A comparison of parametric with semiparametric analysis of the concentration versus effect relationship of metocurine in dogs and pigs.

We investigated the ability of two pharmacokinetic modeling techniques to estimate the equilibration delay (i.e., hysteresis) between plasma drug concentration and observed drug effect. The data were from 20 animals (15 dogs, 5 pigs) receiving an infusion of metocurine, a neuromuscular blocking drug. An effect compartment model was used to model the hysteresis and characterize the relationship between drug concentration and effect. The effect compartment model requires identification of ke0, the rate constant of drug elimination from the effect compartment. Two methods were used to estimate ke0. The first technique was to fit the plasma metocurine concentration-time curve to a two-compartment pharmacokinetic model and then to use this pharmacokinetic model, along with the neuromuscular blockade vs. time curve to estimate ke0 and the parameters of a pharmacodynamic model (the Hill equation). The second technique was to directly estimate ke0 by a recently described semiparametric technique that does not require either a pharmacokinetic or pharmacodynamic model, although it does assume that drug flux to and from the effect compartment is a first-order process. This semiparametric technique only estimates a single parameter, ke0. The results from the new semiparametric analysis technique were similar to the results from the parametric analysis. In the few animals where the results differed, the semiparametric analysis produced a better description of the data.