Lead/Presenter: Fran Cunningham, VA Center for Medication Safety
All Authors: Aspinall SL (VA Center for Medication Safety, VA Center for Health Equity Research and Promotion), Sylvain N (VA Center for Medication Safety), Zhang R (VA Center for Medication Safety) Dong D (VA Center for Medication Safety) Echevarria K (VA Pharmacy Benefits Management Services) Glassman PA (VA Center for Medication Safety) Goetz M (VA Greater Los Angeles Healthcare System) Cunningham FE (VA Center for Medication Safety)
Objectives:
Use of fluoroquinolones has been associated with increased risk of rare but serious adverse drug events (ADEs). The Food and Drug Administration has released several warnings recommending use of alternative antibiotics when feasible. The objective of this active surveillance project was to evaluate risks of serious ADEs in outpatient Veterans treated with fluoroquinolones versus common alternatives (amoxicillin/amoxicillin-clavulanate and azithromycin).
Methods:
This was a retrospective cohort analysis of Veterans with new outpatient prescriptions for oral levofloxacin, ciprofloxacin, or moxifloxacin (Group 1), amoxicillin or amoxicillin/clavulanate (Group 2), or azithromycin (Group 3) between 6/1/2008 and 5/31/2018. Patients could have multiple prescriptions ?30 days apart. Primary outcomes were hospitalization or ED visit due to acute myocardial infarction (AMI), ventricular arrhythmia, aortic aneurysm/dissection, Achilles tendon rupture/tendinitis, or peripheral neuropathy over the 30 days following release of the prescription, and 30-day all-cause mortality. Cox proportional hazards models assessed the association between fluoroquinolones and ADEs versus the other antibiotics.
Results:
After propensity score matching for baseline demographics, comorbidities, antibiotic indication, and concomitant medications, there were 795,709 well matched outpatients in Groups 1 and 2 and 551,269 in Groups 1 and 3. Based on preliminary results, patients in Group 1 had a greater risk for aortic aneurysm/dissection compared with Group 2 (HR = 1.74; 95% CI 1.27-2.40) or Group 3 (HR = 2.17; 95% CI 1.41-3.34). Patients in Group 1 also had a greater risk for AMI versus Group 2 (HR = 1.52; 95% CI 1.35-1.71) or Group 3 (HR = 1.41; 95% CI 1.22-1.62) and a greater risk for 30-day all-cause mortality versus patients in Group 2 (HR = 2.28; 95% CI 2.19-2.37) or Group 3 (HR = 2.02; 95% CI 1.93-2.12). There was no significant difference in the risk of ventricular arrhythmia between Group 1 and Groups 2 (HR = 1.12; 95% CI 0.85-1.48) or 3 (HR = 1.33; 95% CI 0.92-1.90).
Implications:
We identified a possible signal of an increased risk of aortic aneurysm/dissection, AMI, and all-cause mortality with fluoroquinolone use versus amoxicillin/amoxicillin-clavulanate and azithromycin in Veterans. Our findings warrant further formal investigation.
Impacts:
Results of this project may be used to help develop guidance on fluoroquinolone prescribing to ensure the safe and appropriate use of these antibiotics.