Pharmacokinetics/Pharmacodynamics

Determination of the calculated peak concentration:

Cmax = C1 (eke(t))

where Cmax is the calculated peak concentration and t is the time between the C1 and the end of the

intravenous infusion in hours.

Determination of the calculated trough concentration:

Cmin = C2 (e-ke(t))

where Cmin is the calculated trough concentration and t is the time between C2 and the beginning of the

next dose.

Determination of the drug Vd:

Vd = ×

dose

tinf × ke

(1 – e-ke(tinf))

Cmax – Cmin × e-ke(tinf)

where tinf is the duration of the drug infusion.

Determination of the new dosing interval:

τ = + tinf

ln

ke

( )

Cmax,desired

Cmin,desired

where τ (tau) is the new dosing interval, Cmax,desired is the desired peak concentration for the new dosing

regimen, and Cmin,desired is the desired trough concentration for the new dosing regimen.

Determination of the new dose:

dose = Cmax,desired × ke × Vd × tinf ×

(l – e-ke(τ))

(l – e-ke(tinf))

One drawback of this method is that it assumes that the PK variables obtained (e.g., vancomycin trough)

correlate with PD variables (AUC/MIC). Although this may be true in many cases, some advocate

incorporating PD into individualized drug dosing. Alternative methods use PK variables from previous

patients (population PK) to estimate the PK in an individual patient. These methods usually require

complicated mathematical calculations, many of which are not practical for clinical use. Many software

programs are now available for clinician use in patient care that simplify the process of carrying out

complex calculations. Bayesian modeling has also been proposed to address the issues posed when using

population PK in patient groups that may not be well represented in the population. Using software with

population PK variables from non-critically ill patients to interpret or predict PK in critically ill patients

could result in errors in designing an appropriate drug-dosing regimen.

A position paper provides a more detailed review of the PK, PD, and TDM of commonly used antimicrobials

in critically ill patients and notes recommendations that certain antimicrobials should be monitored

on this list because they have routinely been monitored in critically ill patients. The recommendations

also include the β-lactam class of antibiotics, linezolid, and the antifungal voriconazole. Finally, the

document provides PK and PD target variables for these drugs. The greatest barrier continues to be

access to local laboratory analysis and results within a clinically relevant time frame. The reader is

referred to the position paper for a more in-depth review.