Pharmacokinetics linear, constant-coefficient

Suppose the compartmentai matrix is a constant matrix (i.e. all are constants). In this situation one can write K instead of K(X,p,t) to indicate that the elements of the matrix no longer depend on (X,p,t). As will be seeiy there are several important features of the K matrix that will be used in recovering pharmacokinetic parameters of interest. In addition, as described in Jacquez and Simon (5) and Coveil et al. (4), the solution to the compartmentai equations (a system of linear, constant-coefficient equations) involves sums of exponentials. 

The notion of residence times can be very important in assessing the pharmacokinetics of a drug. The information about residence times available from a linear, constant-coefficient compartmental model is very rich, and will be reviewed in the following comments. 

What happens in the compartmental model framework Here the most common way to deal with pharmacokinetic data is to fit them first by a sum of exponentials, since, in a linear, constant-coefficient system, the number of exponential phases in the plasma level-vs-time curve equals the number of compartments in the model. 

In conclusion, noncompartmental models and linear, constant-coefficient models have different domains of validity. When the domains are identical, then the pharmacokinetic parameters estimated by either method should, in theory, be equal. If they are not, then differences are due to the methods used to estimate them. 

---