flip-flop kinetics

From: ABoeckmann <alison@c255.ucsf.edu> Date: Fri, 2 Apr 1999 16:04:23 -0800 (PST) Subject: flip-flop kinetics I agree with Nick's comments on Peter's question. I'd like to add that I know of no reason why flip-flop should be any more or less likely when one uses an analytic solution (ADVAN2) vs. a numerical solution (integrating differential equations in ADVAN6). But the predictions from the two ADVAN's will inevitably have tiny numeric differences, and the estimation step *might* follow a slightly different path to the minimum, which *might* lead to a flip-flop with one or both methods. To avoid flip-flop, you can include observations of the drug in the depot (absorption) compartment. Probably you do not have this data. Or, model your parameters so that KA>K E.g., KA = K + a term that must be positive This can be done with both ADVAN2 and ADVAN6. ======================