How to fit pharmacokinetic data

Numerous studies analyze blood samples to determine the concentration of a drug after administration to volunteers or patients. Depending on one hand on the nature of the active pharmaceutical ingredient and the drug formulation and on the other hand on the way it has been administered, data collected contain valuable information regading the processes in which the drug is involved in the human body.

Extracting this information is of utmost importance to researchers. It requires testing various models against the data, determining pertinent model parameters through fitting procedures and deciding what is the most meaningful and reliable model. Pharmacokinetic models have been developed for several decades starting from mathematical expressions that are based on the underlying processes.

Even before the implementation of computers, fitting algorithms had been established to allow calculation of best model parameters. For the most part, the trend was to convert the data in a form that would generate a simple shape, e.g., a straight line. This apporach has several advantages, such as visual comparison and ease of calculation. Data transformations are associated with changes in the statistical weighting of individual data points.

Fitting algorithms and computer advances have made direct fitting to complicated (non-linear) models possible and fast, but the process is not automatic. Starting values for model parameters can be crucial for the determination of meaningful final parameters. It rests with the operator to decide if the model tried is the most appropriate. There are statistical criteria for unbiased decisions, but experience is a very important asset for the successful operator. Herein, we utilize PK models based on the concept of finite time of absorption, the so called Physiologically Based Finite Time Pharmacokinetic Models (PBFTPK).

You can test your data in the form below.

Your name:
E-mail address:
Paste your data here, preferably in two columns (or three columns if you have error estimates) separated by tabs or single spaces.

Comments, data units, suggestions:


Biopharmaceutics, Pharmacokinetics, Pharmacodynamics Group