Abstract
BACKGROUND: It is important to optimize dosing schemes of antibiotics to maximize the probability of therapeutic success. The recommended pharmacokinetic/pharmacodynamic (PK/PD) index for piperacillin/tazobactam therapy in clinical studies ranges widely (50%-100% fT(>1-4×MIC)). Dosing schemes failing to achieve PK/PD targets may lead to negative treatment outcomes. OBJECTIVES: The first aim of this study was to define the optimal PK/PD index of piperacillin/tazobactam with a hollow-fibre infection model (HFIM). The second aim was to predict whether these PK/PD targets are currently achieved in critically ill patients through PK/PD model simulation. PATIENTS AND METHODS: A dose-fractionation study comprising 21 HFIM experiments was performed against a range of Gram-negative bacterial pathogens, doses and infusion times. Clinical data and dose histories from a case series of nine patients with a known bacterial infection treated with piperacillin/tazobactam in the ICU were collected. The PK/PD index and predicted plasma concentrations and therefore target attainment of the patients were simulated using R version 4.2.1. RESULTS: fT (>MIC) was found to be the best-fitting PK/PD index for piperacillin/tazobactam. Bactericidal activity with 2 log(10) cfu reduction was associated with 77% fT(>MIC). Piperacillin/tazobactam therapy was defined as clinically 'ineffective' in ∼78% (7/9) patients. Around seventy-one percent (5/7) of these patients had a probability of >10% that 2 log(10) cfu reduction was not attained. CONCLUSIONS: Our dose-fractionation study indicates an optimal PK/PD target in piperacillin/tazobactam therapies should be 77% fT(>MIC) for 2 log(10) kill. Doses to achieve this target should be considered when treating patients in ICU.