Estimation of statistical moments and steady-state volume of distribution for a drug given by intravenous infusion.

Although it is generally recognized that estimates of the area under the drug concentration vs. time curve (AUC) after a dose is rather insensitive to curve-fitting procedures, little is known about estimates of mean residence time (MRT) or volume of distribution at steady-state (Vss), both of which can be derived from area estimates. This question is of particular concern when a drug is given as a short-term constant rate i.v. infusion since the infusion phase is often ignored and blood sampling restricted to the postinfusion period. Two nonexperimental methods for approximating concentration data during infusion termed the linear approximation method and the imaginary bolus method were found to be useful under certain conditions. Both methods provide reasonable estimates of AUC and the area under the first moment of the drug concentration-time curve (AUMC) for drugs with a wide range of pharmacokinetic characteristics. The imaginary bolus method was found to be the better of the two for estimation of MRT and to be widely applicable for this purpose. Vss proved to be highly sensitive to the approximation methods; although the imaginary bolus method is superior to the linear approximation method for estimating Vss, it does not work particularly well for drugs with pronounced multicompartment characteristics. In general, accurate estimation of Vss of drugs given by short-term i.v. infusion requires that at least one drug concentration be determined during infusion.