We propose to implement, test and apply new methodologies to investigate and quantify population and individual characteristics of pharmacokinetics in typical patient populations using routinely available patient data. Population characteristics are the average relationships between observable patient features (such as sex, age and renal function) and the parameters of an appropriate pharmacokinetic model (such as volume of distribution or clearance). Individual characteristics are the parameter values that best describe an individual's kinetics. Routine patient data is that clinical and laboratory data obtained solely for the purposes of patient care. The novel methodologies we will implement, test and use are: (1) data analysis methods for estimating population characteristics from routine data. Routine data is costless, and more representative than any other available data. New and important influences on pharmacokinetics can be quantified using it. The drugs investigated here will be chosen from quinidine, theophylline, phenytoin, digoxin, lithium, gentamicin, methotrexate, phenobarbital, and perhaps others. (2) Empirical Bayesian methods that use population characteristics, observable patient features, and past pharmacokinetic (drug concentration) data from the individual as feedback to estimate his individual characteristics. This methodology can directly aid patient care, as well as aid the discovery of new influences on pharmacokinetics. The drugs of interest here are the same as above.
Showing the most recent 10 out of 55 publications
