Search

Background: The difference in type of antibiotics and susceptibility of Bacteroides fragilis to antibiotics may influence warfarin anticoagulation. However, these influences have not been clarified in clinical settings., Objectives: This study aimed to investigate association the between the prothrombin time-international normalized ratio (PT-INR) and concomitant use of antibiotics in a real-world population of warfarin users., Methods: This was a single-center cohort study using data from health records and included patients who received β-lactams (BLs)/fluoroquinolones (FQs) during ongoing warfarin treatment (2011-2015) at Hamamatsu University Hospital in Japan. Antibiotics were categorized into those to which B fragilis is susceptible (BL sus , FQ sus ) and those to which it is not (BL non , FQ non ) and into those given orally (BL po , FQ po ) or intravenously (BL iv , FQ iv ). Outcomes were excessive PT-INR and changes in PT-INR, defined as the ratio (INR ratio) and difference (ΔINR) of maximum PT-INR and baseline PT-INR. Excessive PT-INR was graded as INR ratio of >1.5 or >2.5., Results: A total of 1185 warfarin users were included. The proportion of INR ratio >2.5 in FQ iv was higher than in BL iv (95% CI: 1.59-46.5). The proportions with an INR ratio of >1.5 in BL sus and FQ sus were higher than in BL non (1.72-14.1) and FQ non (1.05-9.36), respectively. ΔINR values in FQ po , FQ iv , and FQ sus were higher than those in BL po , BL iv , and FQ non , respectively., Conclusions and Relevance: Concomitant use of FQs, or of antibiotics to which B fragilis is susceptible is associated with higher risk of excessive anticoagulation. These findings would contribute to safe and proper antibiotic treatment in warfarin users.

Fluoroquinolones and macrolides may, due to a potential drug‐drug interaction, increase the concentration of any concomitantly administered direct oral anticoagulant (DOAC) and thereby increase the risk of severe bleeding. However, clinical evidence for such an effect is scarce. The present study aimed to evaluate the association between the use of fluoroquinolones or macrolides and bleeding events in patients with concomitant DOAC use. This was a nationwide cohort study including 19 288 users of DOACs in 2008‐2018 using information from Swedish national health registers. We compared the incidence of bleeding events associated with use of fluoroquinolones or macrolides using doxycycline as a negative control. Cox regression was used to calculate crude and adjusted hazard ratios (aHRs) in time windows of various length of follow‐up after the start of antibiotic use. The incidence rates for fluoroquinolones and macrolides ranged from 12 to 24 and from 12 to 53 bleeding events per 100 000 patients in the investigated time windows. The aHRs (95% confidence interval) for use of fluoroquinolones and macrolides were 1.29 (0.69‐2.44) and 2.60 (0.74‐9.08) at the concomitant window, 1.31 (0.84‐2.03) and 1.79 (0.75‐4.29) at 30 days, and 1.34 (0.99‐1.82) and 1.28 (0.62‐2.65) at 150 days, respectively. With regard to fluoroquinolones, the present study suggests that the risk of bleeding when combined with DOACs, if any, is small. Codispensation of macrolides in patients on DOACs was not associated with an increased risk of bleeding. However, due to the small number of macrolide users, the results must be interpreted with caution. [ABSTRACT FROM AUTHOR]