data splitting/interoccasion variability

Dear NMUSERS, I have some questions on data splitting/interoccasion variability. I have Carboplatin PK data from 69 Patients from different studies. 9 of these patients were observed on more than 1 cycle (7 Patients 2 cycles, 1 Patient 3 cycles, 1 Patient 4 cycles). On all occasions (cycles) full sampling was performed. 1) Will estimation of IOV be successful with data from only 12 additional cycles? 2) How should the dataset be spitted in order to perform an internal "validation" (e.g. leaving all the data from the 9 patients with several occasions in the model development set + a random sample out of the remaining 60 patients?). I am looking forward to your comments. Best regards, Andreas. ____________________________ Andreas Lindauer University of Bonn Department of Clinical Pharmacy An der Immenburg 4 D-53121 Bonn phone:+49 228 73 5781 fax: +49 228 73 9757