3 Most treatment recommendations are therefore based on observational studies and clinical experience. Fewer than 1600 participants have been enrolled in randomised trials of antibiotic therapy for the treatment of this infection over the past 50 years. 2 Despite the frequency and severity of S aureus bacteraemia, the optimal antibiotic treatment is uncertain. 1 When S aureus enters the bloodstream it can disseminate to cause metastatic, deep-seated infection of almost any organ, with an associated mortality of approximately 20%. Staphylococcus aureus bloodstream infection, also known as bacteraemia, is one of the most common and serious community-acquired and hospital-acquired bacterial infections worldwide. From randomisation to 12 weeks, no evidence of differences in serious (p=0♱7) or grade 3–4 (p=0♳6) adverse events were observed however, 63 (17%) participants in the rifampicin group versus 39 (10%) in the placebo group had antibiotic or trial drug-modifying adverse events (p=0♰04), and 24 (6%) versus six (2%) had drug interactions (p=0♰005). By week 12, 62 (17%) of participants who received rifampicin versus 71 (18%) who received placebo experienced treatment failure or disease recurrence, or died (absolute risk difference −1♴%, 95% CI −7♰ to 4♳ hazard ratio 0♹6, 0♶8–1♳5, p=0♸1). Standard antibiotics were given for 29 (IQR 18–45) days 619 (82%) participants received flucloxacillin. 301 (40%) participants had an initial deep infection focus. 47 (6%) had meticillin-resistant infections. 485 (64%) participants had community-acquired S aureus infections, and 132 (17%) had nosocomial S aureus infections.

Between Dec 10, 2012, and Oct 25, 2016, 758 eligible participants were randomly assigned: 370 to rifampicin and 388 to placebo.

Set in Stone by Angela Brown