Page 13 - 70_03

VOL. 70 (3), 663-679, 2004 PHARMACOKINETICS AND INDIVIDUALIZED DRUG THERAPY

looks fine when two blood samples were taken and the concentratio-
ns were measured; the trough level should exceed 500 ng/mL for this
infection.

Suddenly, however, the last blood sample was very low and you
start to wonder what has happen. The first question is of course to
ask the patient if he really has been taken the drug. During the
interviews with the parents it was evident that the mother usually
had been supervising the medication, but during the last weeks the
boy had been trusted to manage his own medication. He assured
that he had been taken the drug, but the measured low concentra-
tion got him to confess that he has not been taken the drug for two
weeks and an Bayesian estimate fitted very well to this kind of non
compliance as shown in Fig. 7.

FIGURE 7. A Bayesian estimate of drug dosing taking into consideration that the
last 8 doses in Fig. 6 have not been taken by the patient.

Phenytoin, as you know, is widely used as an antiepilectic drug,
but it is difficult to find the correct dose, because it is very easy to
saturate the enzymes for its elimination. Just a small increase in
dose can lead to serious side effects and a too small dose does not
protect against seizures. In Fig. 8 is shown a patient where the
measured plasma concentration looks fine, but in reality, if the pa-
tient continues with the prescribed dose 250 mg two times daily it
will lead to serious side effects.

673