OD has always struck me as a circular argument. I know the parameter estimates, and now I’m trying to find a design to re-estimate the parameters. Please comment. Also, what if you don’t have a good estimate of the parameters a priori?

–Workshop participant from MetrumRG’s COOPED UP webinar, “A Gentle Introduction to Optimal Design for Pharmacometric Models”

I agree that this can be a circular argument, but in practice we usually have a model in place (or at least one or two models hypothesised) from an earlier phase of development and we’re trying to design a new study with this model (or models) in mind.

If there’s a large amount of uncertainty in the current parameter estimates, we can apply optimality criteria more robust than plain old D-optimality (e.g., ED-optimality [1], which puts the “ED” in “PopED”; or HCD/HClnD-optimality [2]).

There are also approaches for cases when there is uncertainty in the model structure itself and we need to be able to discriminate between competing structures [3].

Tim

[1] Pronzato, L., & Walter, É. (1985). Robust experiment design via stochastic approximation. Mathematical Biosciences, 75(1), 103-120.

[2] Foo, L. K., & Duffull, S. (2010). Methods of robust design of nonlinear models with an application to pharmacokinetics. Journal of Biopharmaceutical Statistics, 20(4), 886-902.

[3] Waterhouse, T. H., Redmann, S., Duffull, S. B., & Eccleston, J. A. (2005). Optimal design for model discrimination and parameter estimation for itraconazole population pharmacokinetics in cystic fibrosis patients. Journal of Pharmacokinetics and Pharmacodynamics, 32(3-4), 521-545.