• Primary efficacy: Improved or cure with or without antibiotic modifications, defined as: survival at 7 days post randomization with resolution of fever and resolution of hypotension [ Time Frame: 7 days ] [ Designated as safety issue: Yes ]
  
• Primary safety: 30-day all cause mortality [ Time Frame: 30 days ] [ Designated as safety issue: Yes ]
  

• Improved or cure without antibiotic modifications [ Time Frame: 7 days ] [ Designated as safety issue: No ]
  
• Modification of the anti-staphylcoccal treatment within 1 week of treatment onset for perceived failure of therapy [ Time Frame: 7 days ] [ Designated as safety issue: No ]
  
• Survival at 7 days post randomization without the need for modification of the anti-staphylcoccal antibiotic [ Time Frame: 7 days ] [ Designated as safety issue: Yes ]
  
• Bacteriological failure, defined as persistent isolation of Staphylococcus aureus with the same phenotype 7 days after or more after treatment onset [ Time Frame: 7 days ] [ Designated as safety issue: No ]
  
• Need for surgical intervention or other invasive procedures [ Time Frame: 30 days ] [ Designated as safety issue: No ]
  
• Need for central catheter removal [ Time Frame: 30 days ] [ Designated as safety issue: No ]
  
• Persistent bacteremia [ Time Frame: 30 days ] [ Designated as safety issue: No ]
  
• All-cause mortality in ICU and in-hospital [ Time Frame: 30 days ] [ Designated as safety issue: Yes ]
  
• Adverse events [ Time Frame: 30 days ] [ Designated as safety issue: Yes ]
  
• Durations of fever, assigned antibiotic treatment, mechanical ventilation, ICU and hospital stay [ Time Frame: 30 days ] [ Designated as safety issue: Yes ]
  
• Resistance development [ Time Frame: 30 days ] [ Designated as safety issue: No ]
  

Staphylococcus aureus (SA) is a major pathogen causing community-acquired and health-care associated infections. In hospitals, SA infections are associated with a significant burden; in-hospital mortality during the last 15 years following SA bacteremia in Beilinson hospital was 38% and did not decrease in recent years. Resistance to beta-lactams is widely prevalent in hospitals (57% of all SA isolates causing bacteremia at our center). The drug of choice currently recommended for these infections is a glycopeptide (vancomycin or teicoplanin).

Cotrimoxazole (trimethoprim-sulfamethoxazole) is a relatively 'old' drug commonly used for urinary tract infections. In-vitro, it is active against SA, including methicillin-resistance Staphylococcus aureus (MRSA) strains and its activity against SA is bacteridicidal. Trimethoprim alone is bactericidal against SA, while sulphamethoxazole alone is relatively inactive and their combination is synergistic both in-vitro and in-vivo. The prevalence of cotrimoxazole-susceptible SA varies locally. At our center, 97% of SA strains causing bacteremia in 2004 were susceptible to cotrimoxazole. Community-acquired MRSA, prevalent in the United States as a cause for severe skin and soft tissue infections, has not been described in Israel.

Several reasons exist to search for antibiotics other than vancomycin for MRSA infections. Vancomycin is less effective that beta-lactam drugs for SA infections susceptible to both agents. It is the last resort antibiotic for MRSA infections out of the currently recommended bactericidal antibiotics for invasive infections. Use of vancomycin has led to the development of SA strains with partial or complete resistance to vancomycin (VISA and VRSA, respectively). Vancomycin use is associated with the appearance of vancomycin-resistant enterococcus (VRE) species. Nosocomial infections with VISA and VRE are difficult to treat and may spread rapidly in the hospital. 10 Finally, vancomycin cannot be administered orally.

Limited evidence supports the efficacy of cotrimoxazole for MRSA infections, with paucity of data for high-burden invasive infections. Cotrimoxazole is probably inferior to vancomycin for methicillin-susceptible SA. ; thus we may infer indirectly its inferiority to methicillin and drugs alike for MRSA infections. Cotrimoxazole may be less effective than glycopeptides and oxacillin for left-sided endocarditis. No evidence exists to support the use of cotrimoxazole empirically for the treatment of suspected SA infections in the hospital.

We plan an open label single-center pragmatic randomized controlled trial to compare cotrimoxazole to vancomycin. We will include patients with documented or highly suspected MRSA infections, according to pre-defined risk factors. We chose to target this patient population to assess the efficacy of cotrimoxazole both empirically and for documented infections.