Monte Carlo simulations: Maximizing antibiotic pharmacokinetic data to optimize clinical practice for critically ill patients

Infections in critically ill patients continue to result in unacceptably high morbidity and mortality. Although few data exist for correlating antibiotic exposure with outcome, antibiotic dosing is likely to be highly important for maximizing resolution of infection in many patients. The practical and financial difficulties of performing pharmacokinetic (PK) studies in critically ill patients mean that analyses to maximize data such as Monte Carlo simulation (MCS) are highly valuable. MCS uses computer software to perform virtual clinical trials. The building blocks for MCS are: firstly, a robust population PK model from the patient population of interest; secondly, descriptors of the effect of covariates that influence the PK parameters; thirdly, description of the susceptibility of bacteria to the antibiotic and finally a PK/pharmacodynamic (PD) target associated with antibiotic efficacy. Probability of target attainment (PTA) outputs can then be generated that describe the proportion of patients that will achieve a pre-specified PD target for an MIC distribution. Such analyses can then inform dosing requirements, which can be used to have a high likelihood of achieving PK/PD targets for organisms with different MICs. In this issue of JAC, Zelenitsky et al. provide a very useful example of MCS for interpreting the optimal methods for dosing meropenem, piperacillin/tazobactam, cefepime and ceftobiprole in critically ill patients