Could someone explain the variance-covariance matrix set-up of the random effects tab? I am new to population PK modeling and would like to use the Phoenix NLME interface. I understand that the default setting is to select the diagonal covariance model that assumes all random effects are independent. My data showed that CL and V were highly correlated, but Ka was not correlated to any other parameter. How would I proceed from this point? I see there is an option to “split” or “break” the parameters, but it is not clear what this means. Any help would be appreciated. Thanks!
To do this, you will need to click on the “Break” button between nKa and nV, nCl. Then, deselect the “Diag” check box for the nV, nCl group.