OBJECTIVES: Recent evidence suggests that there is an association between infection and venous thromboembolism (VTE). Here, we examined the risk of VTE after Staphylococcus aureus bacteraemia (SAB) compared to the risk in control subjects.

DESIGN AND SETTING: Register-based nationwide observational cohort study of hospitalized patients and matched control subjects from the general population in Denmark between 1995 and 2008.

RESULTS: Amongst 15 669 SAB cases and 156 690 controls, 182 and 511, respectively, experienced VTE within 1 year. The overall incidence rate (IR) of VTE amongst cases was highest within the first 30 days [IR of deep vein thrombosis (DVT), 39.3 (95% confidence interval (CI) 28.9-53.4)/1000 person-years (PYs); IR of pulmonary embolism (PE), 16.3 (95% CI 10.1-26.2)/1000 PYs]. IRs of DVT were particularly increased amongst cases with a previous diagnosis of VTE, community-acquired infection, a history of injection drug use and in younger age groups. The overall hazard ratio of VTE for cases compared to controls declined from 15.6 (95% CI 10.3-23.5) in the first 30 days after SAB to 4.5 (95% CI 3.2-6.2) from 181 to 365 days after infection. The increased risk of VTE amongst SAB patients persisted after excluding cases with identified VTE risk factors.

CONCLUSIONS: There was a particularly high risk of VTE during the first month following an episode of SAB. The risk declined over time, but remained at a threefold increased level compared to control subjects, suggesting that there are shared risk factors for SAB and VTE. Patients with SAB and well-documented risk factors for VTE may benefit from thromboprophylaxis.