to 16h or 12 to 24h for dataset #2; ...); or c. others? Based on preliminary analysis of dataset #3, the average elimination half-life is 3h. Thank you in advance for your kind advice. Best wishes, Kok-Yong Seng DSO National Laboratories Republic of Singapore