A role for transient infections in the aetiology of venous thrombosis (VT) has been suggested. This study aimed to determine whether individuals who receive antibiotic treatment (as a proxy for infections) have an increased risk of first and recurrent VT and whether infections should be seen as a provoking risk factor for VT. We used the self-controlled case series method to study the risk of first VT during antibiotic prescriptions. The risk of recurrent VT during antibiotic use was estimated by of time-dependent Cox-regression. A total of 2547 patients with a first VT were included and followed for a median of 5·9 years for recurrence (1999-2010), in whom 114 first events occurred during antibiotic use. We found a five-fold increased risk of first VT during antibiotic treatment: [incidence-rate-ratio 5·0; 95% confidence interval (CI), 4·0-6·1]. Antibiotic use was associated with a 2·0-fold (95% CI, 1·1-4·0) increased risk of recurrent VT. Patients with an unprovoked first VT who used antibiotics shortly before this event, had a similar risk of recurrence as patients with a provoked first VT (adjusted hazard ratio 1·1; 95% CI, 0·7-1·7). Individuals who receive antibiotics have an increased risk of first and recurrent VT and infections should be considered a provoking risk factor for VT.
Keywords: antibiotics; epidemiology; infection; recurrence; venous thrombosis.
© 2017 John Wiley & Sons Ltd.