Comparison of probability of target attainment calculated by Monte Carlo simulation with meropenem clinical and microbiological response for the treatment of complicated skin and skin structure infections

Abstract

Monte Carlo simulation is often used to predict the cumulative fraction of response (CFR) for antibiotics, but the relevance of these predictions to outcomes in humans has not been well studied. We compared the CFR for meropenem 500 mg every 8 h against pathogens causing complicated skin and skin structure infections from a randomised, multicentre clinical trial with clinical response (CR) and microbiological response (MR). A population pharmacokinetic model was utilised to estimate pharmacokinetic parameters for 96 clinically evaluable patients with pathogen and minimum inhibitory concentration (MIC) data available. A 1000-subject Monte Carlo simulation was performed to estimate bacteriostatic (20% of time serum concentration above the MIC ( T > MIC)) and bactericidal (40% T > MIC) exposures for comparison. Only the bactericidal CFR versus the CR was not statistically different (92% CR versus 91.9% CFR; 95% confidence interval of the difference, −7.7% to 4.2%), whilst bacteriostatic CFRs overestimated actual CR and MR. This study demonstrates that the use of Monte Carlo simulation to predict the CR of meropenem in complicated skin and skin structures is accurate.