On the Single-Point, Single-Dose Problem

Abstract

The problem of predicting dosage requirements for individual patients has received considerable attention in recent years. The use of average values of pharmacokinetic parameters for predicting dosage rates or steady-state drug concentrations has obvious limitations due to inter-patient variability. In order to incorporate individual variability into such predictions, the “single-point, single-dose” method was introduced by Slattery et al. [1]. In this method, one measurement of serum drug concentration is made at a sampling time, t _s , after a test dose is given. This single measurement is then used in a linear, one compartment model to predict relevant pharmacokinetic variables. Slattery et al. [1], as well as other investigators [2–4], have indicated an optimal sampling time of ${t_s} = k_0^{ - 1}$ where k ₀ is the average population value of the elimination rate constant. These analyses generally utilize a localized linearization argument, which is strictly valid only in the limit as the population variance of the rate constant approaches zero.