Doing the Same with Less: A Randomized, Multinational, Open-Label, Adjudicator-Blinded Trial of an Algorithm vs. Standard of Care to Determine Treatment Duration for Staphylococcal Bacteremia

Abstract Background The appropriate duration of antibiotics for staphylococcal bloodstream infection (BSI) is unknown. An algorithm to identify patients with staphylococcal BSI who can be safely treated with shorter courses of therapy would improve care and reduce total antibiotic use. Methods Adult patients with staphylococcal BSI were randomized to treatment based on algorithm-based therapy (ABT) or to standard of care (SOC). Co-primary outcomes were clinical success, as determined by a blinded Adjudication Committee, and serious adverse event (SAE) rates. The prespecified secondary outcome measure was antibiotic days by treatment group, among patients without complicated BSI. Prespecified durations of therapy in ABT were: S. aureus BSI (SAB): uncomplicated: 14 days; complicated: 4–6 weeks. Coagulase-negative staphylococci BSI (CoNSB): simple (1 positive blood culture) (0–3 days), uncomplicated (>1 positive blood culture) (5 days), complicated (7–28 days). Outcomes were compared using intention-to-treat principles. The target sample size was 500 patients, to ensure 90% power for establishing noninferiority within a margin of 15%. Results Between April 2011 and March 2017, 509 adults with suspected staphylococcal BSI at 16 sites in the US and Spain were randomized to ABT (N = 255) or SOC (N = 254). There were 116 patients with SAB (23%) and 385 (76%) with CoNSB (Figure 1). Overall success rate in the ABT group was 82.0% vs. 81.5% in the SOC group, difference 0.5%, 95% CI −5.2% to 6.1%. SAEs were reported in 32.9% of ABT vs. 28.3% of SOC patients (OR 1.2, 95% CI 0.9 to 1.8). Among evaluable patients without complicated BSI, mean duration of therapy was 4.4 days in the ABT group vs. 6.4 days in the SOC group (difference −2.0 days, 95% CI −3.3 to −-0.7, P = 0.003). Among patients with uncomplicated SAB, treatment durations were similar (15.3 days in ABT vs. 16.3 days in SOC, difference −1 day, 95% CI −3.89 to 1.91, P = 0.497), whereas for uncomplicated CoNSB, duration was shorter in the ABT group (5.3 days in ABT vs. 8.4 days in SOC, difference −3 days, 95% CI −4.87 to −1.34, P < 0.001). Conclusion The use of a treatment algorithm for staphylococcal BSI was associated with significant reductions in duration of antibiotic therapy in patients without complicated BSI, without significant differences in overall success or SAEs. Disclosures V. Fowler Jr., NIH: Investigator, Contract HHSN272200900023C