Hi! EMA considers C[sub]min[/sub] at the end of the dosing interval as the metric required in steady-state studies (rather than the global minimum within tau)! Of course %PTF (Fluctuation% in PHX/WNL-speak) is also affected. For a first workaround see this post at the BEBA-Forum. I have no idea how to deal with situations where t[sub]last[/sub] # tau.
A workaround if you are able to estimate lambda-z in steady state: I assume a dosage interval (tau) of 24 hours. Run the NCA as usual (don’t forget to give ‘Tau’ in the Setup/Dosing). I hardcoded 24 hours in the example below. For more flexibility you can merge the dosing worksheet beforehand and get the Tau from there and use it as a variable. 1. Cmin estimated at tau: Run a custom transformation on NCA/Output Data/Final Parameters Pivoted; formula: Clastexp(-Lambda_z(24-Tlast)), name: C_TAU Note: In the spirit of AUC_TAU the value is estimated at exactly t=tau; IMHO that#s a better method than doing ‘apples-and-oranges-statistics’ on Clast. If the last value is missing (e.g. a broken vial) and you still get a reasonable estimate of lambda-z, problems in comparing different time points (bias!) are avoided. But: Describe the procedure in the protocol - don’t hide it in an SOP! 2. %PTF (%Fluctuation in PHX/WNL’s terminology): In the transformation above: Next >> 100*(Cmax-C_TAU)/(AUC_TAU/24), name: PTF 3. Don’t know why MRT_TAU is not included in the standard output; therefore: Next >> Arithmetic transformation x/y; map AUMC_TAU to x and AUC_TAU to y, name: MRT_TAU