Objectives: In order to explore the pharmacodynamic need for continuous versus intermittent (three times a day) administration of ceftazidime in critically ill patients, a pharmacokinetic computerized device was used to simulate concentrations of ceftazidime in human serum after 6 g/day.
Methods: Efficacy was measured as the capability of simulated concentrations over time to reduce initial inoculum against four strains of Pseudomonas aeruginosa. MICs of the strains matched NCCLS breakpoints: one susceptible strain (MIC = 8 mg/L), two intermediate strains (MIC = 16 mg/L) and one resistant strain (MIC = 32 mg/L). C(max) was 119.97+/-2.53 mg/L for intermittent bolus and C(ss) (steady-state concentration) was 40.38+/-0.16 mg/L for continuous infusion. AUC(0-24) was similar for both regimens (approximately 950 mg x h/L). Inhibitory quotients were three times higher for the intermittent administration whereas t > MIC was higher for continuous infusion (100%) versus intermittent administration (99.8%, 69% and 47.6% for the susceptible, intermediate and resistant strains, respectively).
Results: Against the susceptible and intermediate strains, no differences were found between both regimens with > or = 3 log10 reduction from 8 to 24 h. Against the resistant strain, only the continuous infusion achieved this bactericidal activity in the same time period, minimizing the differences between resistant and susceptible strains. Significantly higher initial inoculum reduction at 32 h was obtained for the continuous versus the intermittent administration (83.35% versus 38.40%, respectively).
Conclusions: These results stress the importance of optimizing t >MIC, even at peri-MIC concentrations, of ceftazidime against resistant strains. Local prevalence of resistance justifies, on a pharmacodynamic basis, electing for continuous infusion versus intermittent administration.