Methods: Antimicrobial MIC results via Phoenix® and Etest® for Enterobacteriaceae and Pseudomonas aeruginosa isolates were obtained from the clinical microbiology laboratory and processed into unit-specific antibiograms. MIC distributions for aztreonam (ATM), cefepime (CFP), meropenem (MEM), and piperacillin/tazobactam (PTZ) were generated from the data. Pharmacokinetic parameters for each agent were derived from healthy subject literature. PD target attainment was modelled utilizing Monte Carlo Simulations for 4 different critical care units (Cardiovascular [CVICU], Medical [MICU], Neurosurigical [NSICU], and Surgical/Trauma [SICU]). Desired %fT>MIC was 40% for MEM, 50% for PTZ, and 60% for ATM and CFP.
Results: 508 isolates (103 P. aeruginosa and 405 Enterobacteriaceae) across 4 units were included. SICU had the most isolates (178) and Enterobacteriaceae (150) isolates, while MICU had the most P. aeruginosa (38) isolates. In general, ATM had the lowest probability of achieving the PD target when compared to the other agents. All CFP and MEM regimens evaluated against Enterobacteriaceae were predicted to achieve the PD target >90% of the time. In contrast, CFP and MEM at 2g q8h prolonged infusion over 3 hours against P. aeruginosa were predicted to achieve the PD target >90% of the time in the NSICU, but not in other units. Decreases in overall PD target attainment for all agents were driven by P. aeruginosa MICs, as probability of target attainment was significantly higher for Enterobacteriaceae across all regimens and units. Prolonged infusion regimens outperformed the equivalent bolus regimen for each corresponding drug regimen.
Conclusion: Differing MIC distributions among critical care units may lead to different optimal antimicrobial dosing regimens to achieve the desired PD target. Empiric beta-lactam monotherapy may be inadequate for P. aeruginosa infections in critical care units.
W. C. Rutter,
D. S. Burgess, None
See more of: Poster Abstract Session