Staphylococcus aureus bacteremia: A treatment algorithm
An algorithm to define the duration of therapy for patients with Staphylococcus aureus bacteremia was prospectively tested. Recommendations for all patients with SAB were as follows: all removable foci of infection should be removed, surveillance blood cultures should be obtained on day 2-4 of antibiotic therapy, and TEE should be performed on day 5-7 of antibiotic therapy. Each patient was classified according to the following clinical schema: 1) Simple bacteremia. All of the following criteria were met: TEE negative for either vegetations or predisposing valvular abnormalities; negative surveillance blood cultures; focus of infection removed; rapid clinical resolution; no prosthetic devices present; no clinical evidence of metastatic infection. 2) Uncomplicated bacteremia. Any of the following criteria were met; TEE revealed predisposing valvular lesions but no evidence of IE; growth of S. aureus in a surveillance blood culture; non-removable superficial skin focus of SAB; clinical signs of infection after 72 hours of antibiotic therapy. 3) Endocarditis-according to the Duke Criteria. 4) Extracardiac disease- deep tissue infection in a patient with a TEE negative for vegetations. Therapeutic recommendations were made according to these categories. A 7 day course of intravenous antibiotics was recommended for simple bacteremia; a 14 day course for uncomplicated bacteremía; a 4 to 6 week course for IE; and a 4 to 8 week course plus appropriate surgical debridement for patients with extracardiac disease. Our recommendations were followed in 109 patients. None of the 43 patients (39%) who received short course therapy relapsed or died due to SAB. Patients receiving short course therapy according to algorithm recommendations were significantly less likely to relapse than patients with catheter associated SAB in whom our algorithm recommendations were not followed (0/43 vs 10/57 patients, p=0.03). 66 patients were treated with 4-8 week courses of intravenous antibiotics for IE (32 patients) or extracardiac infection (34 patients). 14 of these 66 patients (21%) failed therapy: 7 (11%) relapsed and 7 died due to SAB. Our algorithm identifies patients with SAB suitable for short course therapy and detects many patients with clinically inapparent IE.