Association Between Aortic Aneurysm and Aortic Dissection With Fluoroquinolones Use in Patients With Urinary Tract Infections: A Population‐Based Cohort Study

Abstract

Background Fluoroquinolones are first‐line antibiotics recommended for the treatment of complicated urinary tract infections (UTIs), with frequent reports of adverse effects of aortic aneurysm (AA) and aortic dissection (AD). We examined whether fluoroquinolones can increase the risk of AA and AD in patients with UTIs in the Taiwanese population. Methods and Results We used the National Health Insurance Research Database to identify patients diagnosed with UTIs under single antibiotic treatment of fluoroquinolones and first‐, second‐, or third‐generation cephalosporins. An AA and AD diagnosis within a year constituted the study event. Multivariable analysis with a multiple Cox regression model was applied for comparing the hazard risk of AA and AD between fluoroquinolones and first‐ or second‐generation cephalosporins. Propensity score matching was performed to reduce the potential for bias caused by measured confounding variables. Among 1 249 944 selected patients with UTIs, 28 568 patients were assigned to each antibiotic group after propensity score matching. The incidence of AA and AD was not significantly different between the fluoroquinolones and first‐ or second‐generation cephalosporins (adjusted HR [aHR], 0.86 [95% CI, 0.59–1.27]). However, the mortality increased in the fluoroquinolones group (aHR, 1.10 [95% CI, 1.04–1.16]). Conclusions Compared with first‐ or second‐generation cephalosporins, fluoroquinolones were not associated with increased risk of AA and AD in patients with UTI. However, a significant risk of mortality was still found in patients treated with fluoroquinolones. The priority is to control infections with adequate antibiotics rather than exclude fluoroquinolones considering the risk of AA and AD for patients with UTI.

Do fluoroquinolones increase aortic aneurysm or dissection incidence and mortality? A systematic review and meta-analysis

“…We screened 1,434 abstracts and 38 full-text papers and included 13 studies ( 12 – 19 , 23 , 37 – 40 ) ( Figure 1 and Supplementary Table 1 ). A total of 11 studies ( 12 – 19 , 37 – 40 ) focused on the association of FQs with de novo AAD incidence, and only one study ( 23 ) investigated the association of FQs with all-cause mortality and aortic-specific mortality of patients with existing AAD.…”

“…We screened 1,434 abstracts and 38 full-text papers and included 13 studies ( 12 – 19 , 23 , 37 – 40 ) ( Figure 1 and Supplementary Table 1 ). A total of 11 studies ( 12 – 19 , 37 – 40 ) focused on the association of FQs with de novo AAD incidence, and only one study ( 23 ) investigated the association of FQs with all-cause mortality and aortic-specific mortality of patients with existing AAD. Six studies ( 12 , 15 , 17 , 19 , 23 , 40 ) were cohort studies, three ( 13 , 18 , 39 ) were nest case-control studies, and four ( 14 , 16 , 37 , 38 ) were self-controlled studies, including two case-time-control studies ( 14 , 16 ), one case-crossover studies ( 38 ), and one self-controlled case series studies ( 37 ).…”

“…A total of 11 studies ( 12 – 19 , 37 – 40 ) focused on the association of FQs with de novo AAD incidence, and only one study ( 23 ) investigated the association of FQs with all-cause mortality and aortic-specific mortality of patients with existing AAD. Six studies ( 12 , 15 , 17 , 19 , 23 , 40 ) were cohort studies, three ( 13 , 18 , 39 ) were nest case-control studies, and four ( 14 , 16 , 37 , 38 ) were self-controlled studies, including two case-time-control studies ( 14 , 16 ), one case-crossover studies ( 38 ), and one self-controlled case series studies ( 37 ). Five studies ( 13 , 14 , 18 , 23 , 40 ) were from Taiwan, three ( 17 , 19 , 37 ) from the United States, and the others were from Canada ( 12 ), Sweden ( 15 ), France ( 16 ), Denmark ( 38 ), and Korea ( 39 ).…”

“…Six studies ( 12 , 15 , 17 , 19 , 23 , 40 ) were cohort studies, three ( 13 , 18 , 39 ) were nest case-control studies, and four ( 14 , 16 , 37 , 38 ) were self-controlled studies, including two case-time-control studies ( 14 , 16 ), one case-crossover studies ( 38 ), and one self-controlled case series studies ( 37 ). Five studies ( 13 , 14 , 18 , 23 , 40 ) were from Taiwan, three ( 17 , 19 , 37 ) from the United States, and the others were from Canada ( 12 ), Sweden ( 15 ), France ( 16 ), Denmark ( 38 ), and Korea ( 39 ). Data used in all studies were obtained from various administrative healthcare databases across countries.…”

“…Data used in all studies were obtained from various administrative healthcare databases across countries. Seven studies ( 13 , 14 , 17 , 18 , 23 , 37 , 40 ) limited inclusion to patients aged 18 years or older of which two ( 18 , 23 ) applied an age limit of 20 years, four ( 15 , 19 , 38 , 39 ) included 40 years or older, and one ( 12 ) included 65 years or older. The definition of FQ exposure in all studies relied on the occurrence of a FQ prescription or reimbursement.…”