Clinical implications of β-lactam-aminoglycoside synergism: systematic review of randomised trials

β-Lactam-aminoglycoside combinations are commonly used despite lack of evidence of a clinical benefit. In this study, all randomised controlled trials (RCTs) assessing directly the clinical implications of synergism by comparing a β-lactam with the same β-lactam in combination with an aminoglycoside as empirical or definitive therapy for any type of infection and clinical scenario were compiled. A systematic search was undertaken to identify all trials regardless of language, date or publication status. The primary outcomes assessed were all-cause mortality and clinical failure regardless of antibiotic modifications. Risk of bias was evaluated and its effect was assessed through sensitivity analyses. Two reviewers applied inclusion criteria and extracted the data independently. A fixed-effect meta-analysis was performed. Fifty-two RCTs were identified assessing patients with febrile neutropenia, pneumonia, abdominal infections, bacteraemia, endocarditis or cystic fibrosis. Only five trials were double-blinded. All-cause mortality was similar with monotherapy versus combination therapy [risk ratio (RR)=0.96, 95% confidence interval (CI) 0.78-1.18, 28 trials, 3756 episodes]. Clinical failure regardless of antibiotic modifications was not significantly different (RR=0.88, 95% CI 0.74-1.05, 27 trials, 2500 episodes). Treatment failure including antibiotic addition/modification occurred more frequently with monotherapy (RR=1.20, 95% CI 1.12-1.28, 48 trials, 6643 episodes). There were no significant differences with regard to bacterial or fungal superinfections or development of antibiotic-resistant strains. Combination therapy resulted in a significantly higher incidence of adverse events, mainly nephrotoxicity. Overall, no clinical benefit was found for the use of a β-lactam with an aminoglycoside compared with a β-lactam alone. Treatment with β-lactams as monotherapy entailed more antibiotic regimen modifications in open trials.