The main limitation of the present analysis is that treatment selected by clinicians was confounded by underlying patient characteristics. Probably no method can fully adjust for the differences between patients given cloxacillin or cefazolin empirically and those treated with broader spectrum beta-lactams. We attempted to adjust for these differences using multivariable regression analysis. While the model constructed was highly predictive, we cannot be sure that all differences between patients treated with narrow vs. broad-spectrum 3-Methyladenine solubility dmso
beta-lactams were adjusted for. Once MSSA is identified, a comparison between cloxacillin and cefazolin is more valid, because there is no information regarding whether one or the other treatment is superior. However, this analysis was limited by the paucity of patients treated with cefazolin, resulting in wide confidence intervals. We limited the analysis of definitive antibiotic treatment to the first week because treatment modifications after this period were frequently dictated by patients�� responses to the initial treatment regimen. For similar reasons we did not consider the overall treatment duration and antibiotic combinations with aminoglycosides and rifampin. Combination treatment PDGFR inhibitor
is rare in our centre because there is no evidence for its effect on survival [12,13] and combination treatment did not significantly affect survival in this cohort (data not shown). A distinctive feature of our cohort was that truly community-acquired bacteraemias were rare (12.2%). We used contemporaneous definitions for healthcare-associated bacteraemia and our Ceftiofur
electronic healthcare system allows for a very high precision in identifying patients�� exposure to the healthcare system before hospitalization. In summary, we present the first analysis attempting to compare different beta-lactams in the treatment of MSSA bacteraemia. Empirical treatment with oxacillin or cefazolin was associated with lower 30-day mortality than other beta-lactams. Definitive treatment with cefazolin or cloxacillin resulted in similar 90-day mortality rates. These results are severely hampered by the association between the treatment regimen selected by clinicians and underlying patient characteristics. This study highlights the difficulties in comparing treatments in a non-randomized design. Many questions regarding the treatment of invasive S.?aureus infections in general and MSSA bacteraemia in particular exist, including the type of antibiotic, combination therapy and duration of treatment. These questions should be answered in randomized controlled trials. All authors, nothing to declare. Funding: none. ""A new editorial team for CMI took over in 2008. The official start of this new team was January 2009; however, for reasons beyond our control, we had to commence before this date.