Sr. programmer analyst Department of Medicine, Division of Infectious Diseases, Washington University School of Medicine Saint Louis, United States
Background: Urinary tract infections (UTIs)—including asymptomatic bacteriuria and acute cystitis—are among the leading indications for antibiotics in pregnancy. Concern about the risk of birth defects in relation to certain antibiotics led the American College of Obstetricians and Gynecologists (ACOG) to offer guidance (since withdrawn) about avoiding sulfonamides (e.g., trimethoprim-sulfamethoxazole [TMP-SMX]) and nitrofurantoin in the first trimester, a critical time-period for organogenesis. Little is known about real-world antibiotic utilization patterns among pregnant people with UTI in the U.S.
Objectives: To characterize temporal trends in antibiotic utilization for the treatment of UTI during pregnancy, overall and by subgroups (trimester and geographic region).
Methods: Using the Merative MarketScan Commercial Database, we identified privately-insured, pregnant individuals aged 15–49 years diagnosed with UTI in the outpatient setting from 2007–2020 and dispensed a UTI-related oral antibiotic within 3 days of diagnosis. We identified pregnancy episodes via delivery codes including livebirth/stillbirth, and included the first eligible UTI per pregnancy. We excluded those with recent hospitalization; recent prior diagnosis of UTI, pyelonephritis, or kidney stones; or select health conditions (e.g., cancer, autoimmune conditions). We categorized antibiotics per treatment guidelines, including separate categories for recommended and non-recommended β-lactams due to differential efficacy. We used data visualization and joinpoint regression to characterize temporal trends in antibiotic utilization.
Results: Among 118,292 UTIs in 111,299 people, the distribution of antibiotic agents was 58% nitrofurantoin, 25% β-lactams (recommended), 6% ampicillin or amoxicillin-alone (non-recommended β-lactams), 5% TMP-SMX, 5% fluoroquinolones, and 1% other. From 2007 to 2020, β-lactam use increased from 18% to 44% with the most substantial increase from 2016–2020 (average annual percentage change [AAPC] for 2016–2020, 15.3% [95% CI 10.8%, 20.1%]). Nitrofurantoin decreased from 59% to 45% during the study period, almost entirely during 2017–2020 (AAPC for 2017–2020, -8.6% [95% CI -9.5%, -7.6%]). Antibiotic utilization trends were generally similar by trimester and geographic region, except nitrofurantoin use was slightly lower in the first trimester.
Conclusions: Antibiotic selection for UTI therapy during pregnancy changed dramatically over the study period, with substantial increases in β-lactam use and decreases in nitrofurantoin use. The most extreme changes coincided with the now-withdrawn 2017 ACOG committee opinion. Comparative safety studies are needed to inform treatment guidelines for UTI during pregnancy.
