Objectives
Osteoarticular infection (OAI) is a feared complication of Staphylococcus aureus bacteraemia (SAB) and is associated with poor outcomes. We aimed to explore the risk of OAI and death following SAB in patients with and without rheumatoid arthritis (RA) and to identify risk factors for OAI in patients with RA.

Methods
Danish nationwide cohort study of all patients with microbiologically verified first-time SAB between 2006–18. We identified RA, SAB, comorbidities, and RA-related characteristics (e.g. orthopaedic implants and antirheumatic treatment) in national registries including the rheumatology registry DANBIO. We estimated the cumulative incidence of OAI and death and adjusted hazard ratios (HRs, multivariate Cox regression).

Results
We identified 18 274 patients with SAB (n = 367 with RA). The 90-day cumulative incidence of OAI was 23.1% (95% CI 18.8; 27.6) for patients with RA and 12.5% (12.1; 13.0) for patients without RA (non-RA) [HR 1.93 (1.54; 2.41)]. For RA patients with orthopaedic implants cumulative incidence was 29.4% (22.9; 36.2) [HR 1.75 (1.08; 2.85)], and for current users of tumor necrosis factor inhibitors (TNFi) it was 41.9% (27.0; 56.1) [HR 2.27 (1.29; 3.98) compared with non-users]. All-cause 90-day mortality following SAB was similar in RA [35.4% (30.6; 40.3)] and non-RA [33.9% (33.2; 34.5), HR 1.04 (0.87; 1.24)].

Conclusion
Following SAB, almost one in four patients with RA contracted OAI corresponding to a doubled risk compared with non-RA. In RA, orthopaedic implants and current TNFi use were associated with approximately doubled OAI risk. One in three died within 90 days in both RA and non-RA. These findings encourage vigilance in RA patients with SAB to avoid treatment delay of OAI.

rheumatoid arthritis, biologic disease-modifying antirheumatic drugs, glucocorticoids, infection, Staphylococcus aureus, epidemiology, time-to-event study, mortality, prosthetic joint infection, septic arthritis