A descriptive tool to characterize nonlinear kinetics, with applications to meperidine and lidocaine.

We describe an explorative data analysis tool which can detect and describe the presence of nonlinearities in multiple dose kinetics studies. The method is nonparametric (i.e. not dependent on modeling assumptions), uses regression to estimate the functions representing the kinetics, and makes the detection of nonlinearity a part of the model selection process. Flexible functions (splines) are used to describe the kinetics corresponding to the lowest given dose, and to describe the (possible) departure of the kinetics corresponding to higher doses from the reference kinetics. The estimated kinetics and departures can be examined to offer possible insight into the nature of the nonlinearity. The methodology is applied to the analysis of meperidine and lidocaine kinetics through the lungs and the heart. We find that the lung kinetics of lidocaine and meperidine are linear. However their myocardial kinetics are complex and nonlinear.