Objectives: To document the effects of appropriate and inappropriate empirical antibiotic therapy on mortality in a cohort of patients with bacteraemia due to methicillin-resistant Staphylococcus aureus (MRSA) and to summarize effects with previous studies.
Methods: In the retrospective cohort study, episodes of clinically significant MRSA bacteraemia during a 15 year period were included. Polymicrobial episodes were excluded unless MRSA was isolated in more than one bottle and co-pathogens were given appropriate empirical antibiotic treatment. Appropriate empirical treatment was defined as matching in vitro susceptibility and started within 48 h of blood-culture taking, except for single aminoglycosides or rifampicin. We assessed univariate and multivariate associations between appropriate empirical therapy and 30 day all-cause mortality. Multivariable analysis was conducted using backward stepwise logistic regression. We reviewed all studies assessing the effects of appropriate empirical antibiotic treatment on mortality for MRSA infections and compiled adjusted odds ratios (ORs) using a random effects meta-analysis.
Results: Five hundred and ten episodes of MRSA bacteraemia were included. There were no cases of community-acquired infection. The 30 day mortality was 43.9% (224/510) and was stable throughout the study period. Mortality was significantly higher among patients receiving inappropriate (168/342, 49.1%) compared with those receiving appropriate (56/168, 33.3%) empirical antibiotic treatment, P = 0.001. In the adjusted analysis the OR was 2.15 [95% confidence interval (CI) 1.34-3.46]. Pooling of six studies using adequate methodology for the adjusted analysis resulted in an OR of 1.98 (95% CI 1.62-2.44).
Conclusions: Appropriate empirical antibiotic treatment has a significant survival benefit in MRSA bacteraemia. Treatment guidelines should consider this benefit.