Project Summary/ Abstract Significance: Urinary tract infections (UTIs) are the most common infection during pregnancy. While these infections are typically minor, they are associated with an increased risk of preterm birth, low birth weight, pyelonephritis and sepsis. Despite this, there is insufficient evidence comparing the different antimicrobial treatment regimes. This lack of comparative effectiveness data is particularly concerning given the increasing prevalence of antimicrobial resistance in urinary pathogens. As a result, medical organizations have been unable to provide up to date guidance to patients and providers on best practices for treatment. Specific Aims: The proposed project will not only directly clarify the treatment options; it will also develop tools for future researchers to investigate similar questions. The first aim is to compare the safety and effectiveness of the most common antimicrobials used to treat UTIs during pregnancy, as well as to assess whether antimicrobial choice should be influenced by the patient’s symptomaticity. The second aim of this study is to develop and validate algorithms for the identification of UTIs during pregnancy. Successful development of these algorithms will allow researchers to identify UTIs in large health care data sources. Approach: In Aim 1, we will conduct a cohort study comparing the safety and effectiveness of the five most commonly prescribed antimicrobials for UTIs in US pregnancies. The data source for this study will be the University of North Carolina (UNC) Health System electronic health records (EHR). Approximately 14,000 people who received one of the five most common antimicrobial treatments will be compared on several maternal and perinatal outcomes including: preterm birth, low birthweight, pyelonephritis, spontaneous abortion, stillbirth, maternal sepsis, prescriptions for additional courses of antimicrobials, and antimicrobial switching. We will then examine heterogeneity in antimicrobial effectiveness and safety between people who are symptomatic versus asymptomatic. For Aim 2, we developed two algorithms containing diagnosis and procedure codes that we hypothesize will accurately identify symptomatic and asymptomatic UTIs. We will then manually review patient charts and assess how well the algorithm correctly classifies patients. Fellowship Information: The applicant is a PhD student in Epidemiology at UNC Chapel Hill and a predoctoral fellow at the UNC Center for Pharmacoepidemiology. Mr. Kahrs proposes a training plan that will equip him with the tools to launch a successful career in comparative effectiveness research for pregnant people. This training plan will take place in a thriving collaborative research environment, guided by an established team of interdisciplinary mentors. The outputs from this fellowship will not only help guide clinical care for pregnant people but will assist Mr. Kahrs in his journey towards becoming an independent investigator.