Comparative and Veterinary Pharmacology

conducted in a very homogeneous animal population under well-controlled labora-
tory conditions versus the wider range of PK characteristics that are likely to be
observed under actual conditions of use. For simplicity, the MIC 90 in this case was
assumed to be 1mgmL–1, and the target AUC/MIC was assumed to be 100 h.
Under these conditions, we calculated the AUC/MIC value associated with a 90%
TAR. This is illustrated in Fig.1a, which demonstrates that, as variability increases,
there is a marked change in the AUC/MIC value achieved by 90% of the simulated
population (i.e. the 90% TAR), even though the average AUC/MIC value in the
population remained constant.
The final step shows the impact of population variability on dose selection. In
other words, assuming that the goal is to have a dose that results in a 90% TAR for a
target AUC/MIC of 100 h, how would the estimated dose change as population
variability increases? For simplicity, all estimated dosages are expressed relative to

Relationship between AUC/MIC value associated with 90%

TAR versus %CV

a

b

0 20 40 60 80 100

Percent CV

AUC/MIC for 90%

TAR

0

20

40

60

80

100

120

Change in dose to acheive the same AUC/MIC as estimated

from a population with 15% CV

Percent CV

Ratio of dose needed relative to achieve a

90% TARthe

(expressed relative to

the 15% CV group)

0

0.5

1

1.5

2

2.5

3

0 20 40 60 80 100

Fig. 1(a) The influence of variability (expressed as %CV) on the AUC/MIC value associated with
a 90% TAR. (b) The influence of variability on the dose needed to achieve a 90% TAR for AUC/
MIC = 100 h. TheX-axis represents the %CV of the simulated population. TheY-axis reflects the
ratio of the dose for any %CV relative to that needed when the %CV = 15

Antimicrobial Drug Resistance 243